FEDERAL COURT OF AUSTRALIA

 

Australian Gas Light Company (ACN 052 167 405) v Australian Competition & Consumer Commission (No 3) [2003] FCA 1525

TRADE PRACTICES – mergers and acquisitions – partial acquisition of electricity wholesaler by electricity retailer – whether likely to have effect of substantially lessening competition in a market – National Electricity Market – whether wholesale market in electricity and derivative contracts – whether separate wholesale markets in electricity and derivative contracts – geographic scope – temporal sub-markets – inter-regional supply constraints - retail markets for delivery of electricity to end users – market transactions governed by bidding, dispatch and pool pricing – auction system – pool price volatility – necessity for hedging through derivative contracts – acquisition of partial interest in generator creating natural hedge for retailer – whether retailer likely to reduce hedge contract cover – whether generator with greater exposure to spot price more likely to exercise market power increasing price bids – apprehended cascade of vertical integration – raised barriers to entry into wholesale and retail markets – market foreclosure – market characteristics – application by retailer for declaration of non-contravention of s 50 – onus of proof – ‘likely’ – ‘substantial lessening of competition’ – nature of threshold test – declaratory relief – discretionary factors – whether authorisation proper alternative – differences between administrative authorisation and judicial declaration – whether application for declaration confers illegitimate forensic advantage on applicant – whether deprives regulator of legitimate forensic advantage – declaration granted – subject to undertaking to Court previously offered to ACCC

 

Trade Practices Act 1974 (Cth) s 50, s 87B, s4, s 163A

National Electricity (South Australia) Act 1996 (SA) s 6

National Electricity (Victoria) Act 1997 (Vic) s 6

National Electricity (New South Wales) Act 1977 (NSW) s 6

Electricity - National Scheme (Queensland) Act 1997 (Qld) s 6

Electricity (National Scheme) Act 1997(ACT) s 5

Electricity Industry Act 1993 (Vic)

Electricity Industry Act 2000 (Vic)

Electricity Industry Legislation (Miscellaneous Amendments) Act 2000

Electricity Industry (Residual Provisions) Act 1993 (Vic)

Office of the Regulator-General Act 1994 (Vic)

Essential Services Commission Act 2001 (Vic)

Judiciary Act 1903(Cth) s 39B(1A)

Federal Court of Australia Act 1976 (Cth) s 21, s 23.

Electricity Industry (Prohibited Interests) Regulations 2003

 

 

Australian Gas Light Company (ACN 052 167 405) v Australian Competition & Consumer Commission (No 2) [2003] FCA 1229 cited

Re Queensland Co-operative Milling Association; Re Defiance Holdings Ltd (1976) 25 FLR 169 applied

Re Howard Smith Industries Pty Ltd (1977) 15 ALR 645 cited

Outboard Marine Australia Pty Ltd v Hecar Investments (No 6) Pty Ltd (1982) 44 ALR 667 cited

Broken Hill Pty Co Ltd v Trade Practices Tribunal (1980) 31 ALR 401 cited

Tilmanns Butcheries Pty Ltd v Australasian Meat Industry Employees Union (1979) 42 FLR 331 discussed

Global Sportsman Pty Ltd v Mirror Newspapers Pty Ltd (1984) 2 FCR 82 cited

News Limited v Australian Rugby League Football Ltd (1996) 64 FCR 410 cited

Munro Topple & Associates Pty Ltd v Institute of Chartered Accountants in Australia (2002) 122 FCR 110 discussed

Trade Practices Commission v Ansett Transport Industries (Operations) Pty Ltd (1978) 32 FLR 305 cited

Radio 2 UE Sydney Pty Ltd v Stereo FM Pty Ltd (1982) 62 FLR 437 cited

Trade Practices Commission v Australian Iron and Steel Pty Ltd (1990) 22 FCR 305 cited

Rural Press Limited v Australian Competition and Consumer Commission [2003] HCA 75 applied

Stirling Harbour Services Pty Ltd v Bunbury Port Authority (2000) ATPR 41-752 cited

Universal Music Australia Pty Ltd v Australian Competition and Consumer Commission (2003) 201 ALR 636 cited

Dandy Power Equipment Pty Ltd v Mercury Marine Pty Ltd (1982) 64 FLR 238 followed

Sellars v Adelaide Petroleum NL (1994) 179 CLR 332 cited

Norwest Refrigeration Services Pty Ltd v Bain Dawes (WA) Pty Ltd (1984) 157 CLR 149 cited

Whitehouse v Carlton Hotel Property Ltd ` (1987) 162 CLR 285 cited

Levine Clark [1962] NSWR 686 cited

Re Broadcasting Station 2GB Pty Ltd [1964-5] NSWR 1648

Berlei Hestia (NZ) Ltd v Fernyhough [1980] 2 NZLR 150

Abrasive Materials Pty Ltd v Australian Fused Materials Pty Ltd (1998) 16 ACLC 1172 cited

Singapore Airlines Ltd v Taprobane Tours WA Pty Ltd (1991) 33 FCR 158 cited

Queensland Wire Industries Pty Ltd v The Broken Hill Pty Co Ltd (1988-89) 167 CLR 177 cited

QIW Retailers Ltd v Davids Holdings Pty Ltd (No 3) (1993) 42 FCR 255 cited

Re Tooth Co Ltd and Tooheys Ltd (1979) 39 FLR 1 applied

Commonwealth v Stirling Nicholas Duty Free Pty Ltd (1972) 126 CLR 297 cited

Bass v Permanent Trustee Company Ltd (1999) 198 CLR 334 cited

THE AUSTRALIAN GAS LIGHT COMPANY (ACN 052 167 405 v AUSTRALIAN COMPETITION AND CONSUMER COMMISSION, GREAT ENERGY ALLIANCE CORPORATION PTY LIMITED (ACN 105 266 028) and GEAC OPERATIONS PTY LIMITED (ACN 105 367 888)

V880 OF 2003

 

FRENCH J

19 DECEMBER 2003

PERTH (Heard in Melbourne)

IN THE FEDERAL COURT OF AUSTRALIA
VICTORIA DISTRICT REGISTRY	V880 OF 2003

 

BETWEEN:	AUSTRALIAN GAS LIGHT COMPANY (ACN 052 167 405) APPLICANT
AND:	AUSTRALIAN COMPETITION AND CONSUMER COMMISSION FIRST RESPONDENT   GREAT ENERGY ALLIANCE CORPORATION PTY LIMITED (ACN 105 266 028) SECOND RESPONDENT   GEAC OPERATIONS PTY LIMITED (ACN 105 367 888) THIRD RESPONDENT

 

JUDGE:	FRENCH J
DATE OF ORDER:	19 DECEMBER 2003
WHERE MADE:	PERTH (Heard in Melbourne)

 

THE COURT ORDERS THAT:

 

1. Subject to the Australian Gas Light Company giving to the Court an undertaking in the terms annexed to these orders the Court HEREBY DECLARES that none of the following acquisitions of shares in bodies corporate would have the effect or be likely to have the effect of substantially lessening competition in a market in contravention of s 50 of the Trade Practices Act 1974 (Cth):

 

(a) the acquisition by the Australian Gas Light Company (AGL) of shares in Great Energy Alliance Corporation Pty Ltd (GEAC) pursuant to the GEAC Subscription Deed dated 3 July 2003;

 

(b) the acquisition by GEAC Operations Pty Ltd (GEAC OpCo), a wholly owned subsidiary of GEAC, of the Loy Yang Sale Shares (as that term is defined in the Statement of Claim);

 

(c) the acquisition by AGL of shares in GEAC pursuant to the GEAC Subscription Deed dated 3 July 2003 in combination with the acquisition by GEAC OpCo of the Loy Yang Sale Shares.

 

 

2. The undertaking referred to in Order 1 will be discharged upon written notice to the Court signed by both parties that the ACCC has accepted an undertaking from AGL by way of renewal or variation of or substitution for the undertaking last offered to it by AGL.

 

3. AGL has liberty to apply to the Court upon reasonable prior written notice to the ACCC to discharge or vary the undertaking.

 

4. The ACCC is to pay AGL’s costs of this application.

 

 

 

Note: Settlement and entry of orders is dealt with in Order 36 of the Federal Court Rules.

ANNEXURE TO ORDER

UNDERTAKING UPON WHICH DECLARATION IS MADE

 

1. BACKGROUND

 

1.1 A consortium comprising The Australian Gas Light Company Limited (AGL), The Tokyo Electric Power Company, Incorporated (TEPCO) and certain financial investors including the Commonwealth Bank of Australia (Financial Investors) propose to acquire the electricity generation business in Victoria currently conducted by the partners in the Loy Yang Power Partnership and known as Loy Yang A, with AGL holding a 35% interest, the Financial Investors holding a 30% interest and TEPCO holding a 35% interest in that business.

1.2 This Undertaking has been given to the Commission and/or the Court in order to facilitate the acquisition by the consortium referred to in clause 1.1 of Loy Yang A on the basis outlined in clause 1.1.

2. DEFINITIONS

 

Act means the Trade Practices Act 1974 (Cth).

AGL means The Australian Gas Light Company ABN 95052 167 405 and/or its Related Bodies Corporate.

Aggregate Information means any information:

(a) derived from Confidential Generator Information or Confidential Customer Information (such as a sum, average, statistical analysis, comparison or general qualitative description); but

(b) from which the underlying Confidential Generator Information or Confidential Customer Information cannot reasonably be derived or ascertained.

Business Day means a day other than a Saturday, Sunday or public holiday in Victoria.

Commission means the Australian Competition and Consumer Commission.

Confidential Customer Information means the details (including identity of each counter-party, price, term and volume) of any specific Customer Contracts with the Marketing Management Company, but excluding:

(a) information which is generally known; and

(b) AGL’s own Customer Contracts with the Marketing Management Company.

Court means the Federal Court of Australia.

Customer Contract means an electricity derivative contract or power purchase agreement:

(a) entered into by the Market Management Company; or

(b) considered, proposed or likely to be entered into by the Market Management Company.

Confidential Generator Information means:

(a) details (including identity of each counter-party, price, term and volume) of Dispatch and Marketing Activities; and

(b) details (including quantities, dates and times) of any reductions or expected reductions in the availability of the Loy Yang plant to less than the plant’s Registered Capacity,

but excluding information which is generally known.

Confidentiality Regime means the arrangements described in paragraph 3.6.

Consortium means the structure by which the Consortium Members hold interests in the Loy Yang Business, from time to time.

Consortium Members means The Australian Gas Light Company, Tokyo Electric Power Company and the Financial Investors, or any of their Related Bodies Corporate or successors.

Dispatch and Marketing Activities means:

(a) the determination and management of the scheduling of available capacity at the Loy Yang Plant;

(b) the determination and management of trading, dispatch and re-bidding and contracting strategies;

(c) the placement of dispatch offers and re-bids;

(d) entering into Customer Contracts; and

(e) regulatory policy and dealings with relevant economic and competition regulators (including any issues arising under the National Electricity Code or the Act) for the Loy Yang Business.

Economic Interest

(a) means interests in a company or partnership, including, shares, voting rights, rights to receive dividends, rights to receive other distributions of income or capital, rights to receive a share of proceeds on winding up; but

(b) excludes:

(i) rights to purchase the interest in the Loy Yang Business or the Loy Yang Assets of a Consortium Member seeking to divest its interest the exercise of which are subject to AGL obtaining approval (on a formal or informal basis) from the Commission or the Australian Competition Tribunal; and

(ii) any rights AGL has to prevent approval of decisions in respect of Permitted Matters.

Financial Investors means Commonwealth Bank of Australia or other equity investors in the Loy Yang Business.

Intermediary has the meaning given in the National Electricity Code.

Loy Yang Assets means the electricity generating plant and the coal mine used in the operation of the Loy Yang Business.

Loy Yang Business means the electricity generation business operating in Victoria, known as Loy Yang A.

Loy Yang Consortium Agreements means a Shareholders Agreement, a Partnership Agreement, and any other arrangements between the Consortium Members (or their Related Bodies Corporate) or between entities which manage the Loy Yang Business.

Loy Yang Plant means the electricity generating plant used in the operation of the Loy Yang Business.

Market Management Company means a company appointed in accordance with clause 3.2(a).

NEM means National Electricity Market.

Permitted Matters means the arrangements described in paragraph 3.4.

Related Bodies Corporate has the meaning given in sub-section 4A(5) of the Act.

Risk Management Policy has the meaning given to it in Appendix A.

3. UNDERTAKINGS

 

3.1 AGL undertakes that the combined Economic Interest of AGL in the Loy Yang Assets and/or the Loy Yang Business will not exceed 35%.

3.2 AGL undertakes that the Loy Yang Consortium Agreements, which it will enter into with the Consortium Members, will require the following arrangements regarding the governance of the Loy Yang Business to be put in place and maintained until these Undertakings cease:

(a) the Consortium Members will appoint or will procure the appointment of a company (Marketing Management Company) which will be solely responsible for undertaking Dispatch and Marketing Activities as the agent of the Consortium Members.

(b) AGL will be prohibited from having any Economic Interest in the Marketing Management Company. It also will not enter into any contracts, arrangements or understandings with the shareholders of the Marketing Management Company which would in effect confer on it such an Economic Interest.

(c) The terms upon which the Marketing Management Company is appointed in respect of the matters in paragraph (a) will include a requirement that any dealings between it and AGL are to be conducted at arms length.

(d) AGL will not participate in the appointment or supervision of the executive management of the Marketing Management Company.

(e) For the avoidance of doubt Dispatch and Marketing Activities which the Marketing Management Company undertakes as the agent of the Consortium Members will not give rise to a contravention of this undertaking provided that paragraphs 3.2(a) to (d) are otherwise complied with.

3.3 AGL undertakes not to be otherwise involved in:

(a) the Dispatch and Marketing Activities of the Loy Yang Business; or

(b) any board or management decision-making (including at the level at which AGL holds its Economic Interest in the Loy Yang Business or at the Marketing Management Company level) in respect of Dispatch and Marketing Activities,

in each case except in so far as the subject matter of the decision-making involves a Permitted Matter.

3.4 For the avoidance of doubt, the Consortium Agreements may provide that certain specific matters relating to the conduct of the Loy Yang Business require the agreement and/or participation of AGL. Those matters are (Permitted Matters):

(a) the financing and capital structure of the Loy Yang Business;

(b) variations to the structure by which the Consortium Members hold interests in the Loy Yang Business;

(c) expansion of the Loy Yang Business and acquisitions, construction and disposals of substantial assets;

(d) appointment or change of auditors for the Loy Yang Business;

(e) distribution policy of the Loy Yang Business;

(f) key corporate changes, including a merger, trade sale, initial public offering, dissolution, suspension or winding up of the Loy Yang Business or constitutional amendments in respect of the Loy Yang Business, except in circumstances where that event would case AGL to breach paragraphs 3.1, 3.2 and 3.3 of these Undertakings;

(g) annual budgets (including capital and operating expenditure) and approvals of expenditures and liabilities outside of budget of the Loy Yang Business;

(h) the Risk Management Policy of the Loy Yang Business;

(i) environmental policies, occupational health and safety policies and industrial relations policies of the Loy Yang Business; and

(j) commencing, defending or settling claims of the Loy Yang Business.

3.5 For the avoidance of doubt, AGL may hold an Economic Interest not exceeding 35% in a company which provides the following services to the Loy Yang Business:

(a) management, labour, engineering and other services for the physical day-to-day operation and maintenance of the Loy Yang Assets; and

(b) services in relation to the matters referred to in paragraph 3.4(i).

3.6 AGL undertakes that the Loy Yang Consortium Agreements will make provision for the adoption of, and compliance with, a Confidentiality Regime which prohibits AGL having access to:

(a) Confidential Customer Information; and

(b) Confidential Generator Information.

For the avoidance of doubt the Confidentiality Regime does not preclude AGL from having access to Aggregate Information reasonably necessary for AGL to assess compliance with policies regarding Permitted Matters, so long as AGL uses the Aggregate Information solely for the purpose of assessing compliance with policies regarding Permitted Matters.

3.7 Notwithstanding paragraph 3.6, if a company is appointed pursuant to paragraph 3.5 then:

(a) paragraph 3.6 does not apply in respect of information which falls within paragraph (b) of the definition of Confidential Generator Information provided to any director appointed by AGL to that company; but only if

(b) any such director has provided a written undertaking to not disclose to AGL, its officers, employees or agents any confidential information within paragraph (b) of the definition of Confidential Generator Information.

APPENDIX A TO AGL UNDERTAKING

 

RISK MANAGEMENT POLICY

 

 

The term Risk Management Policy, in the context of an electricity generation business, means a policy document which has the following characteristics:

1. It is a document which records a formal policy adopted by the highest governance body within a business entity (for example a Board of Directors or Partnership Committee).

2. Its purpose is to preserve the value of the business’s assets and ability to deliver budgeted outcomes by:

(a) setting global limits and controls on the business’s exposure to; and

(b) establishing the internal governance structure and management philosophy for managing,

specific categories of business risk.

3. The specific risk areas dealt with would include:

(a) major asset (physical) risk (for example risks of losses arising from poor maintenance on key productive asserts (turbines));

(b) trading risks (price, volume and credit risk) (for example risk of losses from or total contract portfolio exceeding available plant capacity or a disproportionate exposure to spot prices);

(c) operational risk (for example breakdown in human resources, processes or technology); and

(d) legal and compliance risk (for example poor contract management systems, absence of compliance with Trade Practices Act, Corporations Act or National Electricity Code).

4. The global limits and controls may be quantitative or qualitative in nature and would operate to limit or control the total level of risk (often measured in financial terms) that a specific business activity or division may incur. For example:

(a) trading risk (price and volume) may be subject to limits on the proportion of financial budget forecasts which may be put at risk as a consequence of contract portfolio exceeding available plant capacity or a disproportionate exposure to spot prices;

(b) credit risk (counterparty default risk) would be subject to limits on the financial exposure to counterparties of varying categories of credit worthiness; and

(c) legal contract risk would be subject to requirements on use of ISDA pro forma contracts and ISDA optional force majeure clauses in those contracts.

5. The internal governance structure and management philosophy for managing risks would provide for:

(a) organisational structures which segregated relevant functions (such as staff responsible for trading, transaction confirmation, settlements, risk management, accounting and financial reporting, and internal audit); and

(b) delegations of authorities regarding activities with exposure to specified risks and allocations of responsibility for risk management functions.

6. In relation to trading, credit and legal risk the Risk Management Policy would provide for the delegation to a Risk Management Committee responsibility:

(a) for overseeing risk management operations and procedures (within the parameters of the Risk Management Policy) of activities involving energy market exposure; and

(b) for setting individual trader limits.

7. A Risk Management Policy is not a Risk Management Framework. A Risk Management Framework, in the context of an electricity generation business, would:

(a) be approved by the Risk Management Committee; and

(b) specify the management procedures, rules and specific controls for implementing the Risk Management Policy.

Index

 

Introduction 1 - 10

Electricity – A Special Kind of Product 12 – 19

The National Electricity Market – History

and Origins 20– 49

Outline of the National Electricity Market 50 – 52

The Operation of the NEM – Generation of Electricity 53 - 66

The Operation of the NEM – Transmission 67 - 77

The Operation of the NEM – Distribution of Electricity 78 – 80

The Operation of the NEM – Retail Supply of

Electricity 81 – 95

The Operation of the NEM – Demand 96 – 101

The Operation of the NEM – Bidding, Pricing

and Dispatch 102 – 119

The Operation of the NEM – Spot Price Volatility 120 - 123

Operation of the NEM – Hedging Arrangements

And Electricity Derivative Contracts 124 - 129

The Operation of the NEM – Inter-regional

Hedging and Inter-regional Settlement Residue

Auctions 130 – 133

The Loy Yang Power Station and Coal Mine 134 - 137

The Loy Yang Power Station – Original

Acquisition and Financing – 1997 138 – 140

The Loy Yang Power Station – Existing

Security Arrangements 141 – 144

The Loy Yang Power Station – Events Leading

to the Acquisition Proposal 145 - 159

The Loy Yang Power Station – External Financing 160 – 162

AGL – Gas and Electricity Retailer 163 – 192

AGL’s Vertical Integration Strategy 193 – 216

AGL’s Development of an Acquisition Proposal 217 – 223

Overview of Transaction Documents 224 – 229

Details of Transaction Documents – The Share

Sale Agreement 230 - 231

Details of Transaction Documents – The GEAC

Shareholders Agreement 232 - 252

Details of the Transaction Documents – The MMCo

Agency Agreement 253 - 257

Details of the Transaction Documents – The MMCo

Operational Deed 258

Details of the Transaction Documents – The MM

HoldCo - Shareholders Deed 259 – 261

AGL Seeks Informal Clearance from the ACCC 262 - 269

The Commencement of the Proceedings 270 - 271

The Pleadings - Ownership and Acquisition 272 – 275

The Pleadings – The Electricity Industry in Victoria 276 – 279

The Pleadings – The Relevant Markets 280 – 281

The Pleadings – Competition in the Retail Markets 282 – 285

The Pleadings – Generation and Transmission 286 – 289

The Pleadings – Operation of the NEM –

Inter-Regional Pricing 290 – 294

The Pleadings – Supply, Demand and Pricing in the

Wholesale Market 295 – 298

The Pleadings – Electricity Derivative Contracts and

Spot Price Management 299 – 302

The Pleadings – AGL’s Post-acquisition Control of

Loy Yang 303 – 305

The Pleadings – The Section 50(3) Factors 306 – 307

The Relief Sought 308– 310

The ACCC’s Contentions About the Ways in which

The Proposed Acquisition Will Cause or be Likely

To Cause a Substantial Lessening of Competition 311

The Statutory Framework 312 – 319

A Brief History of Section 50 320 - 336

The Construction of Section 50 337 - 354

The Onus of Proof 335 – 356

The ACCC Case Summarised 357

Whether AGL Will Control or Influence LYP 358 - 376

Market Definition 377 - 387

Matters to be Considered Under Section 50(3) 388 - 402

Projections for Supply and Demand in the NEM 403 - 426

Market Power – Definition and Relevance 427 - 228

Market Power – The Contentions 429 - 431

The Loy Yang Power Station – Market Power

And Pricing in the Summer of 2000/2001 432 -456

Market and Regulatory Response to Pricing

In the Summer of 2000/2001 457 – 469

Pricing in the Wholesale Electricity Market and its

Relationship to the Longrun Marginal Costs of

Generation 470 –493

The Acquisition, the Natural Hedge and the Likelihood

Of Price Increases – The Economic Case 494 – 569

The Effect of the Acquisition Upon Vertical

Integration in the Relevant Market 570 – 594

Retail Markets in Victoria 595 – 598

Conclusions on Whether the Proposed Acquisition

Contravenes Section 50 599

The Grant of Relief – Discretionary Issues 600 – 612

The Form of Relief 613- 618

Annexures 1-8

IN THE FEDERAL COURT OF AUSTRALIA
VICTORIA DISTRICT REGISTRY	V880 OF 2003

 

BETWEEN:	AUSTRALIAN GAS LIGHT COMPANY (ACN 052 167 405) APPLICANT
AND:	AUSTRALIAN COMPETITION AND CONSUMER COMMISSION FIRST RESPONDENT   GREAT ENERGY ALLIANCE CORPORATION PTY LIMITED (ACN 105 266 028) SECOND RESPONDENT   GEAC OPERATIONS PTY LIMITED (ACN 105 367 888) THIRD RESPONDENT

 

JUDGE:	FRENCH J
DATE:	19 DECEMBER 2003
PLACE:	PERTH (Heard in Melbourne)

REASONS FOR JUDGMENT

Introduction

1                     The Australian Gas Light Company (AGL) is a major retailer of electricity in the electricity market. The Loy Yang A Power Station is a major generator of electricity located in the La Trobe Valley in Victoria. It generates electricity for supply to the National Electricity Market (NEM) which covers Victoria, New South Wales, Queensland, South Australia and the Australian Capital Territory. It is owned, together with the Loy Yang Coal Mine, by a consortium of companies which have been endeavouring to sell the business of the station and mine for some time as they are under pressure from external financiers. The operating company is Loy Yang Power Management Pty Ltd (LYPM).

2                     On 3 July 2003, AGL and other members of a consortium formed for the purpose, signed agreements for the acquisition of the Loy Yang Power Station Business (LYP) from its existing owners. Under those agreements AGL is to acquire a 35% interest in a holding company whose subsidiary is to acquire the shares in each of the companies operating the consortium.

3                     In February 2003, AGL approached the Australian Competition and Consumer Commission (ACCC) to seek an informal clearance in respect of the proposed acquisition. It offered an undertaking under s 87B of the Trade Practices Act 1974 (Cth) that certain structural arrangements reflected in the agreements would be maintained to prevent AGL from having any control or substantial influence over the way in which electricity from the power station would be made available, bid, dispatched and contracted on a day to day basis. The arrangements would also prevent AGL from being privy to commercially sensitive information about retail competitors and their dealings with LYP or about LYP’s bidding, dispatch and contracting strategies. Notwithstanding these undertakings the ACCC was of the view that the proposed acquisition raised substantial competition concerns pursuant to s 50 of the Trade Practices Act.

4                     On 5 September 2003, the ACCC wrote to AGL’s solicitors stating that if the acquisition were to proceed it would reserve its position as to what course of action it would take. It foreshadowed that the courses of action open to it would include an action for contravention of s 50 of the Act seeking the full range of available remedies including pecuniary penalties and divestiture. The ACCC published a press release to similar effect on 8 September 2003.

5                     On 15 September 2003, AGL commenced proceedings in this Court seeking declarations to the effect that the proposed acquisition would not contravene s 50. The trial was expedited and heard over twelve and a half sitting days in November. The acquisition is subject to fulfilment of a condition precedent, which may depend upon the outcome of these proceedings.

6                     There are complex issues of fact and law raised in the case. The central question is whether or not it could be said that the proposed acquisition would be likely to have the effect of substantially lessening competition in a market. The relevant markets under consideration are wholesale markets for the sale of electricity and/or electricity derivative contracts and retail markets for the sale of electricity.

7                     The competition issues in the case have focussed on two main questions:

1. Whether the acquisition of a 35% interest in LYP by AGL will provide it, as a retailer, with a ‘natural hedge’ which will lead to an equivalent reduction in the amount of protection it seeks against spot market price volatility by reducing the extent of its hedge contract cover. The ACCC contends that such a reduction would lead to a thinning of the hedge contracts market and a resulting incentive and tendency on the part of LYP and other generators of electricity to exercise market power and seek to lift prices in the spot market.

2. Whether the acquisition, as a form of partial vertical integration, would encourage further vertical integration by other market participants reducing the number of independent wholesalers and retailers, exposing remaining retailers to vertical market foreclosure and raising barriers to entry for new participants.

There are some related arguments that AGL will exercise some degree of influence or control over LYP, that in any event LYP will respond to AGL’s needs and/or that there will be risk sharing agreements entered into between AGL and LYP. There is also an argument that competition in the retail market could be reduced.

8                     The case raises questions of prediction and prognosis in a complex commercial environment subject to a regulated auction system for electricity across the NEM and different regulatory systems affecting pricing for smaller customers in the various States participating in the NEM.

9                     The proceedings have pitted leading economic and industry experts against each other, for the most part as advocates of the positions relied upon by the parties. The Court has been exposed to sophisticated econometric evidence seeking to mathematically model bidding and pricing behaviour in the relevant markets. It has been necessary, paying due respect to the amount of thought and expertise which underlies those contending positions, to keep firmly in mind that the Court’s decision must be based upon the best view it can form of the commercial realities of risks, incentives and behaviour and so on of the operation of the competitive process with and without the proposed acquisition.

10                  For the reasons which follow, I am satisfied that the proposed acquisition is not likely to have the effect of substantially lessening competition in any relevant market. That conclusion is reinforced by the structural arrangements which are the subject of the undertaking offered by AGL to the ACCC and which AGL has in turn offered to the Court. The declaration which I make is subject to that undertaking with certain amendments to delete provisions which imply ongoing monitoring by the Court or preset terms and conditions for its variation or discharge. I have framed the orders so that the undertaking given to the Court can be discharged upon the ACCC accepting a renewal of the undertaking originally offered to it by AGL or a variation or substituted undertaking which they can agree between themselves and which would give the ACCC an ongoing monitoring role. It may be in the interests of both parties to reach some such accommodation.

Electricity - A Special Kind of Product

11                  Reflecting the scientific understanding of the 18^th Century, Samuel Johnson’s Dictionary of the English Language (1755) defined electricity as ‘a property in some bodies, whereby, when rubbed so as to grow warm, they draw little bits of paper, or such like substances to them’. The New Oxford English Dictionary (1998) defines electricity as:

‘… a form of energy resulting from the existence of charged particles (such as electrons and protons) either statically as an accumulation of charge or dynamically as a current.’

For present purposes, and applying the second limb of the Oxford English Dictionary definition, it is sufficient to define electricity as energy transmitted by the flow of electrons through a conductor.

12                  The flow of current which constitutes the flow of electrical energy can be produced by the movement of a conductor across a magnetic field. The kinetic energy of that movement is converted into the electrical energy of the current flow. So mechanical energy may be converted to electrical energy which may in turn be transmitted through conductors over a distance and applied, by consumers of the energy, to the operation of a large variety of electrical appliances.

13                  The generators which produce electricity for household, business and industrial use do so essentially by rotating a coil of wire through a magnetic field. A current is induced in the wire and flows into an external transmission system. The rotation may be produced mechanically by steam, which is itself generated by burning coal or gas fuel. Other sources of mechanical energy can also be used to rotate a conductor. These include water used in hydro generators and wind in wind generators. Engines driven by petrol or diesel fuel may be used. In Tasmania there is a proposal for a wood waste generator. Electrical energy may also be produced by conversion of solar energy. That process does not operate upon the same physical principle as that used in the mechanically powered generators which provide the bulk of electrical energy.

14                  In the language of physics the rate at which energy is produced by a system or process is called power. The basic unit of measurement of electric power is the watt. A watt represents the production of one joule of energy per second. A common commercial unit of power is the Megawatt which comprises a million watts. The capacity of a generator of electricity may be measured by reference to its power output in Megawatts. The energy it delivers is measured by the product of that capacity and the time for which it operates at that capacity. So a generator with a one Megawatt capacity will deliver one Megawatt hour (MWh) of electrical energy for every hour it operates.

15                  Electrical energy can be measured and, being measurable, can be bought and sold according to the quantities delivered like any other product. However it has some important features which affect the ways in which it can be traded. It cannot be stored except to a limited, and for present purposes inconsequential, extent in batteries. Its supply must match demand. If it does not then the generation and transmission systems may become unstable and dangerous. It also has some characteristics of a fluid flow. Once the flow from a particular source passes into a common transmission system it cannot be distinguished from electricity fed by another source into that system. So it is not possible to identify electricity used by a consumer connected to the transmission system as originating from one or another generator.

16                  Australia makes heavy use of electrical energy for industrial, business and domestic purposes. The electrical supply industry comprises the activities involved in generating electrical energy, transporting it to customers for their use, and charging them for their consumption. Important functional elements of the industry are:

1. Generation – the production of electricity.

2. Transmission – the transmission of electrical energy across high voltage wires (capable of carrying more than 66 Kilo Volts) in networks designed for the bulk transfer of electricity at high voltages. Some very large end users such as aluminium smelters take electricity at high voltages directly from the transmission network.

3. Distribution – the low voltage transportation of electricity from transmission networks to most end-use customers.

4. Retail – the sale of electricity to customers and the provision of support and information services to them.

5. Interconnection – the connection of electrical supply systems in different regions by transmission lines passing from State to the other. Transmission lines taking electrical energy from one relevant State or region to another are referred to as interconnectors.

17                  The operators of generators are commonly regarded as the ‘wholesalers’ of electricity. Although properly an incident of the wholesale function, the transmission of electricity from generators in Australia is effected, for the most part, through transmission networks which are separately owned and which may include interconnectors to link one region to another. Electricity is sent from the transmission networks into distribution networks for ‘retail’ delivery to consumers. Distribution and retail sales may be carried on by the same or distinct operators. Importantly the so-called wholesale and retail functions are not reflective of physical delivery of energy from generators to retailers and from retailers to consumers. There can be no storage of electrical energy by retailers for distribution to end users. The electrical energy from the generator flows directly to the customer whether or not a retailer is involved. The designations ‘wholesale’ and ‘retail’ for the purposes of market analysis are best attributed to the financial arrangements and transactions between generators, retailers and end-users.

18                  This broad description is applicable to the electrical supply industry in New South Wales, Victoria, Queensland, South Australia and the Australian Capital Territory.

19                  Before embarking upon further detailed consideration of the way in which the electrical supply industry operates it is necessary to consider the recent history of its regulatory reform and privatisation and the creation of the NEM.

The National Electricity Market – History and Origins

20                  The electricity supply industry in Australia was largely in the hands of publicly owned utilities until the 1990s. Through them, State governments carried on the generation, transmission, distribution and retailing of electricity. Each State had its own separate electrical supply systems with only limited interconnection.

21                  An Industry Assistance Commission report delivered in 1989 following a reference from the Commonwealth Treasurer in 1988, described the gas and electricity industries as particularly inefficient – Industries Assistance Commission, Government (Non-Tax) Charges Vol 1 1989. In 1989 the newly formed Industry Commission was asked by the Commonwealth Treasurer to report on the institutional, regulatory or other arrangements subject to influence by governments which led to inefficient resource use in the electricity and gas sectors and to advise how such inefficiencies might be reduced or removed. The report, which was delivered in 1991, found an urgent need for reform of the electricity and gas sectors – Industry Commission, Energy Generation and Distribution Vol 1, 1999 p 2. The lack of commercial discipline imposed by competition was found to be the primary source of inefficiency. The Industry Commission’s recommendations for reform of the electricity sector underpinned its future development.

22                  At a special Premiers’ conference held in 1991, agreement was reached that a National Grid Management Council (NGMC) should be established to consider arrangements for an interstate electricity network. The Council was to prepare a draft protocol covering the planning, operation, development, monitoring and extension of the Eastern and Southern Australian Electricity Grid – Special Premiers’ Conference, Communiqué 30-31 July 1991. Following the establishment of the Council of Australian Governments (COAG) in May 1992, to initiate, develop and monitor the implementation of policy reforms of national significance, the NGMC was required to submit its recommendations to that body. In January 1993, the NGMC recommended the establishment of a competitive market in the trading of electricity. It made a number of recommendations about features of the national electricity market which included:

1. Direct customer to generator access.

2. Non-discriminatory access to the interconnected transmission network.

3. No barriers to interstate trade or to entry for new participants in generation or retail supply.

4. Uniform trading rules across South and Eastern Australian ESI

See National Grid Management Council, National Electricity Market and Common Trading Arrangements, An Information Paper, January 1993.

23                  The NGMC recommendations were followed a few months later by those of the Hilmer Inquiry. The Hilmer Inquiry was established as an independent committee of inquiry by the Prime Minister in October 1992 following agreement by COAG on the need for a National Competition Policy. Its objective was to develop a national framework for competition policy. Among its recommendations were:

1. The removal of immunity from the Trade Practices Act 1974 (Cth) of government owned businesses.

2. The structural reform of public monopolies which included:

. separation of regulatory and commercial functions of public monopolies;

. separation of natural monopoly and potentially competitive activities; and

. separation of potentially competitive activities into a number of smaller independent business units.

3. Access to essential facilities through the adoption of a new legal regime under which firms could be given the right of access to specified ‘essential facilities’ on fair and reasonable terms.

4. Price oversight for monopolies to deal with circumstances where all other competition policy reforms had proven inadequate.

5. The establishment of the National Competition Council (NCC) to oversee the proposed competition regime and the ACCC to undertake the regulatory functions previously performed by the Trade Practices Commission and the Prices Surveillance Authority.

24                  The reforms recommended by the Hilmer Inquiry were adopted in broad terms by COAG in April 1995. Three intergovernmental agreements emerged from that process, they being known as:

1. The Conduct Code Agreement

2. The Competition Principles Agreement

3. The Agreement to Implement the National Competition Policy and related reforms which included commitments to reform the electricity industry.

25                  In May 1996, New South Wales, Victoria, Queensland, South Australia and the Australian Capital Territory entered into an agreement known as the National Electricity Market Legislation Agreement under which each of the participating jurisdictions agreed to enact a National Electricity Law with South Australia as the lead jurisdiction. The National Electricity Law is a schedule to the National Electricity (South Australia) Act 1996 (SA). It is applied as a law of South Australia by s 6 of that Act. It has been applied and adopted as the law of Victoria by s 6 of the National Electricity (Victoria) Act 1997 (Vic). That section provides:

‘The National Electricity Law set out in the Schedule to the National Electricity (South Australia) Act 1996 of South Australia, as in force for the time being –

(a) applies as a law of Victoria;

(b) as so applying may be referred to as the National Electricity (Victoria) Law.’

Like provision is made in s 6 of the National Electricity (New South Wales) Act 1997 (NSW), s 6 of the Electricity – National Scheme (Queensland) Act 1997 (Qld) and s 5 of the Electricity (National Scheme) Act 1997 (ACT). The National Electricity Law provides in s 6 that the Ministers of the participating jurisdictions may approve a code of conduct called the National Electricity Code (the Code) as the initial Code for the purposes of the Law.

26                  The National Electricity Law refers to NECA, the National Electricity Code Administrator Limited and to NEMMCO, the National Electricity Market Management Company Limited. NECA is a company incorporated under the Corporations Act 2001 and limited by guarantee. Its members are the five participating jurisdictions in the NEM and the State of Tasmania. It is governed by a Members Agreement and its Memorandum and Articles of Association. Under cl 2.1 of the Members Agreement, which is reflected in cl 1.5.2 of the Code, it has a number of objectives, the principal objective being to supervise, administer and enforce the Code. It has objectives related to the collection and dissemination of information and the administration of the ongoing development of, and changes to, the Code in order to achieve market objectives. Clause 1.5.3 of the Code requires NECA to monitor and report on compliance with the Code and its adequacy, to enforce the Code, to establish procedures for dispute resolution concerning provisions of the Code and to manage changes to the Code.

27                  NEMMCO is described in cl 1.6.1 of the Code as ‘[t]he body corporate responsible for operating and administering the market in accordance with the Code.’ It is required under cl 1.6.3 to operate and administer the market in accordance with the provisions of Ch 3 of the Code, to maintain power system security, to undertake its co-ordination of power system planning responsibilities and to register Code participants.

28                  The provisions of the Code do not have the force of statute although the National Electricity Law provides for civil penalties for breaches of it. The Code was authorised by the ACCC under s 88 of the Trade Practices Act in 1997 and those parts of it dealing with transmission and distribution networks were accepted as an Access Code under s 44ZZAA of the Trade Practices Act.

29                  The National Electricity Market is the name for the institutional arrangements for trading electricity in Queensland, New South Wales, Victoria and South Australia. It is established by the operation of the National Electricity Law and the National Electricity Code. Section 9 of the National Electricity Law prohibits any person:

1. from owning, controlling or operating a generating, transmission or distribution system unless the person is registered as a Code Participant in accordance with the Code.

2. other than NECA and NEMMCO, from administering or operating a wholesale market for electricity generating units or loads.

3. purchasing electricity from NECA or NEMMCO unless the person is registered as a Code Participant.

The effect of the provision is that all electricity industry participants in participating jurisdictions are to participate in the NEM.

30                  The Code sets out rules to be observed by all Code Participants participating in the NEM. NECA administers the Code and enforces it in accordance with the National Electricity Law. Chapter 3 of the Code contains the rules governing the operation of the wholesale spot price market which is operated by NEMMCO. NEMMCO registers all Code Participants and operates the wholesale spot market through a centralised dispatch process.

31                  Each of the participating States and Territories of the NEM developed complementary reforms within their own borders. In Victoria, the government decided to privatise the electricity industry in conjunction with its competitive reforms. This resulted in the Electricity Industry Act 1993 (Vic). That Act, in effect:

1. Split the vertically integrated State Electricity Commission of Victoria into four functional activities for the purposes of privatisation. Those activities were generation, transmission, distribution and retail.

2. Entrenched the competitive reforms through the splitting up of the State Electricity Commission by introducing restrictive cross-ownership rules – Electricity Industry Act 1993 Part 13.

3. Introduced a licensing regime under which new privatised entities operating in the electricity supply industry in Victoria would be regulated.

32                  The Electricity Industry Act 1993 (Vic) provided for the transfer of the generating assets of the State Electricity Commission to an entity established under s 7 of the Act and called ‘Generation Victoria’. The transmission assets of the State Electricity Commission were transferred to an entity called ‘Power Net Victoria’ (previously known as ‘NationalElectricity’) set up under s 13 of the Act. A new entity called ‘Electricity Services Victoria’ took over the distribution assets of the State Electricity Commission under s 20. The assets, liabilities and staff of Generation Victoria and Electricity Services Victoria were transferred to various generation and distribution companies owned by the State of Victoria in anticipation of privatisation. A licensing regime under which privatised entities would be regulated was established under ss 159 to 169 of the 1993 Act.

33                  The licensing regime created by the Act prohibited the generation, transmission, distribution, supply or sale of electricity without a licence from the office of the Regulator-General or a licence exemption (s 159). Retail and distribution functions were identified and separately licensed (s 161). The State Electricity Commission had had only one point of contact with customers for distribution and retail functions and in order that billing systems and customer records did not need to be duplicated, combined distribution and retail businesses were formed.

34                  The Victorian transmission system, together with the Victorian portions of interconnectors linking Victoria to South Australia and Victoria to the Snowy Mountains Hydro-electric Scheme, were transferred to Power Net Victoria under the pre-privatisation arrangements effected by the 1993 Act. The transmission asset holder, Power Net Victoria, was privatised and sold to GPU Power Net pursuant to an agreement under s 153U of the Act. The Victorian Electricity Transmission System is now owned by SPI PowerNet. The assets that form part of the shared transmission network are operated by VEN Corp under a Network Services Agreement which SPI PowerNet is required to enter into under its transmission licence.

35                  Initially transmission prices were regulated by a Tariff Order made under s 158A of the 1993 Act. However since the Code came into operation and transitional provisions expired, transmission prices have been regulated under Chapter 6 of that Code as modified by Victorian Jurisdictional Derogations which are set out in Chapter 9 of the Code. They provide for an alternative pricing regime for VEN Corp. The obligations to connect generators, distributors and other parties to the transmission system are regulated by Chapter 5 of the Code.

36                  Following its creation, Generation Victoria’s assets were divided among five key generating companies which were privatised between 1996 and 1999. These were Loy Yang Power, Yallourn Energy, Hazelwood Power, EcoGen Energy and Southern Hydro. Privatisation agreements were made pursuant to s 153U of the Act. In 1994, a company called VPX was established under s 41A which had functions for Victoria analogous to those now undertaken by NEMMCO in the national energy market. VPX conducted a Victorian spot price market and determined, on the basis of generator bids, which generators should be dispatched, ie which generators were to supply electricity to the market from time to time. It was responsible for the reliable operation of the system as a whole and the transmission system in particular. Its spot market functions were subsumed by NEMMCO with the commencement of the NEM. Its electricity transmission system planning and shared electricity transmission network operation functions have been transferred to the State owned corporation, VEN Corp.

37                  As with the assets of Generation Victoria, those of Energy Services Victoria were divided among a number of distribution and retail entities which were progressively privatised during 1995, again pursuant to s 153U.

38                  From October 1994, customers were required to purchase electricity and distribution services from their local distributor. Victoria was notionally divided into five areas for this purpose. Government owned retailers responsible for each of those areas were Solaris Power, United Energy Limited, PowerCor Australia Limited, Eastern Energy and CitiPower Pty Ltd. In 1995, these retail businesses were sold by the Victorian government. Solaris Power is now AGL Electricity Ltd and Eastern Energy is now TXU Electricity Ltd. These companies today to provide distribution services in their allocated areas.

39                  Under the regulatory structure, customers were progressively enabled to choose which retailer would provide their energy even though they would continue to be connected to the Grid via their local distributor. The end product of this process was called ‘full retail contestability’. The timetable for establishing full retail contestability was set out in Regulations called The Electricity Industry (Non-Franchise Customers) Regulations 1995.

40                  Initially retail prices were regulated under a Tariff Order. It was evidently envisaged that following the introduction of contestability retail prices would not need to be regulated because there would be competition between the retailers. Prices for all larger customers, and those small customers who choose to enter into a contestable contract offered by a retailer are not regulated. Provision has been made however for certain minimum customer protections to be mandatory for those customers consuming less than 160MW hours annually.

41                  By 1997, consumers using more than 750MWh per year were allowed to chose between retailers. Other customers were assigned to an Incumbent Retailer based upon their geographic location. Each of the Incumbent Retailers and LYPM and the administrator of the State Electricity Commission of Victoria were party to transitional hedge contracts which corresponded to the volume of electricity by non-contestable customers. The contracts came into existence before the privatisations and their commercial terms were stipulated rather than negotiated. Along with like contracts between other generators and the Incumbent Retailers they were classified as Vesting Contracts. They enabled each of the Incumbent Retailers and the State Electricity Commission of Victoria to supply electricity with very limited exposure to the risk of pool price volatility. So far as LYPM was concerned, they provided a significant degree of certainty of revenue flow during the transitional period. At that time any retailer was able to supply contestable customers. They could be supplied with electricity on a hedged or unhedged basis. LYPM itself negotiated and entered into a number of hedge contracts with retailers for those customers.

42                  Mr Kenneth Thompson, the General Manager Marketing for LYPM, set out in his affidavit a timetable under which Victorian electricity customers became contestable. It was as follows:

Date	Customer Load Level	Number of Customers
December 1994	Customers with loads in excess of 5MW	47
July 1995	Customers with loads in excess of 1MW and less than 5MW	330
July 1996	Customers with loads in excess of 750MWh/Yr and less than 1MW	1,500
July 1998	Customers with loads in excess of 160MWh/Yr and less than 750MWh/Yr	5,000
December 2000*	All remaining customers, subject to there being no significant technical or economic constraints	1,957,300

* Subsequently deferred until 2001

This historical background and the relatively recent completion of the transition to full contestability gives some indication of the evolving character of the markets which are in issue in this case. The Court is not dealing with mature well-established markets occupied by participants who have had long experience in their operation.

43                  Following the completion of the privatisation process, a new Electricity Industry Act 2000 (Vic) was enacted. The regulatory and licensing regime established by the 1993 Act, including the cross-ownership provisions, were re-enacted by the Electricity Industry Act 2000 (Vic), albeit there were amendments to them. The 1993 Act was amended by the Electricity Industry Legislation (Miscellaneous Amendments) Act 2000 through the repeal of provisions relating to the privatisation process and provisions overlapping with the new Electricity Industry Act 2000. The 1993 Act was renamed the Electricity Industry (Residual Provisions) Act 1993 (Vic).

44                  The Victorian electricity industry is now principally regulated by the licensing regime established under the Electricity Industry Act 2000. The Essential Services Commission (the Commission) issues generation, transmission, distribution and retail licenses subject to various conditions. The Commission is the successor to the office of the Regulator-General and was established under the Essential Services Commission Act 2001 (Vic). The conditions imposed upon licenses vary according to their type. Certain aspects of distribution and retailing such as the quality and reliability of supply, complaints handling and customer billing (in the case of retailing) are also regulated by the Distribution Code and Retail Code. These have been developed by the Essential Services Commission and are enforced by the inclusion of conditions in distribution and retail licenses under s 21 of the Electricity Industry Act 2000, which require compliance with those Codes.

45                  Distribution prices initially were set by a Tariff Order and then by the office of the Regulator-General under the Office of the Regulator-General Act 1994 (Vic). Prices are currently set under a five-year determination of the former office of the Regulator-General. On the expiry of that determination, the Essential Services Commission will make a new determination. As to domestic and small business retail customers, prices continue to be regulated under the Electricity Industry Act 2000 (Vic). From January 2001, customers whose aggregate consumption exceeds 40MWh have been able to choose their electricity supplier under an Order in Council of 13 November 2000 pursuant to s 23 of the 2000 Act. Since 13 January 2002 domestic and small business consumers of electricity have been able to choose their electricity retailer pursuant to an Order in Council of 11 January 2002 under s 35 of the 2000 Act.

46                  In the present scheme, electricity retailers are required to make standing or deemed offers to domestic and small business customers whose usage does not exceed 160MWh/year. The retailer must make these standing or deemed orders to all domestic and small business customers within the designated geographic region allocated to that retailer. Retailers have obligations in relation to their standing or deemed offers. They are required to publish the applicable terms and conditions, including any tariffs, and those terms and conditions must not be inconsistent with the Electricity Retail Code. Any proposed changes to the terms and conditions cannot take effect for at least two months from the time of publication. A retailer may propose an increase at any time. The Governor in Council, who acts on the advice of the Minister, has a reserve power to fix tariffs applicable under standing or deemed offers which he gets from s 13 of the 2000 Act. The Governor in Council can revise tariffs up or down as he or she thinks fit.

47                  The Essential Services Commission is required to investigate and report to the Minister on any proposed increase in standing tariffs by electricity retailers. This requirement is imposed under a Standing Order. Benchmarks have been established to indicate the costs of key components for the retailing of electricity and appropriate margins. Ordinarily the publication of a proposed increase by a retailer triggers an investigation by the Essential Services Commission which may prompt an exercise of the Minister’s reserve power to fix tariffs. The Minister is not obliged to act on the Commission’s report.

48                  An important variation to the cross-ownership restrictions effected by the 2000 Act is that cross-ownership may be permitted if the ACCC formally or informally approves the transaction leading to the cross-ownership. Under the Electricity Industry (Prohibited Interests) Regulations 2003 certain prescribed interests can be disregarded when determining whether there is a prohibited interest for the purposes of s 68 of the Act. Under the Regulations, retailers can have a minority but significant shareholding of up to 35% in one significant generator where that retailer does not hold a relevant interest in the shares of two or more corporations, each of which is entitled to generating capacity of more than 200MW.

49                  The preceding description of the privatisation of the electricity industry in Victoria, which I find as a fact, reflects the content of an outline provided to the Court by AGL and evidence from Mr Kevin Thompson of LYPM which was not in dispute in these proceedings so far as it described those matters. Similar restructuring has occurred in other States in the NEM although the privatisation of the restructured businesses has been largely confined to Victoria and South Australia.

Outline of the National Electricity Market

50                  The NEM commenced operation on 13 December 1998. It incorporates the States of New South Wales, Victoria, Queensland and South Australia and the Australian Capital Territory. Although Tasmania is not presently in the NEM there are plans for it to be included following the construction of a link to the mainland known as Basslink. Western Australia and the Northern Territory operate separate electricity systems independent of South East Australia and are not part of the NEM. The following description of the operation of the NEM is based, for the most part, on the evidence of Dr Daniel Price, the Managing Director of Frontier Economics Pty Ltd, who provided expert evidence for AGL in these proceedings. His full account of the NEM was contained in a report entitled ‘Outline of the National Electricity Market’ which was dated 10 October 2003. Mr Greg Denton, an expert witness called by the ACCC, regarded Dr Price’s description as ‘sufficiently accurate’ save for his failure to mention what he called the ‘ancillary services market’. The descriptive elements of Dr Price’s outline were not in contention between the parties and the summary that follows, supplemented by references to the evidence of Mr Denton and Mr Nethercote, constitutes findings of fact about the operation of the market unless otherwise indicated.

51                  The NEM is governed by the Code which contains the rules for its operation. NEMMCO has responsibility for the day-to-day operation and administration of the power system and the spot price market. It also publishes an annual Statement of Opportunities which provide technical and marketing data and which projects supply and demand in the NEM.

52                  The NEM consists of five interconnected regions. A region is defined in cl 3.5.1 of the Code as ‘… an area served by a particular part of the transmission network containing one or more major load centres or generation centres or both’. The regions at the present time are Queensland, NSW, the Snowy region, Victoria and South Australia. Interconnectors, which are the transmission lines crossing between the different regions, allow generators in those regions to compete against each other.

Operation of the NEM – Generation of Electricity

53                  Generator output and capacity can be measured at the generator terminal before the power station’s own electricity requirements and transformer losses are taken into account. The capacity so measured is known as ‘name plate capacity’. The energy measured at the generator terminal is called ‘generated’ output. Output and capacity can also be measured at the transmission connection point after the diversion of some electricity to the power station’s auxiliary load and losses from the generator transformer is taken into account. Capacity measured at the transmission connection point is known as ‘sent out’ capacity. Energy measured at this point is known as ‘sent out energy’.

54                  Another important attribute of a generator is its capacity factor. This measures the actual generation as a proportion of potential generation. The lower the variable costs of production, then the higher the capacity factor will be.

55                  Generators can be classified according to their capacity factor. There are three important classifications:

1. Base load generators. These usually run with a capacity factor above 75% and typically have:

(i) low variable operating costs;

(ii) high capital costs;

(iii) long start up times; and

(iv) a limited stable range of operation.

Base load plants are generally operated continuously for long periods at or near full capacity. The Loy Yang A power station is a base load generator.

2. Mid merit or intermediate plants. These ordinarily have a capacity factor between 15% and 75%. Compared to a base load plant they typically have:

(i) medium operating costs;

(ii) short start up times; and

(iii) medium ramp rates. The ramp rate is the rate at which a power plant may increase or decrease the level of electricity which it generates and is explained more fully below.

Mid merit or intermediate plants usually stop generating during daily low demand troughs and do not operate to full load except during daily demand peaks.

3. Peaking plants. These operate with a capacity factor below 15%. Compared to base load plants, they typically have:

(i) relatively high operating costs;

(ii) higher ramp rates;

(iii) fast start up times.

They are generally operated to meet peaks in demand. They run infrequently and have a relatively low level of capacity utilisation. Open cycle gas fired plants are typically peaking generators.

These classifications also appeared in the evidence of the Chief Executive Officer of LYPM, Mr Ian Nethercote. He described the attribute of a generator known as its ‘ramp rate’ and the event called ‘outage’.

56                  The rate at which a power plant can vary the level of electricity which it generates is known in the industry as the ramp rate. Each of the four generators at Loy Yang A has a ramp rate capability of plus or minus 8 to 10MWs/minute. Each of the generators can only vary the level of electricity it generates with some certainty by up to 8 to 10MW per minute. This means that the Loy Yang A generator has a limited capacity to respond to significant or large sudden changes in market conditions, including demand and supply imbalances. Other types of power plants have a high ramp rate capability. A confidential table of ramp rates for other generators in the NEM was exhibited to Mr Nethercote’s affidavit. As appears from that table, by far and away the highest ramp rates are associated with hydro-powered stations.

57                  The unavailability of a generator for the production of electricity is known in the industry as an ‘outage’. There are planned outages necessary for maintenance and testing requirements. Unplanned outages occur generally because of plant breakdown. There are also environmental constraints on the station’s generation capability. Emission limits imposed by the Environmental Protection Authority may result in generation being reduced from time to time to ensure that particular emissions such as sulphur are within legal limits.

58                  Chapter 2 of the Code sets out and describes various categories of Code Participants and registration procedures. Code Participants are those bound by the Code through registration with NEMMCO, save for NEMMCO itself which is also a Code Participant (cl 2.1.2). Every person who owns, controls or operates a generating unit that supplies electricity to a transmission or distribution system must register with NEMMCO (cl 2.2.1). To register as a generator a person must classify each of its generating units as scheduled or non-scheduled (cl 2.2.1(e)). Generally speaking, a generating unit or group of units with a nameplate capacity of 30MW or more is classified as a scheduled generating unit (cl 2.2.2(b)). A generating unit with a lesser capacity is classified as a non-scheduled generating unit (cl 2.2.3(a)). A generating unit from which sent out electricity is not purchased in its entirety by the Local Retailer or a customer at the same connection point is a market generating unit (cl 2.2.4(a)). That is to say it is a generating unit which sells electricity into the spot market and by virtue of the Code must sell all its electricity into that market. A non-market generating unit sells all of its output to the local retailer or a customer outside the spot market (cl 2.2.5(a)).

59                  If a market generator, in respect of a generating unit, wishes to use it to provide market ancillary services then the market generator must apply to NEMMCO for approval to classify the unit as an ancillary services generating unit.

60                  The market generator must sell all sent out electricity through the spot market and accept payments from NEMMCO for sent out electricity at the spot price available at the connection point as determined for each trading interval in accordance with the provisions of Ch 3 of the Code (cl 2.2.4(c)).

61                  Most large generators in the NEM, such as the Loy Yang A power station, are scheduled market generating units which means that all of their output is sold through the NEM and NEMMCO schedules the dispatch of their electricity. The object of these classifications is to ensure secure and reliable operation of the power system by ensuring that all large generators are centrally dispatched by NEMMCO and to avoid imposing the costs of complying with NEM arrangements on small generation projects. Often non-market, non-scheduled generators are too small to make participation in the NEM economic.

62                  Generation capacity in each of the NEM regions as at August 2003 was as follows:

Scheduled Generating Capacity in the NEM
Region	Generation Capacity (MW)
NSW	12,241
Qld	10,063
Vic	8,326
Snowy	3,676
SA	3,463
TOTAL	37,769

 

Non-scheduled generating capacity was as follows:

Non-Scheduled Generating Capacity in the NEM
Region	Generation Capacity (MW)
NSW	309.80
Vic	221.65
Qld	144.00
SA	54.50
Snowy	0.00
Total	729.95

There are approximately thirty-two registered participants with scheduled generators in the five regions of the NEM. They operate sixty-seven generators of different kinds and capacities between them. A table showing the registered participants, their generators and information about those generators, including their registered capacities and their capacity factors, is set out as Annexure 1 to these reasons.

63                  In the NEM at present LYP has 5.3% of total capacity and generated 8.64% of total generation in the 2002 calendar year. AGL Electricity Pty Ltd has 1.01% of total capacity and generated 0.01% in the 2002 calendar year. However, in the Victorian region LYP has 18.79% of total capacity and generated 29.89% of all electricity in the 2002 calendar year. AGL Electricity Pty Ltd has 1.5% of total capacity and generated 0.01% of electricity in the 2002 calendar year. Annexure 2 sets out the Shares of Scheduled Registered Capacity and Energy in the NEM (at the generator terminals) in one table and similar figures for Victoria in a second table.

64                  As will be seen from the list of scheduled generators in the NEM, the major Victorian generating companies operate one generating outlet each. When the New South Wales Electricity Industry was restructured, generating portfolios consisting of a number of power stations were created. Victoria adopted a different model. Most generating companies created at the time of reform in Victoria owned a single power station. The largest generation portfolios in the NEM are Macquarie Generation with a total capacity of 4,690MW and Delta Electricity with 4,240MW. The Snowy Hydro follows with 3,756MW. Loy Yang Power A is the largest generator in Victoria having 2,000 MW of capacity. There are several companies that have generators in different regions in the NEM. Purchasers and builders of individual power stations in South Australia and Victoria have combined assets to form ‘inter-regional portfolios’. International Power owns generation capacity in both South Australia and Victoria.

65                  Overall the New South Wales generating portfolios have the largest share of NEM capacity. Macquarie Generation has 12.4%. Delta has 11.2%. Loy Yang Power A has a 5.3% share of NEM capacity. It is the tenth largest generator in the NEM. When energy output rather than capacity is calculated, base load generators which produce most of the energy show a relatively high share while peaking generators show a lower share. So although Loy Yang Power A has a 5.3% share of capacity in the NEM, it has an 8.6% share of total energy output. This makes it the third largest generator in the NEM by energy. When Tasmania joins the NEM through the Basslink connection it will comprise a sixth region and will contribute 2,514MW of mostly hydro generating capacity. All the Tasmanian capacity is presently owned by Hydro Tasmania which is in turn owned by the State Government.

66                 Different generators use different types of fuels and the variable cost of electricity generation depends upon the cost of the fuel used and the efficiency with which energy can be extracted from it. Although generation fuel costs are largely fixed, generators often need to purchase some additional amount of fuel outside any long term take or pay contracts. Such purchases are variable costs because their volume and timing is at the discretion of the generator and depends upon its output. Coal generators in the NEM have a lower variable cost than gas-fired plants and usually have a higher capacity factor. Victorian brown coal is cheaper than New South Wales and Queensland black coal. Gas-fired generation in the NEM is more expensive than coal-fired generation and therefore usually operates with lower capacity factors. Hydro plant has low variable costs but its energy input is limited by the availability of water and the inability to produce electricity at high capacity factors. For this reason, hydro plants operate as mid merit or peaking plants rather than as base load plants. Liquid fuel-fired generators, powered by fuel such as oil and distillate, are the most expensive plants in the NEM.

Operation of the NEM – Transmission

67                  The transmission of electricity in the NEM involves the utilisation of transmission networks within NEM regions and interconnectors between those regions. The Code identifies two types of transmission companies, namely Transmission Network Service Providers (TNSPs) and Market Network Service Providers (MNSPs). TNSPs own, operate and/or control high voltage transmission assets that carry electricity between generators and distribution networks. The revenue earned by these providers is regulated. MNSPs are unregulated interconnectors between two regions of the NEM and they earn revenue from trading in the NEM.

68                  Regulated interconnectors transport power from one region in the NEM to another region. The flow of energy between regions through these interconnectors will depend upon regional price differences. So regions of relatively cheap generation will be net exporters of power. Regulated interconnectors are owned and operated by TNSPs. The unregulated interconnectors are designed to take advantage of price differentials between regions and to earn revenue by arbitraging them. The flows through those interconnectors are determined by the dispatch bids which they submit to NEMMCO. If their bids are higher than price differentials between the regions they connect, then they will not be dispatched. As with generators, the bidding behaviour of unregulated interconnectors depends on market conditions. The process of bidding for dispatch of electricity is discussed below.

69                  All five regions in the NEM are connected by at least one regulated interconnector. There are two unregulated interconnectors. One is Directlink, between New South Wales and Queensland. The other is Murraylink, between Victoria and South Australia. It has recently applied to the ACCC to convert to a regulated interconnector.

70                  As the following table prepared by Dr Price shows there is substantial transfer capacity between New South Wales and Victoria via the Snowy. There is relatively less transfer capacity available between New South Wales and Queensland and between Victoria and South Australia.

Table: Inter regional transfer capabilities

 

Regions Capability

 

Exporting	Importing	880 MW (700MW QNI + 180MW Directlink)
NSW	Qld
Qld	NSW	1130 MW (950MW QNI + 180MW Directlink)
Snowy	NSW	3200 MW winter 2800 MW in summer
NSW	Snowy	850MW (varies from 200 MW to 1200 MW)
Snowy	Vic	1900 MW
Vic	Snowy	1100 MW
Vic	SA	680 MW (460MW Heywood + 220MW Murraylink)
SA	Vic	420 MW (300 MW Heywood + 120MW Murraylink)

This table is based upon information in the NEMMCO Statement of Opportunities 2003, Chapter 4 at p 4-8.

71                  The term ‘network capability’ is defined in the NEMMCO Statement as the technical capability of the network to transfer power measured in MW. To operate a network in excess of its capability can cause damage to plant or interruption of supply. There are therefore power flow limits imposed on the transmission network through network limit equations which are developed by TNSPs. NEMMCO has a due diligence process which is designed to ensure that these equations define secure power system operating conditions. NEMMCO converts the TNSPs’ network limit equations into a form called ‘network constraint equations’ that can be used in the NEMMCO dispatch optimisation process.

72                  The Code distinguishes between inter-and intra-regional constraints or limits. However intra-regional limits may reduce the capacity of an interconnector and therefore impact upon inter-regional constraints. As a general proposition the capability of the network to transmit power from generation centres to demand centres is limited by a number of factors relevant to the physical limits of specific plant and their stable operation. These factors include thermal limits of the transmission network, voltage levels on generation or transmission equipment, the ability of the power system to remain secure and stable, power system frequency control, and limits on secondary components of the power system including protection systems and monitoring equipment. The network capability therefore describes the power flow limit which can be permitted on the transmission lines and takes into account contingencies which could occur which could cause a specific limit to be reached.

73                  The network capability also necessarily affects the amount of electrical energy that can be transferred across interconnectors from one region to another. The transfer limits between regions can vary throughout the day and the year. The limit of each interconnector depends upon a number of factors including demand in the importing and exporting regions, generator output, the number of units in service at a power station, flows on other interconnectors and intra-regional demand and flows. So the inter-regional transfer capability between Snowy and Victoria depends upon the level of Snowy generation, demand in Southern New South Wales and Broken Hill and the status of Snowy generating units.

74                  When the flow on an interconnector reaches the limit of the inter-regional transfer capability, the interconnector is said to be ‘constrained’. The constraint is in effect imposed by the centralised spot price determination and dispatch process discussed later in these reasons. It means that beyond the level of constraint the regions may operate, to some degree, independently for pricing purposes. Where a transmission constraint has been reached generators within the importing region may be able to sell at a price greater than that in the exporting region. So instead of a single spot price operating across the interconnected New South Wales and Victorian system there will be separate prices for each State. The price differences caused by transmission constraints can constitute price signals or guides which show the value of greater interconnection between the regions. Price differences can also be a guide to the value of demand reduction in the importing region and/or additional generating capacity in various regions.

75                  When the NEM was created it linked a number of electricity supply industries which had previously operated independently. Interconnection between the regions has been strengthened in recent years reducing the incidence of constraints. The figures given by Dr Price for the 2002 calendar year did not take into account the commissioning of Murraylink in October 2002 and the augmentation of the Snowy/Victoria interconnect in December 2002.

76                  The prices which the TNSPs can charge retailers, direct customers and generators for the provision of their network services are known as ‘Transmission Use of System’ charges (TUoS) and these are regulated by the ACCC. The ACCC sets a maximum revenue stream for each TNSP for a number of years, which is known as a regulatory period. Under cl 6.2.4 of the Code these regulatory periods must be at least five years. In setting the maximum revenue stream the ACCC is required to take into account the asset base, capital depreciation, capital expenditure and operations and maintenance expenditure of each of the providers. Mechanisms offering customer savings on regulated TUoS charges are possible under the Code. Market participants connected to the network can negotiate discounts with providers if they can demonstrate cost savings by feasibly ‘bypassing’ the transmission network. Generators located within a distribution network close to customers are entitled to a TUoS rebate from the local distributor within whose area the generator is located if they can demonstrate investment savings in the distribution network that they would have incurred had they not located close to customers.

77                  Mechanisms for the expansion of the capacity of the transmission network depend upon whether the investment in such expansion is to be regulated or unregulated. Regulated interconnector augmentation as well as augmentations to the transmission system within a region are assessed under the ACCC’s regulatory test specified in the Code. Such augmentations are only approved if they maximise net benefits compared to a range of alternatives. If approved, the cost of these investments, including a return on capital, may be reflected in TUoS charges. The augmentation of unregulated interconnectors may be affected by market participants in much the same way as generation investment as a response to inter-regional price signals. There is however no scope for unregulated intra-regional transmission augmentation in the NEM.

The Operation of the NEM – Distribution of Electricity

78                  Distribution refers to the delivery of electricity across low voltage networks in the NEM. Companies that operate distribution networks in the NEM are registered as Distribution Network Service Providers (DNSPs). There are currently 13 registered DNSPs across the NEM. The following table, prepared by Dr Price, sets out the registered service providers in each region of the NEM, the number of customers each serves and whether each such distributor is ‘stapled’ to a retail business:

Distribution companies in the NEM

 

DNSP

Number of customers served

Stapled to a retailer?

NSW

 

EnergyAustralia	1,400,000	Yes
IntegralEnergy	790,000	Yes
Australian Inland Energy and Water	20,000	Yes
Country Energy	750,000	Yes

ACT

ActewAGL

130,000

Yes

Victoria

AGL Electricity	261,000	Yes
CitiPower	269,000	No
Powercor Australia	600,000	No
TXU	533,000	Yes
United Energy	583,000	No

Queensland

Energex	1,100,000	Yes
Ergon Energy	560,000	Yes

South Australia

ETSA Utilities

755,000

No

79                  Since the NEM commenced, a number of stapled distribution and retail businesses have been separated to create independent distribution-only and retail-only businesses. There has also been a number of distribution company mergers in New South Wales and Queensland to take advantage of costs savings.

80                  State-based authorities regulate distribution activities in each State in accordance with a framework set out in the Code. Distribution revenue is regulated along lines similar to those applicable to transmission revenue. Although there are variations between States, in general their regulators periodically determine the maximum revenue for DNSPs having regard to the same criteria as the ACCC applies to TNSPs. With each revenue reset, which is usually every five years, the parameters are reassessed and the provider assets can be revalued up or written down depending on their utilisation.

The Operation of the NEM – Retail Supply of Electricity

81                  The introduction of retail competition has been an important aspect of reform in the retail sector of the electricity industry. The retail function in the NEM does not refer to any underlying physical delivery of electricity from retailer to customer. The electricity flows from generators through transmission and distribution lines to the end users. That flow is not controlled by the retailer. The retail function rather describes the assumption, by the retailer, of liabilities to the generator in respect of electricity for which the retailer is paid by the consumer. As noted earlier, in relation to the development of contestability in the NEM, consumers can use the electricity retailer of their choice and are not required to purchase electricity from a retailer with supply responsibility in a particular geographical area. Full retail competition has been introduced in New South Wales, Victoria, South Australia and the ACT. For the present Queensland will not introduce full retail competition for customers with an electricity usage of less than 200MWh per annum or electricity bills of less than $20,000 per annum.

82                  In Victoria consumers who use less than 160MWh per annum can negotiate a contract with a retailer or remain on a regulated tariff. The State Government has established ‘deemed’ tariffs and ‘standing offers’ for this class of consumer as ‘safety net tariffs’. The consumer who negotiates a contract with a retailer cannot return to a regulated tariff. The Essential Services Commission reviews annual price increases proposed by retailers for the tariffs to ensure they are not charging monopoly prices. The price oversight role will cease when competition for small consumers is considered effective.

83                  Safety net tariffs in South Australia are provided for consumers using less than 160MWh per annum. Under the Electricity Act 1996 (SA), AGL in South Australia is required to establish a ‘standing contract price’. Consumers there can choose to negotiate a contract with a retailer and return to the standing contract price provided by the franchise retailer. The Minister can refer the published prices to the Essential Services Commission of South Australia for inquiry.

84                  Consumers using less than 160MWh per annum are also protected by a regulated retail tariff in New South Wales. They can however choose to negotiate a contract with a retailer and opt back to the regulated retail tariff at any time.

85                  There is a safety net regulated tariff in the ACT. The tariff structure there is designed to end in 2006 and applies to customers using less than 100MWh per annum. ACT customers are free to negotiate a contract with a retailer and elect to return to the regulated tariff at any stage provided by the franchise retailer at least until the end of the transitional period in 2006.

86                  Retail prices for small consumers are still regulated in Queensland.

87                  In general, the maximum retail tariffs which are set periodically by regulators take into account energy purchasing costs based on a benchmark energy purchase price, the costs of meeting statutory greenhouse gas reduction targets, retail operating costs including billing and revenue collection, the retailer’s profit margin, network charges, losses and market fees.

88                  The steps necessary to establish a retail electricity business were described comprehensively by Andrew Bonwick, Chief Executive Officer and Managing Director of Australian Energy Services Pty Ltd. That company operates as a retailer in the NEM under the name Powerdirect. Prior to assuming his position with the company Mr Bonwick was Director, Sales and Marketing at Yallourn Energy, which is a Victorian electricity generator and retailer trading under the name AusPower. The steps described by Mr Bonwick were as follows:

1. Raise finance for the project by attracting investment in a business proposal which will, inter alia, articulate the target market.

2. Establish a physical base for the business. This primarily comprises employees and infrastructure. The most significant infrastructure aspects are billing and marketing systems. In the initial stages billing systems can be established through predominantly small-scale personal computer generic software programs or outsourced to one of a number of external providers in Australia or overseas. Marketing functions such as telemarketing or direct marketing can be similarly outsourced. Outsourcing of telemarketing requires no geographical nexus between retailer and contractor and may be contracted to marketers also involved in marketing in industries other than energy.

3. Establish a regulatory infrastructure. This involves obtaining all necessary licences and approvals such as a retail licence from the Essential Services Commission of Victoria. The process can take up to six months and constitutes the longest lead-time for any aspect of retailer start up. Tactical capability must also be established. This involves establishing call centre practices, eg through compliance programs, to minimise regulatory risk and avoid potential breaches of licence conditions.

4. Procurement of sufficient quantities of green energy should be organised in order to comply with applicable State and Commonwealth laws. Green procurement has a complex regulatory aspect and a simple acquisition aspect. Under the Federal Mandatory Renewable Energy Target introduced in January 2001, an annual volume of electricity is prescribed that retailers must purchase from identified energy sources. There are overlapping State schemes. In New South Wales the Greenhouse Gas Abatement Scheme was introduced early in 2003. Under this Scheme retailers must seek to attain a target level of reduced greenhouse emissions in New South Wales. This can be achieved by purchasing greenhouse gas abatement certificates. A regulation is pending in Queensland to require the purchase of gas emission certificates to attempt to facilitate generation of electricity by gas to a level of 15% of total generation. Despite the complexity of overlapping State and Federal regimes the actual procurement of green energy is straightforward and not time or effort intensive.

5. Financial hedge contracts must be entered into to protect against exposure to volatility in the market for the acquisition of physical electricity.

6. It is then necessary to establish a customer base. A start up retailer does not require a significant customer load in its initial stages. Powerdirect now has a customer base of 12,000, primarily 40-160MWh per annum customers, and generating a net profit of $2 million on a revenue of $53 million. The relationship between customer load and volume of customers supplied will depend upon the target market. A retailer supplying to customers who consume more than 160MWh annually will have many fewer customers for the same load as one serving only customers consuming less than 40MWh/annum.

89                  In addition to the matters referred to by Mr Bonwick, Dr Price identified the need to establish information technology systems which would interface with the central NEMMCO customer registration and transfer systems and with the systems of other network and retail businesses operating in the area in which the retailer wishes to compete. Customer information technology systems must be created or modified to produce bills that meet regulatory requirements on timing and information provision. The retailer must also ensure that there are appropriate linkages between the marketing and trading systems in place in respect of the tariffs offered and the volumes of energy for which the retailer is responsible.

90                  On the evidence, the hurdles to entering into the business of electricity retailing are reasonably low and the requirements for a licence are not onerous. Dr Hieronymus, an expert witness called by AGL, said:

‘Electricity retailing is inherently an easy entry business. It is not patent protected. Power, particularly to larger customers trades as a commodity with business going primarily to the lowest cost seller. There is no specialized knowledge that cannot be purchased or hired away from incumbents.’

91                  I accept that the field of potential entrants into electricity retailing includes organisations with a strong retailing capability and associated expertise in the development and operation of billing and customer information systems. I also accept that such organisations could include telecommunication companies and banks and credit card companies which have sophisticated customer information and billing systems and strong brand awareness. It should be noted however that when Dr Price was asked by the Court how hard it is for a generator to set up its own retail operation he replied that it is ‘pretty hard’. Asked what the barriers are, he said:

‘The barriers are, one, they don’t have a relationship with customers at the moment, other than a few large customers. The current retailing systems are pretty closely tied in with the distribution systems that exist and they have no experience with that.’

I accept what Dr Price said in that respect. The difficulties he identified in the way of generators developing retail operations were related to securing appropriate systems and expertise. That being so it is not difficult to imagine a generator developing a retail operation in conjunction with another organisation possessing retailing capability and associated expertise.

92                  There are 23 licensed retailers operating in the NEM, 15 of whom are licensed in at least two regions. Of the 8 retailers licensed in a single region, five are generating companies with assets in that region. They do not actively participate in retailing in the NEM. There are 13 licensed retailers in Victoria, of which three hold Victorian retail franchises (AGL, Origin and TXU). Seven are retailers who have franchise areas in other NEM regions or proposed NEM regions. One retailer was launched by and continues to be wholly owned by Yallourn Energy trading as AusPower, and two are retailers without franchise or generator tie-ups, namely Powerdirect and Victoria Electricity. The latter company is a New Zealand owned retailer.

93                  As Dr Price observed, and I accept, retailers in each region of the NEM face competition from a large number of companies with substantial financial backing and significant expertise in providing customer services. Attached to these reasons as Annexure 3 is a copy of a table he prepared setting out details of corporate groups with retail licences in the NEM.

94                  The Court was provided with confidential evidence in relation to the percentage of customers in the NEM supplied by competing retailers following the introduction of full retail competition. This information was based upon details provided by NEMMCO of the quantity and value of energy purchased on the spot market by all registered market participants, including generators, retailers and large customers purchasing electricity directly from the spot market. Dr Price also relied upon details of the number of customers served by each retailer measured according to the NMI identifier assigned to each customer in the NEM. It is apparent from that information that the major retailers in Victoria are AGL, Origin and TXU. In the NEM the major participants appear to be AGL, Country Energy, Energex, Energy Australia, Ergon and Origin.

95                  There is a phenomenon in the retail sector known as ‘customer churn’. This describes the process by which a customer switches to a new tariff or to a new retail supplier. There is limited information available on churn in the NEM. NEMMCO does, however, provide daily information on customer transfers for New South Wales and Victoria. That information is not available from South Australia, the ACT or for the NEM as a whole. A table of customer transfers in New South Wales and Victoria since full retail competition is set out below.

Customer transfers since FRC in NSW and Victoria

 

Region	Small customer transfers	Large customer transfers	Total customer transfers	% total customers in region
NSW	119,936	13,245	133,181	4%
Victoria	213,367	12,442	225,809	10%

This data does not reflect customers changing tariffs.

 

The Operation of the NEM – Demand

96                  Energy demand is the sum of the expected energy use over a period and is usually measured in Gigawatt hours or GWh. Peak demand is the maximum instantaneous demand over a period and is usually measured in MW.

97                  Electricity demand today is determined by the consumption patterns of consumers. Not surprisingly it is generally lowest at night and highest in the morning and early evening. Weather has an important effect on its level and pattern. On a hot day in summer, demand will be higher because of the increase in air conditioning load. It will also increase on cold days in winter because of the increase in the heating load.

98                  The time of peak demand (the maximum demand for the year) varies between regions in Australia. NEM regions experiencing peak demand in summer include Victoria, South Australia and Queensland. The summer peaks in Victoria and South Australia occur at roughly the same time as these States are subject to similar weather patterns. There have been occasions on which summer demand has peaked in New South Wales at the same time as in Victoria and South Australia. The only NEM region which is currently peaking in winter is New South Wales and that is forecast to become a summer peaking region in the future. There is a pattern of volatile summer maximum demands and higher average winter demands common throughout the NEM.

99                  Demand forecasts are based on analyses prepared by organisations responsible for transmission planning in each region and by their advisors. In Victoria, the responsible body is VEN Corp. In South Australia, it is the Electricity Supply Industry Planning Council. In Queensland the responsibility resides with Powerlink which is the Queensland regional TNSP. In New South Wales and the ACT, it is the New South Wales Regional Transmission Network Provider called Trans Grid and in Tasmania, Transend Networks, which are responsible for transmission planning. NEMMCO seeks to ensure consistency between forecasts by issuing the responsible body in each region with a set of forecast guidelines and a consistent set of economic forecasts.

100               According to the NEMMCO Statement of Opportunities 2003, electricity consumption at times of peak demand is affected by factors which include:

1. The level of economic activity.

2. The rate of population growth.

3. The impact of environmental policies.

4. The price and availability of substitute sources of energy such as natural gas for heating.

5. Temperature, in that demand rises with the use of air conditioners and space heating; and

6. Technological innovations and changing consumer behaviour such as, for example, increased use of personal computers.

- NEMMCO Statement of Opportunities 2003, Chapter 2, p 2-2.

101               The Code provides for specified scheduled loads to bid into the NEM and receive dispatch instructions from NEMMCO (cl 2.3.4(d)-(g) and cl 3.8.7). This allows loads that are price sensitive to switch off if the price rises to a particular level or switch on if it falls to a particular level. There are three scheduled loads in the NEM. They are all hydro pumps operating at times of low prices to replenish reservoirs. The water is used to generate electricity at times of high prices.

The Operation of the NEM – Bidding, Pricing and Dispatch

102               Supply and demand in the NEM is matched at every point in time by a centrally co-ordinated dispatch process. This is also the mechanism for setting the spot price for electricity.

103               The logical commencement of the process is the making of offers to NEMMCO by generators specifying the quantities of electricity they are willing to produce for various prices. NEMMCO collates this information and produces a plan or schedule establishing the energy output level of each generator. It puts generator offers in order from the lowest to the highest price, with the lowest price offer being dispatched first and the highest price offer being dispatched last. The order or stack of offers is known in the NEM as the generation ‘merit-order’. A simplified representation of the process was set out in Figure 24 of Dr Price’s Outline of the NEM which is reproduced as Annexure 4. He assumes for the sake of illustration that the NEM comprises just Victoria and New South Wales and that there are only two power generators in each State, each capable of producing 100MW of power. The combined demand for electricity in demand is assumed to be 250MW. In order to meet that demand NEMMCO would stack generation offers from the cheapest New South Wales generator – Generator A, to the most expensive New South Wales generator – Generator B, and then dispatch only those generators required to meet demand. In the example given, all of the output offered by the New South Wales Generator A, the Victorian Generator A and half of the output offered by Victorian Generator B would be dispatched. As the Victorian Generator B is the most expensive plant dispatched, its offer price would be used to determine the spot price at that time. This spot price would also prevail across the interconnected NEM. All dispatched generators earn the spot price irrespective of their offer price. All customers buying electricity from the spot market will be required to pay that price.

104               Dr Price’s figure, reproduced at Annexure 4, assumes that the capacity of the transmission links between the regions will allow the cheapest power stations to be dispatched to the output levels which they have offered. This may not always be the case. Sometimes there is not enough transmission capacity to allow the cheapest power stations to supply existing demand in a given region. In that event, NEMMCO overrides the schedule of generator offers and sends instructions to the next most expensive power station not limited by transmission constraints to operate to a level that meets demand. The power station dispatched out of the ‘merit order’ will generally set the price in the region in which it is dispatched. This means the spot price varies across the NEM regions at the same point in time.

105               Generators are required to provide NEMMCO with their offers by 12.30pm each day for every half hour period for the following day. The trading day is assumed to commence at 4am. The offers are to specify:

1. The price at which the generators are willing to produce electricity; and

2. The quantity they are willing to offer at each price.

Generators are required to provide ten price choices, known as price ‘bands’. These are subject to the following restrictions:

1. There must be at least one negative price band. This meets the case where more generation is offered at a zero price than is required to meet demand. In that event, some generators may want to pay (ie receive a negative price) in order to avoid being switched off by NEMMCO. This is because of the expense and time involved in switching some generators such as coal-fired generators on and off.

2. The price offered by a generator must strictly be increasing with each price band being no lower than its predecessor.

The preceding process normally yields a series of prices rising slowly in the first six or seven price bands where generators expect to be dispatched with significantly higher price offers in the last one or two price bands. A chart prepared by Dr Price as Figure 25 in his outline, illustrates this and is Annexure 5 to these reasons.

106               Generators are not permitted to change the prices they have offered in each band once their final bids have been submitted. Those prices will apply over the trading day. They can change the volume they are willing to supply at each price band and trading interval up to 5 minutes before dispatch. This is a process known as rebidding. It allows management of the risk that they may not be able to meet quantities previously promised, eg by reason of a plant failure. Clause 3.8.22 of the Code provides, inter alia:

‘(a) Prices for each price band that are submitted in dispatch bids, dispatch offers and market ancillary service offers are firm and no changes to the price for any price band are to be accepted under any circumstances.

(b) Subject to clauses 3.8.22(c) and 3.8.22A, a Market Participant may vary its available capacity, daily energy constraints, dispatch inflexibilities and ramp rates of generating units, scheduled network services and scheduled loads, and the response breakpoints, enablement limits and response limits of market ancillary services.’

Market Participants are required under par (c) to provide a brief verifiable and specific reason for a rebid.

107               A new clause 3.8.22A was introduced into the Code in 2003 to protect against manipulative rebidding following price increases in the summer of 2000/01. It requires, inter alia, that:

‘(a) Market Participants must make dispatch offers, network dispatch offers, dispatch bids and rebids in good faith.

(b) In clause 3.8.22A(a) a dispatch offer, network dispatch offer, dispatch bid or rebid is taken to be made in good faith if, at the time of making such an offer, bid or rebid, a Market Participant has a genuine intention to honour that offer, bid or rebid, if the material conditions and circumstances upon which the offer, bid or rebid were based remain unchanged until the relevant dispatch interval.’

As it involved a significant change to the Code the new rule was the subject of an authorisation application by NECA to the ACCC under Pt VII of the Trade Practices Act. A determination granting authorisation was made on 4 December 2002 (X 32). The determination is of interest because of some of its findings with respect to the NEM and the degree to which price spikes in the market should be kept in perspective as an aspect of its ongoing development. It will be referred to later in these reasons. The change was inspired in part by the use of rebidding to lift prices in the summer of 2000/01. As will be seen later, Loy Yang Power used the bidding and rebidding process in a time of high demand to increase spot prices and consequently the prices of hedge contracts.

108               The dispatch of generated electricity to meet demand is effected by NEMMCO using purpose designed dispatch software. This is a linear program known as the National Electricity Market Dispatch Engine. It determines the most efficient combination of electricity and ancillary service providers to meet the requirements of the market. The object of the dispatch process is to minimise the cost of meeting demand based on generator bids. The process is complex and must take into account a number of constraints which include generator technical constraints and interconnect constraints. Using the dispatch software, a price is calculated for every five-minute period or dispatch interval. The five-minute price is called the ‘dispatch price’. To limit information requirements, while maintaining suitable price signals, settlement is undertaken on a half-hourly price, rather than a five-minute price. The half-hourly price is known as the ‘trading interval price’ or ‘spot price’. It is a time weighted average of the six dispatch interval prices that occur in the half-hour period known as the ‘trading interval’.

109               The spot price is determined for each region in the NEM at a point known as the Regional Reference Node. Spot prices will vary between Regional Reference Nodes according to interconnector constraints and the losses associated with transporting electricity between regions. Each generator receives a spot price for energy supplied by it that is referable to the price at its Regional Reference Node. That is known as the Regional Reference Price. Spot prices may not exceed the maximum price that could theoretically be charged for electricity in the NEM. That is now $10,000/MWh – Code 3.3.17 and 3.9.4. It represents the cost to consumers of not being supplied with electricity, known as the ‘Value of Lost Load’ (VoLL). It is imposed where there may be inadequate supply against demand and therefore involuntary load shedding.

110               There are various constraints which affect the operation of the system just described. One of these, which has already been discussed, is the finite capacity to transfer power from one point to another through the transmission system. The spot price determination will account for transmission constraints. Mr Greg Denton, for the ACCC, pointed out in his evidence that, within the Dispatch Engine, over 13,674 constraint formulae are used. These are mathematical equations designed to ensure that the physical limits of the transmission system are not exceeded. The constraint formulae are necessary because the physical transmission system is not modelled, just Regional Reference Nodes and notional transmission lines joining them. The formulae limit the output of a particular generator or generators and/or the flow between regions so that all lines are operating within their rated capacity. When it is necessary to restrict the output of a generator or group of generators by reason of such a constraint, the constraint formula is said to ‘bind’, ie it is binding on the Dispatch Engine solution. A binding constraint may be intra-regional in character which would result in part of the capacity of a generator or generators in the region not being dispatched even if their bids were lower than the spot market price at the time. In that event, the spot price would be determined by reference to bids from the unconstrained generators.

111               To build on his figure illustrating the bidding process Dr Price produced a similar figure showing the effect of inter-regional transmission constraints. The simplified example used in the earlier figure was elaborated by the introduction of a 25MW interconnector between New South Wales and Victoria and the assumption that demand for electricity in New South Wales is 150MW and in Victoria 100MW. To meet demand in the least costly way, NEMMCO must now dispatch 100MW of NSW Generator A, and 100 MW of Victorian Generator A which meets the full 100MW Victorian load. It must also dispatch so much of Victorian Generator B as the interconnector will allow to be exported to meet NSW demand. This amount is 25MW. This will need to be combined with 25MW of NSW Generator B. The figure, which is Figure 26 in Dr Price’s outline, is Annexure 6 to these reasons.

112               In this example, there will not be a single spot price across the interconnected New South Wales and Victorian system. There will instead be separate prices for New South Wales and Victoria. Under these conditions the New South Wales price will be set by New South Wales Generator B. Victoria’s price will be set by the most expensive power station dispatched in Victoria, which is Victorian Generator B. So the New South Wales price will be higher than the Victorian price. New South Wales consumers will pay the New South Wales Generator B offer price and the New South Wales Generators will earn that price. Victorian consumers will pay the Victorian Generator B offer price and Victorian Generators will earn that price.

113               In addition to the effect of interconnector constraints, electricity transported along the network from power stations to consumers suffers losses because of the electrical resistance of the conductors. Approximately 10% of electricity is lost in this way. The actual amount lost depends upon flow across the network but may also be dependent on such variables as ambient temperature and the voltage of the lines as high voltage lines tend to lose less energy than low voltage lines.

114               The NEM spot prices are adjusted to take into account the effect of these transmission network losses. Losses associated with the transfer of electricity between two points are approximated using marginal loss factors which estimate the electrical losses of each additional increment of electricity transmitted between connection points. If there are no constraints upon transmission between regions, the spot prices will vary between the Regional Reference Nodes only by reference to the marginal loss factors. Mr Denton gave an example in his evidence. The Victorian market spot price might be based on a bid from a New South Wales generator of $40/MWh, if that were the most expensive bid accepted by the Dispatch Engine to meet load at that time. In that case the Victorian spot market price might be $42/MWh while the New South Wales spot price might be $40.50/MWh. The difference would represent the losses incurred with the transportation of electricity from the generator.

115               Two types of transmission marginal loss factors exist in the NEM. These are intra-regional loss factors and inter-regional loss factors. Intra-regional loss factors represent the marginal electrical losses associated with transporting electricity to or from a load or generator to the Regional Reference Node. These loss factors are calculated by NEMMCO for each connection point and are fixed for twelve months. Inter-regional loss factors represent marginal electrical losses associated with transporting electricity between Regional Reference Nodes. NEMMCO uses equations which estimate the losses between regions. These are fixed for twelve months. The inter-regional loss factors are calculated for each dispatch interval using those equations. The generator bids are adjusted by reference to these loss factors so that the level of losses is accounted for in determining which generators are dispatched.

116               On the other hand, when an inter-regional constraint binds, regional spot prices may diverge on the basis of the different generator bids available to the Dispatch Engine in each region. Mr Denton says, and I accept, that if a constraint limits the flow of power from New South Wales to Victoria the price in Victoria could be based on the bids from generators that were able to supply the Victorian Regional Reference Node. In this scenario, and absent the competitive threat from the New South Wales generators, the price in Victoria can rapidly escalate to many thousands of dollars. In Mr Denton’s view the different treatment of intra-regional and inter-regional constraints in the NEM reinforces its regional character.

117               Mr Denton’s evidence was that a small number of price spikes in each year, caused by inter-regional constraints, can result in quite different average prices at each Regional Reference Node. He set out a table of average prices in each NEM region for the four financial years from 1999-2003. In 1999/2000 the price in South Australia was, on average, 132% higher than the Victorian price. In 2000/2003 South Australia’s average was 9.3% higher than Victoria, while the New South Wales price was 19.5% higher. Mr Denton contended that transmission constraints were a significant contributing factor to these price differences.

Financial Year Average Prices ($/MWh)
Financial Year	QLD	NSW	SNOWY	VIC	SA
1999-2000	45.25	28.88	27.79	26.11	60.61
2000-2001	42.19	38.36	37.72	45.39	57.33
2001-2002	35.34	34.76	31.59	30.97	31.61
2002-2003	37.77	32.90	29.82	27.54	30.10

118               I accept the general proposition that transmission constraints can be and have been a significant contributing factor where price differences have occurred between regions. I do not accept the more sweeping conclusion offered by Mr Denton that for this reason ‘… competition in the physical supply of electricity and related ancillary services primarily occurs at a regional level in the NEM’. In particular I do not accept that that contention holds good for the foreseeable future. In so concluding I have regard to the relatively early developmental stage of the NEM which has been in existence for just six years operating in a complex matrix of State regulatory laws and the market pricing mechanisms adopted under the Code.

119               When a generator is dispatched to meet demand it will receive the spot price at the connection point. Consumers will pay the spot price at their respective connection points to their retailers. The retailers in turn pay NEMMCO the spot price for electricity delivered to their customers. The spot price at each transmission network connection point is calculated as the spot price at the Regional Reference Node adjusted by the intra-regional loss factor for that connection point. The Loy Yang A power station, for example, has an intra-regional loss factor of 0.9636. If the price for a trading interval at the Victorian Regional Reference Node were $20/MWh and Loy Yang A was dispatched for that trading interval, it would earn $19.27/MWh, calculated as the product of 0.9636 and $20. The spot prices for electricity in the NEM are variable within every twenty-four hours according to changes in demand. As earlier noted, they tend to be lowest overnight and relatively higher during the day. They can vary by factors of many hundreds between half-hours over a trading day. Mr Nethercote, the Chief Executive Officer of LYPM, said that in his experience the NEM is characterised by sustained periods of consistent pricing interspersed with a few short periods of price spikes. The phenomenon of sudden spikes or dips in the pool price is referred to as spot price volatility. The factors affecting spot price volatility within a region can include generator unit trips and failures, transmission line outages and network plant repairs and NEM information technology failures and anomalies. External events such as extreme weather can result in short periods of extremely high demand. This is mostly relevant to South Australia and Victoria in the summer. Fluctuations in demand can also be caused by bushfires and other natural events.

The Operation of the NEM – Spot Price Volatility

120               The volatility of spot prices necessarily gives rise to risks for generators and retailers. Generators will be at risk from low spot prices, while retailers will be exposed to high spot prices. Generators need to maintain their cash flows in order to meet operational, maintenance and fuel costs and financial charges. Retailers need to manage their gross margin. For base-load generators with high fixed costs such as Loy Yang A, there is a particular need for a degree of certainty that they will receive revenue to cover their costs.

121               By way of example LYPM is responsible for lodging bids with NEMMCO for the dispatch of Loy Yang A. A pool operations trading function is carried out by a small group entitled ‘The Energy Trading Team’. This team is ultimately overseen by the General Manager of Marketing and managed by the Manager, Energy Trading. It comprises a trader with primary responsibility for the submission of bids and market analysis, a Trading Analyst and other marketing group members and operations staff. The objective of LYPM’s pool trading strategy was said by Mr Nethercote to be:

‘… to optimise revenue outcomes for the business within the market rules established under the NEC.’

122               Among matters considered by the Loy Yang Energy Trading Team in determining its short term and long term bidding strategies are:

(a) NEMMCO supply and demand forecasts (including short term projected assessment of system availability and medium term).

(b) Daily NEMMCO data recording the previous day’s bidding and dispatch data which is provided to market participants. In this respect, according to Mr Nethercote, LYPM pays particular attention to:

(i) the movement of pool prices and the apparent reasons for any movement;

(ii) the volumes bid and dispatched by other generators in the NEM;

(iii) the price bands at which those volumes are bid and dispatched;

(iv) which generators are ‘setting’ the pool price; and

(v) changes in the bidding behaviour of other generators in the NEM.

(c) Transmission interconnector limits.

(d) Weather forecasts.

(e) Outages, scheduled and unscheduled.

(f) Forecast constraints.

(g) Anticipated bidding behaviour by other market participants.

(h) Historical LYPM data recording the outcome of previous bidding strategies.

(i) Reports of consultants on long term strategies to optimise revenue for LYPM.

(j) Any new generation plant pending in the NEM and any generation being contemplated by potential new entrants.

(k) Any changes to interconnection being contemplated.

(l) Government policy; and

(m) Pending changes to the NEC.

The results of these analyses and factors considered significant in respect of bidding strategies are collated into weekly and monthly reports.

123               Factors determining the amount of capacity that LYPM bids into the NEM are as follows:

1. Plant factors – namely Loy Yang A’s generating capacity at any point in time.

2. Planned outages by LYPM.

3. Operational constraints.

4. Environmental constraints due to impurities in coal supplies.

5. LYPM’s hedge contract commitments. If insufficient volume is dispatched to meet LYPM’s hedge contract commitments it will be exposed to the risk of negative difference payments if the spot price exceeds its contract spike prices. The operation of the hedging process is described below.

6. Ancillary market conditions.

The judgments informing strategies and day-to-day biddings are clearly complex, multi-dimensional and necessarily based on incomplete information. They are inherently uncertain. This is a characteristic of the pricing process in the NEM that has implications for the extent to which any single operator can enjoy and reliably exercise market power.

 

The Operation of the NEM – Hedging Arrangements and Electricity Derivative Contracts

124               The way in which generators and retailers protect themselves from movements in spot prices is by using financial or derivative contracts. These are financial contracts under which one party makes a commitment to pay a counter-party according to spot price outcomes. They do not require physical delivery of electricity as a contract term. However, they create financial incentives which often mean that the participants will try and deliver electricity to meet their contracted demand. The generator will always be paid the relevant spot price for energy dispatched. Depending upon whether it has hedged and the nature of the hedge contract, it will receive additional money from the retailer with whom it has contracted or pay out money to that retailer. There is very little reliable, publicly available information about trading in electricity derivative contracts.

125               The kinds of electricity derivative contracts used by participants in the NEM to hedge their exposure to spot prices are as follows:

1. Swap contract: In this class of contract a party agrees to settle a fixed or variable quantity of electricity at an agreed price which is known as a ‘strike price’ with another party. If the spot price is greater than the strike price of the contract, one party will pay the difference by reference to the number of MW of electricity covered under the agreement.

2. Cap Agreement: This is an agreement in which one party agrees to settle a fixed or variable quantity of electricity at a maximum price with another party. The seller will pay the buyer the difference if the spot price is greater than the strike price.

3. Floor Agreement: In this class of agreement a party agrees to settle a fixed or variable quantity of electricity at a minimum price with another party. The seller compensates the buyer if the spot price is less than the strike price.

4. Collar Contract: In this class of contract the floor and the cap are created together, one party buys a cap and the other buys a floor.

5. Swaption Agreement: This is an option to enter into a Swap contract at a future date, at a predetermined price. The subclasses ‘call swaption’ and ‘put swaption’ respectively designate swaptions in which the buyer has the right to ‘buy’ or to ‘sell’ the fixed price payments.

6. Asian option: This is also known as an ‘average rate option’. It is a contract to make payments if the average price for electricity varies from the strike price over a specified period. If the average price in that period exceeds the option price of an Asian call option then the seller pays the buyer the difference. The buyer pays a premium to enter into the contract.

7. There are many forms of similar financial hedging instruments, the variety of which has been growing. By way of example weather derivatives recognise that extreme weather conditions are often associated with high and volatile prices.

126               The two principal mechanisms for entering into hedging contracts in the NEM are:

1. Over the counter (OTC) contracts. These contracts involve a bilateral agreement with a known counter-party. They can be negotiated directly with other market participants, eg retailers or generators, or arranged through a broker who will enter contracts with standard terms and conditions. According to Mr Nethercote’s evidence, which I accept, the majority of the smaller hedge contracts involving trades below 50MW are traded through registered brokers. Large volume and shaped hedge contracts are generally negotiated bilaterally between a generator and a retailer or other end user.

2. Exchange traded contracts. Here parties enter into a standardised contract with an exchange. The buyers and sellers of the contracts are not known to each other.

127               The financial settlement mechanism, that is the contract billing and payment system, differs between the OTC and exchange traded contracts. Exchange traded contracts are settled daily depending upon the change in the value of the position compared to the previous day. If the contract is held to maturity, it is cash settled. Such contracts are settled through a clearing house which generally assumes the credit risk or the parties dealing with the exchange. OTC contracts are settled in arrears and periodically by participants or an outsourced settlement company. This is done by comparing the contract strike price to the spot price. The parties to these contracts bear each others credit risk. The majority of hedging requirements engaged in in the NEM are met through OTC contracts.

128               There are two derivatives exchanges in Australia. The Sydney Futures Exchange, trades futures and options on equity derivatives, interest rate derivatives and commodity derivatives which include electricity futures. The Australian Stock Exchange Futures Exchange trades grain and electricity futures. Exchange traded derivatives are traded on the centralised exchange in contract units. Sydney Futures Exchange products are based on base load and peak load energy bought and sold in a calendar quarter in the NEM. The ASX Futures Exchange offers a standard off-peak contract whereas the SFE offers a standard base load contract. Unit of electricity futures contracts in both Exchanges is 1MWh over a specified number of hours determined by the length of the quarter. Transactions on the exchanges are sometimes referred to as ‘screen trades’ because information in relation to their trading is available on the SFE or ASX trading screens and the Electricity Brokers Screens on the Reuters service.

129               OTC derivatives in Australia import the terms and conditions of the International Swaps and Derivatives Association (ISDA) Master Agreements. There is an electricity addendum to the ISDA Swap Agreements which is widely used in Australia. However, such agreements and in particular force majeure agreements may be customised according to negotiation between the counter parties.

Operation of the NEM – Inter-regional Hedging and Inter-regional Settlement Residue Auctions

130               The hedging contracts in the NEM refer to prices at a particular Regional Reference Node. However, market participants do enter into inter-regional financial contracts. Such contracts can lead to price exposure for a market participant if the price at its local Regional Reference Node differs from the price at the Regional Reference Node where the contract is struck. The extent of inter-regional price risk depends on the frequency of the constraints between regions and the divergence between regional prices at those times. Dr Price has observed an increase in the extent to which market participants whom he advises are entering into inter-regional hedges since the commencement of the NEM. He says and I accept that the increase in the resort to inter-regional hedges is linked to the growing familiarity of market participants with patterns of inter-regional price differences and their causes and with the increase in interconnection capacity.

131               The rules of the NEM allow for the management of inter-regional price risk or ‘basis risk’ when inter-regional hedges are written. This is done by what is called an Inter-Regional Settlement Residue Auction conducted by NEMMCO. Inter-Regional Settlement Residue is the revenue which NEMMCO accumulates at settlement as a result of the difference between regional prices. NEMMCO periodically auctions the rights to shares of Inter-Regional Settlement Residues in future periods in order to allow participants to manage inter-regional pricing risk. These auctions do not provide a perfect hedge for inter-regional price differences. The reduction in the transfer capacity of a interconnector because of maintenance will reduce the Inter-Regional Settlement Residue to be shared between participants.

132               Another mechanism for the management of inter-regional price risk is a capacity swap between generators. In such arrangements generators in different regions agree to make a portion of their generating capacity available to each other to trade in the other region. So a New South Wales generator might give a Victorian generator 100MW of its capacity to trade at the NSW Regional Reference Node in exchange for the Victorian generator giving the NSW generator 100MW of the Victorian generator’s capacity to trade at the Victorian Regional Reference Node. Market participants may also enter into a cap contract with a generator in the region which is most likely to be importing at times of constraint in order to limit price exposure. Participants take inter-regional positions using any of these risk management options. Some participants take inter-regional positions without hedging inter-regional price risk.

133               In Dr Price’s opinion, which I accept, the extent to which participants attempt to mitigate inter-regional price risk generally depends upon their assessments of the likely risk of inter-regional price differentials including their size and duration and the risk preferences of the participant.

The Loy Yang Power Station and Coal Mine

134               The Loy Yang A Power Station is a brown coal-fired steam driven power station which with the associated Loy Yang coal mine is located in the La Trobe Valley about 165 kilometres south east of Melbourne. It comprises four 500MW turbo-generators. Although designated as 500MW the generators have been augmented so that the maximum generation capacity of the power station is now 2,050MW. Further augmentations are planned to increase that capacity to 2,150MW over the next five years. Of the electricity generated by the station, 10% is used in operating it, the coal mine and a raw water pumping station. This means that the nominal maximum generation available for dispatch is about 1,840MW. Coal-fired plants such as Loy Yang A are physically large and have large generation capacity in contrast with other types of power plants such as gas-fired power plants. They have high capital costs relative to other types of plant. The characteristics and operation of Loy Yang A were described in the evidence of Mr Nethercote who is the Chief Executive of LYPM. The descriptive elements of this evidence were not in dispute and I accept them.

135               The source of the mechanical energy to turn the generators in Loy Yang A is steam. Coal is burnt to boil water and produce high pressure jets of steam which turn the fan blades connected to the generator. The power station uses about 60,000 tonnes of brown coal each day and this is obtained exclusively from the mine. Coal from the mine is also used at the Loy Yang B power plant which is owned and operated by Edison Mission Energy. The coal delivered to Loy Yang Power A is delivered on a ‘just in time’ basis. This is because there is only a limited storage capability at the station. Its main bunker can store up to 75,000 tonnes of coal which is less than one day’s supply for Loy Yang A and Loy Yang B together when they are operating at full load. There can be operational constraints affecting the mine’s capacity to deliver the necessary quantity of coal and if there is such a constraint generation of electricity from either or both of the stations may be reduced to conserve the amount of coal in storage. The quality of the coal provided varies significantly due to its change in composition in various digging locations in the mine. This can have an impact on generation capability if the coal is of poor quality. Differing numbers of generators may have to be operated to produce the same amount of electricity.

136               The desired operating range of Loy Yang A is achieved by maintaining a stable level of generation output at or above a minimum generation level of approximately 300MW per unit. The level of output for Loy Yang A has been consistently high as is demonstrated by comparing capacity factor against available capacity factor for the years 1999 through to 2003. The available capacity factor is the total potential generation in MWh divided by the product of the period of hours and the maximum dependable capacity in MW. The capacity factor is the total actual generation in MWh divided by the product of the period of hours and the maximum dependable capacity. Results in 2002 were significantly lower than for the other years and this was explained by a major generator failure in Unit 4 which gave rise to a four-month unplanned outage.

137               Loy Yang A, in common with coal-fired plants generally, has a relatively long start up time and higher ‘start up’ costs to refire the boiler from a cold condition compared to other types of generators such as gas-fired generators. For this reason it is not suited to intermittent operation. It answers the description of a base-load generator in the classification of generators mentioned earlier.

The Loy Yang Power Station – Original Acquisition and Financing - 1997

138              Initially the Loy Yang A power station and mine were owned by the former State Electricity Commission of Victoria (SECV) which commenced the construction of the power station in 1977. The generators were brought on line between 1984 and 1988. Following the restructuring of the electricity industry, the power station and mine were transferred to the corporatised entity Generation Victoria.

139               On 12 May 1997, the power station and mine were sold for $4.855 billion to a consortium of companies which formed a partnership, and which still own and carry on the business of operating the power station and mine, (LYP). The partners are:

1. Horizon Energy Holdings Ltd which has a 24.63% interest in the partnership and CMS Generation Horizon Energy Holdings Ltd which has a 25% interest. Both are owned by CMS Generation Investment Company 1, CMS Generation Loy Yang Holdings 2 Ltd and GMS Generation Loy Yang Holdings 1 Ltd.

2. Horizon Energy Investment (No 2) Pty Ltd which has a 25% interest in the partnership and is owned by Horizon Energy Investment Ltd.

3. NRGenerating Holdings (No 4) BV which has a 25.37% interest in the partnership and is owned by NRGenerating International BV.

The capital structure of the business consisted of $3.55 billion of debt and $1.3 billion of equity. The LYP is operated by LYPM as agent for the Loy Yang partners under an agency established by a Deed of Appointment of Partnership Representative between the partners and LYPM dated 17 April 1997. The partnership commenced its operations at Loy Yang A on 12 May 1997.

140               The assets of the partnership were, and still are, subject to a charge and other securities in favour of a syndicate of banks whose agent is the Australian and New Zealand Banking Group Ltd. A Security Trust Deed entered into between the banks and the Loy Yang partners is in evidence. In addition there is a number of loan agreements between LYPM and its bankers. These embody covenants for the repayment of interest and principal at certain times. The Bullet Payment due to the banks was extended to 11 July 2003 then further to 11 November 2003 and has evidently since been further extended.

The Loy Yang Power Station – Existing Security Arrangements

141               It is not necessary for present purposes to set out the full scope of financial arrangements between the Loy Yang partners and their bankers. However, the Security Trust Deed is of some significance because of the constraints it imposes upon hedging policies and risk management policies adopted by the Loy Yang partners. The Deed obligates them to adopt risk management hedging in respect of interest due under the loan and the prices to be paid to them for the sale of electricity. The latter obligation requires the Loy Yang partners to ensure that the prices to be paid to them for the sale of electricity are hedged in accordance with a specific Hedging Policy (9.16(b)). The Hedging Policy is defined in cl 1.1 of the Deed by reference to initial electricity hedging guidelines provided by the partners to the ANZ Bank as amended from time to time pursuant to cl 9.16(e). The content of the guidelines included the Loy Yang partners’ overall strategy for marketing electricity and for entering into electricity hedging arrangements, strategies to offset the decrease in the level of cover provided by the Vesting Contracts as a result of tariff customers choosing to become contestable customers, provision for hedging with other generators to cover planned and forced outages and a policy for hedging against ‘firm to VoLL contracts’.

142               Alterations to the Hedging Policy are subject to the approval of the ANZ Bank as agent for the consortium of lenders. A criterion of approval set out in cl 9.16(e) is that any new Hedging Policy not adversely affect the ability of the Loy Yang partners to satisfy certain ratios. These are to be satisfied as a condition precedent to the obligations of the lenders and security providers in respect of the loan transactions including the Security Trust Deed. They include a forecast Loan Life Cover Ration (LLCR) determined according to cl 9.20 of the Deed. The ANZ, as agent for the lenders, would reject a new hedging policy if it determined, on the advice of a defined person called the Economics Adviser that the policy would adversely affect the ability of the Loy Yang partners to satisfy the ratios in cl 7(k)(i).

143               The Loy Yang partners also covenanted, under cl 9.8 of the Deed, that each partner will do everything which it is able to do to ensure that the Project is carried on in accordance with ‘Prudent Practice”. The term ‘Prudent Practice’ is a defined term. Paragraph (a) of its interpretation in cl 1.1 defines it, at a particular time, by reference to practices, methods and acts engaged in or approved by a significant portion of the electricity utility industry or the coal mining industry in Victoria prior to the reference time. Paragraph (b) of the definition sets out its application in respect of hedging and other risk management agreements and narrows that application to accord with practices, procedures and acts specified in the general definition which relate to the specific portion of the electricity utility industry operating in the relevant electricity market ‘in Victoria and Australia’ and having regard to a number of issues which are then specified in the definition. They are:

(i) The prevailing supply demand balance in Victoria and the interconnected market regions (New South Wales, the Australian Capital Territory, South Australia and eventually Queensland and Tasmania);

(ii) The market positioning of the project’s major competitors (eg Loy Yang B, Hazelwood and Yallourn);

(iii) The extent of market power of the dominant participants (especially two of the New South Wales generators, Macquarie Energy and Delta Electricity);

(iv) The encouragement of new entrants which would result from higher pool prices and rising contract prices over an extended period;

(v) The likelihood of regulatory intervention in the marketplace arising from any perception of abuse of market power;

(vi) The actual availability of, and credit worthiness of counterparties to, such hedging and other risk management agreements;

(vii) The borrower’s obligations, and demonstrated past and continuing ability, to pay and repay the Secured Money as and when payable or repayable.

Under cl 9.41 of the Security Trust Deed the partners are required, no later than thirty days before the end of each financial year, to provide the ANZ Bank as agent for the lenders, with a copy of their proposed budget for the next financial year prepared on a quarterly and annual basis. The ANZ Bank is required to consult with the other banks in relation to that budget.

144               As appears from these provisions, the Loy Yang partners’ financiers have the capacity to exert a significant degree of control over both their hedging policy and their prudential practices. Under cl 10.1, dealing with events of default, a change in control of the partnership or the Loy Yang partners without the prior consent of the ANZ Bank as agent for the lenders would be a default incident.

The Loy Yang Power Station – Events Leading to the Acquisition Proposal

145               In the middle of 2002 the Loy Yang partners decided to sell or restructure their partnership. A $500 million debt bullet payment was due in May 2003, six years after the original acquisition. None of the shareholders was able to meet that payment. This led to the shareholders agreeing to consider what Mr Fraser, in a paper to the AGL Board on 4 December 2002 called, ‘all exit options ahead of the maturing Bullet Payment’ .

146               In December 2002, the members of the Loy Yang partnership opened negotiations with Genting Berhad (Genting) and Sime Darby Berhad (Sime) known as the Malaysian Power Consortium in which Genting and Sime each proposed a fifty per cent investment for the purchase of Loy Yang. The first draft of an agreement was sent to the Malaysian Power Consortium in mid-January 2003 and negotiations on it continued until March 2003. The negotiations were terminated when Genting advised the Vendors that Bank Negara approval had been refused so that the consortium was unable to continue with its proposed acquisition. As appears later in these reasons, an offer by a Genting related consortium was made on 11 September 2003 but for reasons explained by Mr Ian Kay, the Managing Director of Horizon Energy Investment Management Ltd, was not capable of acceptance or completion by the Loy Yang partnership.

147               International interest in the acquisition of the power station and coal mine continued. In July 2002, Vinay Kolhatkar, the Head of Private Capital, Equities & Advisory, Institutional Banking of the Commonwealth Bank of Australia was contacted in relation to a possible purchase of LYP. The contact came from a representative of a merchant banker acting on behalf of an international utility company. Conversations with the merchant banker did not lead anywhere but Mr Kolhatkar decided to look into the possibilities of a CBA investment in the Loy Yang business. In October 2002, he initiated a search for other partners to invest with CBA in LYP. One candidate was AGL which was an existing client of CBA.

148               The CBA gave a presentation to representatives of AGL on 16 October 2002. It involved the suggestion that both the CBA and AGL should be co-sponsors of the new consortium to own LYP. The presentation was fruitful and on 21 November 2002, AGL and the CBA executed a Commercial Term Sheet in respect of the proposed acquisition.

149               The CBA and AGL were joined in January 2003 by the Tokyo Electric Power Company Incorporated (TEPCo). TEPCo is Japan’s largest energy utility and the world’s largest privately owned energy company. It services people in and around Tokyo. It has a customer base of about 27,000,000 and sells about 275,500GWh of electricity each year. It has 40,000 employees and is a fully vertically integrated electricity company with generation transmission distribution and retail businesses. It owns power generation assets in Japan which have an aggregate generation capacity of 60,000MW. Three of these are coal-fired power stations generating 2,712MW. According to its general manager, TEPCo regards Australia as a stable market economy with relatively low investment risk. It has a long standing trade relationship with Australia from which it sources coal and gas to fuel certain of its thermal power plants in Japan. AGL and CBA entered into a Commercial Term Sheet with TEPCo on 9 January 2003. By this stage they were describing their consortium as the Great Energy Consortium. The acquisition project was named Project Entropy.

150               According to the Commercial Term Sheet of 9 January 2003 it was intended that TEPCo would provide high level technical services to the consortium under a Technical Services Agreement. The negotiations and execution of significant power purchase agreements with AGL were said in the Commercial Term Sheet to be ‘fundamental to Project Entropy’. AGL had, and still has, an existing hedge position with Loy Yang Power under an agreement called the Deemed Profile Hedging Agreement which is discussed later in these reasons. This agreement was to be taken into consideration and would contribute to the final power purchasing agreements underpinning the acquisition of Loy Yang A.

151               CBA’s general approach to its investment in the consortium was described by Mr Kolhatkar in cross-examination. CBA does not regard the investment as long-term. Its intention is to treat $59 million of its investment as short-term, leaving a further $25 million in the consortium for about four years. Mr Kolhatkar agreed with the proposition that in four years time the CBA could have ‘… a very modest investment in the consortium’. However he rejected the suggestion that the CBA would be a passive investor. I acceptthat whether or not the Bank reduces its level of investment to $25 million in the short-term, it is unlikely to be a mere passive investor in an industry which requires such careful risk management particularly in respect of bidding and pricing.

152               TEPCo’s approach to its investment in the consortium was explained by Toshiro Kudama who is the General Manager of the Business Development Group of the International Affairs Department of the company. According to Mr Kudama’s affidavit, TEPCo’s approach to foreign investment focuses upon assets that have the potential to deliver strong financial results. It seeks associations by way of joint venture or similar structures with strong local partners or partners who have an understanding of local market conditions, a strong and established presence in the local market and an understanding of, and experience in, the energy industry and particularly the electricity industry. TEPCo also seeks investments that provide opportunities for it to utilise its strong technical expertise in the operation and maintenance of thermal power plants.

153               The Loy Yang Business met TEPCo’s foreign investment criteria. The company was also aware that the Loy Yang partners were keen sellers given the financial difficulties that they were experiencing. It regards AGL and the CBA as obvious choices as local partners. It sees AGL as a leading Australian energy company enjoying strong relations with government, understanding local industrial relations issues and having access to capital as a major listed company. The CBA also satisfies its requirements. All of these reasons for TEPCo’s involvement are self-evident. They do not, however, address the question of the nature of TEPCo’s involvement in the consortium if the acquisition proceeds.

154               In my opinion it is clear, even at this stage, that TEPCo will not delegate to AGL or any other consortium member the protection of TEPCo’s interests. It would be surprising, to say the least, if a company of its size, sophistication and resources did so. Nor, in my opinion, will it allow LYP policies and practices to be distorted in favour of AGL interests to the detriment of other shareholders. Mr Kudama said, and I accept, that in negotiating with other consortium members, including AGL, his principal objective was to ensure that TEPCo’s own commercial interests were fully protected as an investor in the LYP and as a shareholder in the company which was to be formed by the consortium for the purpose of the acquisition. He was concerned to ensure that TEPCo’s commercial interests would not be prejudiced or that giving effect to the terms of the Shareholders Agreement ultimately negotiated between the parties would not result in the LYP being operated in any other way than in the best interests of the consortium members as a whole. In the event, he was satisfied that the terms of the arrangement finally agreed upon are entirely compatible with the achievement of TEPCo’s foreign investment objectives and in its best interests.

155               An indication of the care with which TEPCo addresses its own interests is its assessment of a special hedging arrangement called the Deemed Profile Hedging Agreement (DPHA) between AGC and Loy Yang. In February 2003, as part of its due diligence process, TEPCo engaged Deloitte Touche Tohmatsu to report to it on the DPHA. The report was provided on 27 June 2003. Deloittes described the agreement as representing ‘a large portion of Loy Yang’s future contracting levels and hence … key for the Consortium and investors in analysing the value and risks of LYP’. It is not necessary for present purposes to explore the content of the report. It is sufficient to say it was lengthy and detailed and based in part upon modelling of the operation of the DPHA developed by Deloittes. It did not deter TEPCo from proceeding with its involvement in the consortium. It was, however, an indicator of the resources which TEPCo can bring to bear in making its own commercial judgments and supports the conclusion that it is unlikely to subordinate its interests in any way to those of AGL in the operation of the consortium.

156               The final form of the consortium was constituted with additional parties, being the Motor Trades Association of Australia Superannuation Fund (MTAA) and Westscheme. Various documents were executed by the consortium members on 3 July 2003, being the GEAC Subscription Deed, the Loy Yang Share Sale Agreement and the GEAC Shareholders Agreement. An overview and details of the relevant transaction documents appear below. Major conditions precedent to be met before completion included resolution of ACCC issues at least in the view of the partners, and unanimous senior and junior lender consent to the restructuring of their facilities. The Share Sale Agreement had contemplated completion by 2 September 2003. The completion date was subsequently extended to 12 September 2003 and then to 19 December 2003. The extension of the date for satisfaction of the conditions precedent through to 19 December 2003 was required to provide additional time to meet the outstanding conditions precedent, including resolution of the outstanding ACCC issue.

157               Following the execution of the documentation the Loy Yang Partners received an approach from a wholly owned subsidiary of Genting, a Malaysian company, offering to acquire 100% of LYP. The offer was extended on 17 September and 28 October 2003 to 31 December 2003. Ian Kay, the Managing Director of Horizon Energy Investments Management Ltd, regarded the offer as not capable of acceptance or completion by the Loy Yang partners because it was based on an incomplete document only partially negotiated with the Malaysian Power Consortium in March 2003. The offer required full documentation to be negotiated so that its final terms remained uncertain. It also proposed conditions precedent including unanimous approval of senior and junior lenders to a restructure of their debt, the undertaking of due diligence before completion, a written waiver of stamp duty or an agreement to ex gratia relief from the relevant Victorian authorities, shareholder approval and approval by the Foreign Investment Review Board. In any event, the Loy Yang partners were prohibited under the terms of their Share Sale Agreement with GEAC from soliciting any offer or discussing any offer with any third party, including the Genting interests.

158               According to Mr Kay if the GEAC Share Sale Agreement of 3 July 2003 is unable to be completed the Loy Yang partners will face the options of selling the power station and coal mine, entering into a long term extension with their financiers or having a receiver or some other form of insolvency administration put in place if appropriate. I accept that these are the principal options they face.

159               Mr Kay also said in an affidavit in support of expedition of these proceedings and I accept that until the financial position of the LYP becomes more certain, it will remain difficult for Loy Yang A to enter into any new hedge contracts because of counter-party credit concerns. This has been the case for the last twelve months or so. Because of this, Loy Yang A has less contract cover than historically has been the case and its revenues are subject to greater uncertainty. He is also aware of increasing concern among suppliers regarding LYP’s ability to meet its obligations. This impacts upon the LYP’s ability to ensure optimal operating and capital expenditures and has the potential to impact upon the operation of the power station.

The Loy Yang Acquisition – External Financing

160               The external financing of the consortium is still under negotiation. Mr Kolhatkar has been involved in discussions with representatives of the banks presently funding the Loy Yang partnership. There are some thirty-one banks involved. He and representatives of GEAC have been dealing with a steering committee comprising five of the thirty-one banks. According to Mr Kolhatkar’s evidence, which I accept, the consortium members wish to retain the existing syndicate of banks as the external financiers to the project. They are seeking a reduction of the level of senior debt but otherwise terms and conditions similar to those which presently apply to the Loy Yang partners.

161               At the time of the hearing GEAC had proposed a term sheet which had been circulated to the banks but had not been accepted. There is, as Mr Kolhatkar agreed under cross examination, ‘… no deal currently between the GEAC investors and the banks’. He did say that the GEAC representatives had had discussions with the Loy Yang management and had made an active decision not to seek any alteration of the existing Hedging Policy. But he accepted that if the consortium proceeded with the current hedging policy it could be changed by agreement between the parties.

162               It is clear that at the time of hearing no external financing arrangements had been concluded. I am unable to conclude on the evidence whether it is likely that financing arrangements with the existing syndicate of banks will be successfully concluded and if so what its terms and conditions would be. However, it is highly likely, in my opinion, that significant external financing will be required and, having regard to the financial strength of the consortium members, will be procured. It is likely, in my opinion, to impose constraints by reference to risk management and hedging policies similar to those presently in force. The fact that the consortium members wish to maintain such arrangements, which have already proved acceptable to a large syndicate of banks, supports the conclusion that it is highly likely that such constraints will be embodied in any future external financing arrangements.

AGL – Gas and Electricity Retailer

163               AGL was formed in 1837. It was a private company manufacturing gas from coal. Over the years that followed it expanded its operations to supply gas to residential, commercial and industrial customers in New South Wales and the Australian Capital Territory operating under statutory monopolies. The company first participated in the electricity industry upon its acquisition of a 50% interest in the Victorian electricity and distribution business, Solaris, in 1995. Since then it has substantially extended its business into other States and into electricity marketing. A brief chronology indicates the timeframe and scale of that expansion:

1. 1997 – AGL commenced the retail supply of electricity in New South Wales to commercial and industrial customers with an annual demand exceeding 160MWh.

2. 1998 – AGL acquired the remaining 50% interest in the Victorian electricity distribution and retail business Solaris, now known as AGL Electricity Ltd.

3. 2000 – AGL’s gas retail and distribution business in the ACT entered into two joint ventures with the electricity retail and distribution business of the ACT’s electricity, gas, water and sewerage utility. The two joint ventures trade as Actew AGL.

4. In 2000 AGL acquired, upon its privatisation, the South Australian electricity retailing business formerly conducted by Electricity Trust of South Australia.

5. From 2000 AGL supplied retail electricity in South Australia to customers including those consuming less than 160MWh annually.

6. In 2000 AGL commenced the retail supply of electricity in Queensland to medium to large sized commercial and industrial customers with an annual demand for electricity exceeding 160MWh.

7. In 2002 AGL commenced the competitive retail supply of electricity in New South Wales to customers consuming less than 160MWh per annum.

8. In July 2002 AGL acquired the Victorian gas and electricity retailing business, Pulse Energy Pty Ltd, which it has merged with its gas and electricity retailing business in Victoria and which trades as AGL Victoria Pty Ltd.

In addition to the above initiatives involving the electricity industry, AGL has expanded its activities in the retailing of gas.

164               AGL presently operates as a retailer, wholesaler and trader of gas, electricity and energy related services selling electricity and gas to approximately 2,900,000 domestic and small business customers and approximately 5,000 commercial and industrial customers not including those serviced by Actew AGL. It has current scheduled electricity generation interests in Victoria and South Australia, being:

1. The Somerton Plant in Victoria which has a generation capacity of 150MW and is owned by AGL Power Generation (Victoria) Pty Ltd.

2. The Hallett Plant in South Australia which has a generation capacity of 180MW and is owned and operated by AGL Power Generation (SA) Pty Ltd.

AGL is publicly listed. It has three main operating divisions:

1. Energy Sales and Marketing (ES & M) which has responsibility for AGL’s electricity and gas marketing, sales and customer service activities.

2. Agility, which manages energy infrastructure assets including gas pipelines, electricity distribution networks and electricity generation plants owned by AGL; and

3. Networks representing AGL’s ownership interest in gas and electricity distribution networks. AGL Energy Networks consists of two regulated businesses, AGL Gas Networks in New South Wales and AGL Electricity Networks in Victoria.

There is also a Corporate Division responsible for managing AGL investments beyond ES & M. These include its electricity and gas network investments, its interest in the Australian Pipeline Trust, its 50% interest in the liquefied petroleum business Elgas, and overseas interests in New Zealand (NGC Holdings Ltd) and Chile (Gas Valpo). The ES & M Division comprises five operating units:

(a) marketing and residential sales;

(b) business sales;

(c) wholesale energy;

(d) finance and administration; and

(e) business systems group.

Mr Michael Fraser, who gave evidence for AGL, is the Group General Manager of ES & M. Approximately 1,200 AGL employees work within ES & M. Generally I accepted his evidence. He did present as something of an advocate for his company’s cause, seeking to engage the Bench at times with non-responsive answers to counsel in order to make a point which he felt should be made. I do not criticise him for this. He presented as a natural advocate and was honest and forthright in that presentation. But I treated his evidence, when it involved elements of characterisation or evaluation in AGL’s interests, with some caution.

165               Mr Fraser described ES & M’s future strategy by reference to what he called ‘five indices of success’ set out in a document entitled AGL 2003 Corporate Strategy Review and Update and dated 1 May 2003. The indices are:

(a) External influence: Successfully influencing the regulatory and political outcomes across States and fuels;

(b) Wholesale leadership: Securing competitive long-term sources of supply;

(c) Scale Leverage: Translating the benefits of scale leadership into cost leadership;

(d) Retail Effectiveness: Winning the competitive war for customers without damaging margins;

(e) Predictable Performance: Delivering consistent, profitable growth year on year.

166               Mr Fraser described AGL’s marketing and competition strategies in retailing electricity and gas. He said, and I accept, that to maintain its existing customers and gain new customers, AGL constantly monitors the prices and terms upon which other retailers offer to supply customers in the regions in which it competes to assess whether those offers are ‘better’ than AGL’s. In comparing prices and terms he has regard to both price and non-price elements including the provision of auxiliary services, brand and market perception and customer loyalty. AGL’s staff monitor, as far as practicable, the extent to which existing retail customers in Victoria switch to competing retailers. By reason of competitive activity in the market place, AGL loses a number of customers to other retailers through the process of churn discussed earlier.

167               AGL adopts different marketing strategies depending on the customer segment under consideration, including both price and non-price strategies and appeals to customer loyalty. In relation to residential customers the strategies may include the offer of rebates and green energy – generated by means such as wind power which produce relatively lower carbon emissions than other means - and appeals to emotive factors. AGL offers related customer services, a variety of bill payment systems including direct debit and bill smoothing and operates retail stores in New South Wales and South Australia which sell energy appliances. It operates call centres to handle customer inquiries, bill payment, connection and disconnection requests. AGL supplies well over 800,000 Victorian residential and small business customers and well over 4,000 Victorian industrial and commercial customers.

168               AGL prepares weekly reports comparing its performance against that of other retailers. It collects information on other retailers’ sales and marketing activity by reference to their product offers, geographic areas targeted and sales channels used. This is done in a variety of ways including door-to-door and telesale activities, surveys of customers switching from AGL to other retailers and information from AGL employees approached at home by competitors. Mr Fraser continually monitors and compares competitor prices against those of AGL. He has directed staff within AGL to prepare an AGL Electricity Product Analysis listing the types of product, customer volumes, total consumptions and prices for various customer contract offers within Victoria.

169               Retail customers in New South Wales, Victoria, South Australia and the ACT can chose among a number of different types of contracts and tariffs offered by the competing retailers. As a former ‘Incumbent Retailer’ (sometimes called a Host Retailer) AGL must offer to supply electricity to domestic and small business customers in certain regions as a condition of holding a Victorian Electricity Retail Licence. Those offers are governed by tariffs and terms and conditions determined by the Victorian industry regulator, the Essential Services Commission, and published by the licensee in the Victorian Government Gazette at least two months before they take effect. Such contracts are called ‘Standing Offer Contracts’. Although originally there were five Incumbent Retailers mapped into the five distribution regions defined upon privatisation of the Victorian electricity industry, there are now only three remaining, being AGL, Origin and TXU. A significant proportion of AGL’s Victorian residential and small business customers are supplied electricity on the basis of Standing Offer Contracts as at 31 August 2003.

170               Any licensed retailer can offer a market contract which is a contract for supply between the retailer and customer without any requirement for its prior publication. However, under s 36 of the Electricity Industry Act 2000 the terms of such contracts must be consistent with terms and conditions set out in the Electricity Retail Code published by the Office of the Regulator-General on 31 October 2001. The Regulator-General was the predecessor of the Essential Services Commission.

171               As a former Incumbent Retailer, AGL is also required to provide default and deemed terms and conditions and tariffs to residential and small business customers within its previous host retail regions. Deemed contracts are applicable to customers who, following the introduction of full retail competition, neither transferred retailers nor moved on to a market contract. Default contracts apply where a customer occupies premises and starts to take supply of electricity without entering into any contract.

172               Mr Fraser exhibited to his affidavit copies of recent AGL Market Contract Offers. He also exhibited a copy of the first TXU Dual Fuel Market Contract Offer which offered a $100 upfront rebate and a $110 rebate per annum continuing throughout the term of the contract. Another TXU Market Contract Offer offered a $160 rebate upfront with a $110 annual rebate. Also exhibited was an Origin Energy ‘Green Energy’ Market Contract Offer. A recent EnergyAustralia Market Contract Offer included rates the same as standing offer tariffs with up to $200 in rebates over a three-year period. Energy Australia entered the Victorian region in May 2003 and has been increasing its activity since that time. A current Country Energy market contract, also exhibited, offered electricity to customers at rates less expensive than the regulated standing offer tariffs. A Powerdirect Market Contract made similar offers.

173               AGL has experienced significant ‘churn’ in its customers. It has regard to ‘net churn’ statistics which enable it to determine its net customer gains or losses. In a confidential annexure to his affidavit Mr Fraser showed dual fuel energy churn data for customers consuming less than 160MWh annually as at 2 July 2003. He described the retention and acquisition of customers as an important strategic aspect of ACL’s business. No doubt such a strategy has universal appeal to retailers in any market for goods or services. Similar strategies are undertaken by other retailers in the South-East Australian retail markets. He observes, and I accept,that there is a correlation between periods of intense competitor marketing and churn away from AGL. Not long after AGL acquired the Victorian electricity retail business of Pulse Energy in July 2002 its retail competitor, TXU, commenced an integrated campaign to win over former Pulse small customers. The campaign involved door knocks, press advertising and letter box drops. AGL sent its own doorknockers into TXU retail areas to try to win over its customers. It is not necessary to set out the content of the confidential tables demonstrating the volume of customers transferring to and from AGL and other electricity retailers to conclude that there is a significant level of customer transfer between retailers in Victoria.

174               One of the confidential annexures to Mr Fraser’s affidavit set out AGL’s total Victoria and NEM-wide retail customer numbers for 2001 to 2002 and its Victorian and NEM-wide load for that year. This was by reference to National Metering Identifiers which are meters identifying individual sites receiving electricity supply. Although they do not necessarily correlate exactly to physical premises, as one large factory might be allocated several NMIs, they are generally considered a reasonable approximation to retailer/customer numbers. In the period covered by the annexure AGL’s market share, to the extent it was represented by the number of retail customers, was in excess of 20% in the NEM and was about 35% in Victoria. The total volume of electricity sales to Victorian retail customers measured as a proportion of the approximate volume of electricity supplied by AGL to its Victorian retail customers to the total electricity consumed in Victoria was about 25%. Although contained in a confidential annexure these figures were pleaded at par 37 of AGL’s statement of claim. Having regard to these numbers and the percentages of the quantity of electricity which it sold in 2002 in both markets, it is clear that AGL is a major competitor in retail markets for the supply of electricity.

175               The wholesale acquisition of electricity is undertaken in AGL within the Wholesale Energy Group of ES & M. Mr Fraser is responsible to the Managing Director of AGL and the AGL Board for its wholesale energy trading activities. He described AGL’s participation in the NEM and described the general operation of the NEM in terms which have already been covered. AGL is a registered Market Customer and a Generator in the NEM under Ch 2 of the Code. AGL, like other registered retailers, acquires all its electricity requirements for its retail supply business through the NEM. The NEM pool price which it is required to pay as a retailer is, as already observed, very volatile. In calendar year 2002 AGL paid $4,906.09/MWh in one half hour period and negative $228.01/MWh in another in the course of purchasing electricity for its Victorian customers. Mr Fraser produced graphs demonstrating the average of the daily Victorian pool price for 2002 and the daily average minimum and maximum pool prices in Victoria for the same year. These showed substantial spiking at particular times.

176               As he said, the volatility in the pool price represents a substantial business risk to AGL and other retailers. The risk arises from the asymmetry between pool prices and contracts with end use customers which typically have fixed prices. Variations in NEM pool prices could cause AGL to sustain substantial trading losses. Mr Fraser gave an example of a combination of events involving reduced output by generators in one State and extremely hot weather leading to very high pool prices in another.

177               Like other retailers, AGL pursues a number of strategies to manage its exposure to NEM spot price volatility. These include:

1. Entry into electricity hedge contracts with generators and other counter-parties.

2. Acquiring or developing generation assets or entering contracts with generators under which the retailer can determine when and how much electricity a generator will dispatch. The contractual arrangements are called dispatch or master hedges or power purchase agreements. Under such agreements the retailer pays the generator for this capacity and in return is entitled to determine, such to physical constraints, when the plant generates and receives the NEMMCO payments to the generator. AGL’s acquisition of generator capacity is an aspect or outgrowth of its consideration of a ‘vertical integration’ strategy which is discussed later in these reasons.

3. Agreements with customers to adopt part of the wholesale price risk or otherwise influencing customer demand patterns – demand side management.

AGL has pursued all of these strategies. It has built two generators, Somerton and Hallett, in Victoria and South Australia respectively. It seeks to negotiate contractual arrangements with large industrial customers under which they adopt part of the wholesale price risk and vary their demand accordingly.

178               Mr Fraser described the electricity hedge contract as the principal mechanism by which AGL manages its exposure to the volatility of the pool. I accept his evidence in this respect. It is reflected, inter alia, in the risk management policies and systems which AGL, like other energy retailers, adopts. Its most recent policy is the Wholesale Risk Management Policy which it adopted in November 2002. It was developed with the help of Macquarie Risk Management Services which is part of the Macquarie Bank Group and was retained by AGL so it could have the benefit of understanding best practice in risk management learning. The Risk Management Policy adopts controls which allow AGL’s management to run the business within parameters determined by the Board. Risks for which specific controls have been adopted fall into obvious categories and although the Risk Management Policy itself was a confidential exhibit, there is no reason why the categories of risk should not be identified in these reasons. They are:

(a) Market Risk

(b) Credit Risk

(c) Contract and Legal Risk

(d) Operational Risk

(e) Regulatory Risk

There is a Risk Management Committee of which Mr Fraser is Chairman. It has the responsibility of overseeing the risk management operations of any AGL entity with wholesale energy market exposure. That includes the setting of trading limits and ensuring compliance with the Energy Risk Management Policy. An Energy Risk Management Framework sets out the specific controls, procedures and authorities for managing the risks referred to in the Risk Management Policy.

179               Having regard to the nature of the NEM and the evidence of witnesses from participants in the NEM, including Mr Fraser, I am satisfied that risk management is a central feature of the operation of both wholesalers and retailers in the electrical supply industry in the NEM. Its dominance derives from the volatility of the pool price and the very substantial losses that can be sustained if risk is not properly managed. It necessarily impacts upon competitive activity in the NEM and is an important feature of any market to be considered for the purposes of competition law.

180               Within AGL’s Wholesale Energy Group there are two units which undertake electricity trading activities. The first is the Structure Desk which carries out market analysis and forecasting. The second is Energy Trading which manages AGL’s exposure to the NEM spot price. The Risk Group within the Finance and Administration Group of ES & M supervises the compliance of the Structure Desk and Energy Trading with the Risk Management Policy. At the Structure Desk there are about 35 people responsible for a number of tasks including continuous forecasting and monitoring of demand through the NEM in each region and by portfolio of AGL customers. Of key importance to its analysis is the information published by NEMMCO every half hour. That information includes forecasts of demands, whether interconnects between the regions will be constrained and if so by how much, and the pool price that would be achieved based upon the bids made to NEMMCO at that time. The information is referred to as NEMMCO Pre-dispatch Data. NEMMCO also publishes information after dispatch which includes each of the bids by each generator, the actual demand, the actual interconnect constraints and the actual prices. There is communication between the Structure Desk and Energy Trading who manage the relationship between NEM pool price exposure and AGL’s portfolio of hedge contracts.

181               AGL enters into hedge contracts with generators including those outside Victoria. Mr Fraser’s evidence included a confidential table setting out the number of transactions entered into by AGL at the Victorian Reference Node by contracts bought and sold. Another table detailed counterparts who have sold AGL head contracts referenced to a node where they do not own scheduled generation capacity. Although the number of contracts is substantial, the totals do not show the timeframes over which the transactions occurred.

182               In determining the price at which it enters into hedge contracts, AGL has regard to a number of matters which are obvious enough. They include the conduct of generators, past experience of negotiations with generators and retail pricing practices with customers. It also takes account of bid and offer price information projected for swap hedging coverage for periods into the future and published by a computer service.

183               In determining the profile of its electricity hedge contracts, AGL has regard to its customer electricity requirements. Mr Fraser identified four key characteristics of customer demand affecting contracting strategies:

(a) Different customer segments have different demand profiles.

(b) Customer demand, particularly residential and small business customer demand, is materially affected by weather.

(c) It is therefore generally impossible to forecast actual demand.

(d) The retailer must pay NEMMCO for electricity consumed by its customers in a particular region at the pool price determined in that region.

While industrial and commercial customers have a fairly flat and predictable demand profile, residential and small business customer demand is far more volatile. It peaks during certain periods of the day, for example on summer afternoons between 3pm and 7pm. The variability of residential and small business customer demand was illustrated by graphs contained in Mr Fraser’s affidavit. For what appears to be one day in July 2003, they showed customer demand over a 24 hour period. They generally demonstrated that small customer demand profiles were more variable than large customer demand profiles. Variability of demand by season and State was also demonstrated for 2002.

184               Although the Incumbent Retailers in Victoria, prior to the introduction of full retail contestability, had the benefit of ‘vesting contracts’ to hedge the load of residential and small business customers in Victoria. These were terminated on 31 December 2000.

185               The difficulties of accurately forecasting demand mean that retailers, including AGL, frequently find themselves over-contracted or under-contracted for any particular trading interval. Where a retailer is under-contracted at times of high pool prices it will incur substantial costs which cannot be passed on to customers. If it is often over-contracted then its average cost of purchasing electricity will be higher than optimal. This class of risk is described as ‘load shape risk’. Load shape risk is managed by maintaining a portfolio of contracts which provide different types of contract cover (swaps, caps and options) at different levels of demand and price. Generally periods of high demand correlate with periods of high spot prices.

186               Load shape risk may also be managed by negotiating more sophisticated hedge contracts which respond to a variety of circumstances. One such contract is the DPHA which AGL entered into in 2002 with the LYPM. The terms of the contract are confidential but it has been agreed that it can appropriately be described in open Court thus:

‘The DPHA is a long-term swap contract between AGL and LYP. The contract is intended to provide partial hedge cover in respect of AGL’s Victorian customer load comprising customers who consume less than 160 MWh per annum (residential and small business customers). The strike prices are set for the first year and thereafter may be adjusted having regard to market conditions and other confidential criteria. The contract provides for a flexible mechanism for determining volume. The customers load covered by the DPHA represents approximately 45% of AGL’s total Victorian peak demand.’

 

AGL does not put all its eggs in one basket. The extent to which it contracts with any one party is restricted by the terms of its Risk Management Policy intended to prevent concentration of risk with any one counterparty and exposure to credit worthiness risks. In this respect the DPHA required special Board approval.

187               Generally speaking, generators will enter into contracts which specify the reference node of the region in which the generator is located. Contracts specifying regional reference nodes in a region in which the party does not have customer load or a generating plant involve the assumption of what is called ‘basis risk’. This is the financial risk that pool prices at the two reference nodes may diverge because the interconnect between the two regions is constrained. Nevertheless such contracts are sometimes made. There is the mechanism of the inter-regional settlement residue auction which provides a partial hedge against basis risk.

188               AGL has not had a lot of difficulty in obtaining hedges from generators located in Victoria. Mr Fraser referred in this respect to developments within the NEM over the past two years including the augmentation of Snowy interconnect, the construction of the Somerton generator and new Victorian capacity, the commissioning of QNI and additional generation capacity in Queensland and South Australia. In his view a recently proposed 200MW upgrade of the Snowy interconnector and the commissioning of Basslink between Tasmania and Victoria in 2005 and the proposed 70MW upgrade to Loy Yang A will significantly increase capacity into Victoria. On that assumption, a material increase in the prices charged by Victorian generators would be met by retailers hedging Victorian load with generators located in other States. Inter-regional settlement residue units could be used to partially manage the resulting ‘basis risk’. There would always be the possibility for a retailer, such as AGL, augmenting its generation capacity while considering the purchase of transmission rights at an interconnector.

189               Mr Fraser contended that, at present, competitive constraints limit the ability of Victorian generators to inflate the Victorian pool price unreasonably. This was a matter in contention so far as LYP is concerned. An issue to be resolved is whether LYP presently has market power in the sense of an ability to sustainably and profitably increase the pool price. It is however useful here to refer to the evidence collected by Mr Fraser which was said to exemplify generator bidding and dispatch conduct based upon NEMMCO data. These examples of bidding and dispatch behaviour by Victorian generators were said to demonstrate the competitive influence of South Australian and New South Wales generators and the interconnectors on Victorian generators. The examples were as follows:

1. 21 May 2003 - for the half hours ending 1800, 1830 and 1900 Loy Yang Power offered 202MW into price bands greater than $100 up to price bands less than $4,000/MWh. Consequentially 198MW less was dispatched from Loy Yang Power and interconnect flows into Victoria from the Snowy region increased by more than 200MW for the half hours ending 1800 and 1830. The pool price during these periods were $64.84/MWh and $48.96/MWh respectively.

2. 29 July 2003 – in the half hour ending 1830 Edison Mission Energy offered 70MW of its Loy Yang B power station in price bands exceeding $1000/MWh. During that period output from Loy Yang B reduced by 60MW with imported electricity from South Australia increasing by 33MW. The pool price in Victoria for that half hour was $114.61/MWh.

3. 31 July 2003 – Hazelwood Power offered 20MW to NEMMCO at prices in excess of $9,000/MWh and a further 160MW at prices greater than $100/MWh. This increase in the offer price resulted in output from Hazelwood Power being reduced by 66MW in the half hour to 1830 and a further reduction of 58MW in the following half hour. Interconnect flows from South Australia increased during this period by approximately 93MW. In the half hour ending 1930 imports from South Australia decreased by 146MW while output from Hazelwood increased by 84MW as a result of Hazelwood offering an additional 173MW at less than zero. The pool price during this period peaked at $196.10/MWh.

I am satisfied that where interconnector constraints are not applicable the availability of electricity to be dispatched from one region into another has the potential to limit the ability of generators in one region to materially increase prices above those in another region. Mr Fraser accepted that the ability of interstate generators to compete with generators in Victoria for physical dispatch by NEMMCO could be constrained by the capacity of the interconnectors into Victoria. The capacity of interconnectors into Victoria in the annual average peak periods defined as the period between 7am and 10pm during a working day week, represents 40% and for the annual average for all periods 44%. It represents 30% of maximum demand for the financial year 2002/2003.

190               Where price divergence occurs between regions in conditions of constraint, AGL will participate from time to time in the IRSR auction as a risk management tool to support hedging customer load in one region with a hedge contract referenced to the reference node in another region. Entitlements to IRSR units or firm inter-regional hedging instruments, offered by operators of non-regulated inter-regional interconnectors such as Directlink, allow a generator or retailer to offer to buy out-of-region hedging coverage with a mechanism for managing basis risk.

191               By way of investment in physical generation capacity, AGL owns two generation plants which are classified as scheduled market generators under the Code and which make offers into the pool and are dispatched by NEMMCO. These, which have already been mentioned, are Somerton and Hallett with generation capacities of 150MW and 180MW respectively. AGL built Somerton because it deferred AGL Electricity’s need to augment connections to the transmission grid and increased capacity within the network without requiring the construction of a new terminal station. It mitigated the risk in the retail electricity business and provided wholesale electricity risk mitigation in so far as control of Somerton’s capacity would provide a natural hedge for part of ES & M’s customer load. This was an element in AGL’s ‘vertical integration strategy’ considered below. The Somerton Plant began operating in 2002. It comprises four 37.5MW open-cycle gas-fired generating units. AGL Electricity Ltd is registered with NEMMCO as an intermediary in respect of the plant and acts in that capacity in relation to bidding in the plant and receiving revenues that result. It acts under the direction of ES & M. In Mr Fraser’s experience, in-house generation has an important role to play in maintaining an efficient portfolio of risk control products. Generation assets fully integrated into the retail business have advantages over hedge contracts. They enable a retailer to alter the offers of plant into the market in response to changes in the expected demand of its customers. And while hedge contract prices and volumes are agreed, based on prevailing assumptions about retail and wholesale competitive circumstances when the agreement is struck, prices and volumes under those contracts cannot easily be altered each time one of the assumptions changes. With AGL’s own generation it can continuously revise generation practices to account for changes in circumstances. Moreover hedge contracts expose AGL to risks that it cannot readily control such as the financial standing of the counterparty. A significant proportion of ES & M staff time is devoted to the negotiation of hedging contracts. Costs are involved in documenting the agreements that are reached and resolving disputes that may arise.

192               AGL’s interest in developing its own generation capacity informed Board papers dealing with a vertical integration strategy which was the subject of considerable cross-examination and comment in closing submissions. This is discussed below. Notwithstanding the benefits of AGL’s acquisition of generation capacity it determined in 2002, following a strategy review, that its primary physical generation needs lay in the peaking and intermediate range. For base load demand its preferred retail risk cover in Victoria is through hedge contract coverage. The two reasons given for this by Mr Fraser were:

(a) There is excess base load capacity in Victoria entailing that AGL would not need surplus base coverage; and

(b) The DPHA adequately addressed AGL’s base load demands within that State.

AGL therefore concluded that its interests in Victoria would best be served by vertical integration in immediate and peak generation but that base load price risk would best be controlled through hedging contracts. This preference was said to be a function of the ready availability of appropriate contracts for base load capacity in Victoria and of the fact that plants with intermediate and peaking capacity use technology with a very short ramp up period and therefore better respond to sudden spikes in the pool price. According to Mr Fraser, the acquisition of Loy Yang or even a minority interest in Loy Yang was not part of AGL’s optimum Victorian risk management profile.

AGL’s Vertical Integration Strategy

193               AGL’s proposed acquisition of an interest in the Loy Yang power station was incidental to its consideration of a vertical integration strategy in the NEM. The development of such a strategy was considered at Board meetings in 2002 and 2003. The ACCC points to that consideration as an indication of a tendency to vertical integration by market participants which will be pursued by others if the acquisition proceeds. The apprehended anti-competitive effect of consequential integration forms a basis for its contention that the proposed acquisition is likely to substantially lessen competition. The ACCC contends that AGL’s internal Board papers show it to have been concerned, since at least June 2002, about the competitive implications for itself as a retailer if other companies were to vertically integrate.

194               The first of the papers in evidence was prepared by Mr Fraser for a meeting of the AGL Board held on 5 June 2002. It proposed that, in light of the then state of development of the NEM, there were significant commercial and strategic benefits to be achieved through power generation ownership, actual or ‘virtual’. At that time, which was prior to the commissioning of the Somerton and Hallett plants, AGL was a net retailer of electricity with only minimal access to peaking plant. It relied heavily on contracts with counter-parties in the NEM to manage its exposures to the spot market and was more often than not a price taker paying excess premiums to maintain its position in the retail market.

195               Mr Fraser acknowledged that the initial establishment of the NEM had resulted in substantial efficiency gains most of which were passed directly to larger industrial and commercial customers. ABARE have estimated these gains to be in the order of $1.5 billion. But that outcome was achieved in an over-supplied market that masked the synergy benefits that had been lost during the deregulation and disaggregation of the industry. He characterised the summers of 2000 and 2001 as periods when demand and supply were almost at equilibrium and flaws in the market started to become apparent. He said:

‘Certain generators were able to exercise their market power in a tight market by holding back contracts and aggressively bidding their generation assets into the spot market. This had the effect of significantly increasing the spot price of electricity from prior years. A number of market makers, speculators and intermediaries subsequently exited the wholesale market, impacting on the depth of market liquidity.’

This was evidenced in the summer of 2000/01 when Loy Yang’s bidding significantly increased spot prices. Whether or not that reflected an exercise of market power is a matter for determination later.

196               The paper referred to a study commissioned by NEMMCO’s Board in which the Bach Consulting firm had identified three issues having a major impact on the NEM:

. Lack of depth in high voltage transmission interconnection between States. This was said to have the impact of creating supply/demand imbalances within the specific markets, providing regionally located generators with significant market power.

. Regulatory uncertainty and significant political intervention.

. The absence of an integrated liquid, spot, exchange traded and over the counter market.

Under this last heading Mr Fraser said that where liquidity is an assumed measure of reducing risk, illiquid market outcomes would lead to a strategy of vertical integration to manage the significant levels of risk associated with industry participation. His paper went on to point out that AGL’s two major retail competitors, TXU and Origin Energy, had been establishing a portfolio of generation assets. TXU had control of 2,240MW of generation capacity (1,000MW retail) while Origin Energy’s generation portfolio was at 300MW (2,000MW retail) and growing. He referred to Australian National Power, the owners of the Hazelwood, Pelican Point and Synergen power stations, as having been short-listed to acquire the Pulse and Citipower retail assets. If successful they would be provided with a significant vertically integrated energy business with 1.1 million customers and 2,320MW of generation located in Victoria and South Australia.

197               The benefits of vertical integration were then set out:

. Cost savings associated with operating two services together rather than separately.

. Access to markets – ownership of a customer base in conjunction with a wholesale business could be valuable in terms of being able to recognise and quickly exploit, as first mover, new market or product opportunities.

. Diversification of risk – vertical integration could diversify risk by providing a portfolio of businesses. The integration of upstream and downstream businesses created a ‘natural hedge’ against variations in the impact of input price variations on profitability.

. Risk management – vertical integration could minimise exposure to uncompetitive market outcomes particularly in the case of the electricity market if there were a tight supply demand balance and/or concentration of ownership in conjunction with competitors having vertically integrated operations.

198               The paper considered the possibility of acquisitions in peaking, intermediate and base load generation capacity. Mr Fraser expressed concern in the paper that the concentration of ownership/trading rights of existing base load generation could have serious ramifications for AGL and its retail position in the NEM. If either Origin or TXU were to acquire significant base load generation assets in regions where AGL is a predominant retailer, there was a danger that AGL would be exposed to hold out risk as the market tightened. If Australian National Power were to move downstream through the acquisition of Pulse, AGL’s retail assets could be at risk. This had happened with NGC in New Zealand in which it had a majority interest. In a tight market there is a significant value transfer from retail to base load generation. The conclusion of the paper was that in the near to medium term AGL would be able to source adequate hedge contracts in combination with its Somerton and Hallett projects to protect its retail load. It would become increasingly important however to further develop a portfolio of physical and vertical generation assets in order to optimise business outcomes.

199               In cross-examination Mr Fraser agreed that generally speaking one of the factors that AGL would consider in deciding whether or not to pursue its own vertical integration strategy was the pre-existing level of vertical integration in the market between generator and retailer. He and his colleagues had thought about this in the context of the South Australian market. It was a factor that they had also considered in relation to the Victorian market but it was not a particularly important factor.

200               Mr Fraser agreed that at the time the June 2002 paper was written he had a serious concern about the pre-existing concentration of ownership and trading rights in base load generation in regions where AGL was a predominant retailer. He also agreed that it was more advantageous to AGL to be in a market that was less vertically integrated. If AGL were to be as vertically integrated as the next competitor it would not be a matter of concern. If it were not vertically integrated and other players were, then that would be a cause for concern. If AGL were not vertically integrated and others were, it would seriously consider pursuing its own vertical integration strategy.

201               There was reference to NGC, the Natural Gas Corporation, a retailer in the New Zealand market in which AGL had a majority interest. Mr Fraser agreed that one of NGC’s problems in New Zealand was obtaining forward contract cover. A factor which had affected its ability to get forward contract cover was the number of vertical integrations in that market. Another factor was drought in a market heavily reliant upon hydro power. Market participants in New Zealand in these circumstances were unwilling to sell contract cover to NGC. They kept their contract cover to protect their own retail position. After some pressing in cross-examination Mr Fraser accepted that, where vertically integrated generators in New Zealand had a choice, in a drought situation, about who to contract with, his understanding was that they did not want to contract with unintegrated retailers such as NGC. AGL had ultimately sold out of New Zealand and had suffered substantial losses on its investment in NGC. The quantum of the losses was in the vicinity of $275 million. In considering a vertical integration strategy for AGL for South Australia and Victoria in June 2002, Mr Fraser had fresh in his mind the New Zealand experience. He was concerned about the possibility that a similar situation could arise in South Australia or Victoria if AGL were not vertically integrated in either of those markets.

202               The issue was revisited at a meeting of the Board on 4 December 2002 when a further paper was presented. The document, entitled ‘Power Generation – Strategy Update and Opportunities’, referred to the June paper. It identified key investment objectives involved in the pursuit of a vertical integration strategy. These were:

. Reducing the risk profile for AGL and reducing the volatility of its earning.

. Addressing and overcoming imperfections, constraints and market failures which exist in the NEM.

. Building a scale position in the power generation sector and complement this position by developing organisational capabilities within AGL in power generation; and

. Creating a new platform of growth for future investments by AGL.

The issue under review, as part of ongoing strategy development, was the balance between physical ownership and ‘virtual ownership through contractual means’. The work done to that date had focussed on AGL’s market position in South Australia and Victoria. The purpose of the paper for the Board meeting of 4 December was to update the Board ‘… on the emerging themes of AGL’s power generation strategy review and to seek Board approval to actively pursue one of four potential power generation opportunities’.

203               The paper identified some key problems facing AGL. These were:

1. The NEM was not a truly national market so that hedging strategies had to be done by AGL on a State-by-State basis rather than a whole of market basis.

2. Government owned players in New South Wales and Queensland had created an uneven playing field.

3. AGL was facing deteriorating creditworthiness from counter-parties in the NEM.

4. AGL’s pursuit of investments in power generation was consistent with the addition of tangible assets to its balance sheet to match the intangible assets acquired as a result of its growing retail business.

The paper reported that market analysts predicted that the vertical integration business model was a more sustainable business model for an energy company in the longer term. Recent consolidation in the United Kingdom market had seen a ‘… steady path to a more vertically integrated business model by energy market players’. One of the acquisition opportunities mentioned in the paper was Loy Yang Power.

204               A further document relevant to AGL’s consideration of vertical integration options was dated 22 January 2003 and entitled ‘Winning Energy Company Initiative – Vertical Integration/Power Generation Strategy Review’. It was produced by a number of people but its designated ‘Sponsor’ was Mr Fraser. Mr Fraser, when asked in re-examination about the purpose of the review, said that AGL was considering what its strategy ought to be with respect to power generation as a form of investment and as a way of ‘covering off’ wholesale market risk. To this end, AGL was to consider what mixture of contracts and potential acquisitions or construction of plant it should put in place to balance its portfolio and whether it should make any stand alone investments with respect to generation. The Executive Summary, at the beginning of the paper, observed that growth in retailing in the main markets in which AGL operates was constrained by the maturity of the market, continued government ownership and ACCC limitations. Risks were increasing due to use of supplier power and reducing counter-party creditworthiness. Power generation was said to offer an avenue for growth in the non-regulated energy sector and significant benefits in managing AGL’s electricity market exposure. International trends and internal analysis indicated that vertically integrated structures extending electricity retailing into power generation provided long term competitive advantages by capturing value across industry segments and reducing risk by providing a natural hedge.

205               AGL, it was said, needed a carefully crafted strategy to ensure it executed the right types of deals with generators and retained control of key assets up to but within competition regulation restrictions. It needed to ensure it negotiated supply arrangements covering its risks but did not allow others to benefit from reduced market pricing arising from its arrangements. The summary said:

‘In essence, AGL has an interest in seeing volatility in the market that is has the capability to manage and can obtain a margin for doing so.’

206               In reference to the South Australian market AGL, by hedging fully, ensured a reduction in volatility in the spot market that in time would translate to a lower forward contract curve. Competitors would then see a more favourable market in which to increase their market share. There was therefore said to be:

‘… an incentive for AGL to either vertically integrate or to contract in a way that manages risk but does not unduly dampen spot market volatility and associated contract prices.’

AGL’s position as the Incumbent Retailer in South Australia with a 75% market share meant it was forced to deal with virtually all generators and so was subject to hold out risk. This is the risk that a generator will withhold capacity in order to push up prices. Because of this AGL operated as a price taker in all sectors of the market in peak, intermediate and base load. It was, at best, in a position of needing to contract with two out of three generators and at worst faced with dealing with all suppliers.

207               The review document stated that the hedging of AGL’s long-term forward electricity needs was becoming more difficult because of major uncertainties in the market including:

- limitations imposed by the deteriorating creditworthiness of generators

- different approaches adopted by a myriad of regulators controlling the energy industry

- the potential for increased interconnections to reduce constraints between various regions promoting the use of lowest cost generation

- the impact of national and State greenhouse gas regulations on the costs of existing generators and the timing and type of new entrants

- the imposition of price caps in deregulated markets, the continuation of government mandated hedging arrangements, ongoing industry ownership by governments and restricted deregulation in the Western Australian, Queensland and New South Wales markets.

The document stated that AGL’s main retail competitors were already well advanced in implementing vertical integration strategies. The New South Wales and Queensland governments had also adopted vertical integration arrangements for their Government Owned Corporations (GOC) supplying domestic customers. Under the heading ‘Background and Key Study Objectives’, the document observed that under a vertical integration strategy, AGL would obtain control of dispatch of generation assets. Managed in an integrated fashion with its retail portfolio these should yield benefits beyond those available by arms length arrangements such as hedging. The capability had already been demonstrated by the use of AGL’s Somerton and Hallett plants to set market price and produce effects on its power portfolio purchase price. Vertically integrated generation was distinguished from investment in generation that would not be actively managed as part of AGL’s trading activities. It was said:

‘Examples of this type of investment would include minority investment in power projects or investment in plant with its output fully contracted or not in regions where AGL has a retail presence.’

It was foreshadowed that expansion of the retail activities of TXU or Origin in South Australia or the entry of a generator such as International Power to the retail area could exacerbate the transfer of wealth to generators as a result of low market liquidity.

208               If a vertical integration strategy were demonstrated to have significant advantages to AGL the subsidiary issue of whether it should be implemented by asset ownership or contractual arrangements which would provide ‘virtual ownership’ had to be addressed. Considering the quantification and timing of supply side needs in South Australia the paper, relying upon modelling by Frontier Economics, contemplated the acquisition of peak and intermediate generation capacity. Acquisition of base load capacity was not recommended as new plant was currently less attractive than contract purchases. If existing base load capacity could be acquired at or less than assumed contract prices then AGL’s net benefits would increase. As to Victoria, it was said that the acquisition of 1,500MW of brown coal capacity could be justified. The bulk of this supply could be obtained under existing arrangements with LYP.

209               Potential savings to AGL, if it were to restructure as an optimally integrated generator/retailer in South Australia and Victoria, were set out based on Frontier Economic’s modelling. Reference was also made in the paper to AGL’s use of the NEMMCO Settlement Residue Auction Instruments to provide inter-regional cover, particularly into South Australia. This was described as sub-optimal cover due to the inherent basis risk and so avoided as a hedge although participation increased when a market opportunity arose.

210               Competition issues were considered and the possibility of arguments that might be raised by the ACCC against acquisition, namely that the ownership of a generator by a retailer might:

1. prevent its retail competitors from contracting for electricity with the acquired generator, thereby increasing barriers to entry and preventing competitors from obtaining efficiently priced hedges;

2. provide it with a ‘natural hedge’ against spot price volatility. If this provides a cost advantage over new entrants it may deter future competition in the retail sector.

The strength of these arguments was seen as dependent upon issues including generator market and retail market concentration, the relative sizes of the merging generator and retailer and the extent to which market dynamics might change in the short term including the building of new generation capacity/interconnectors and the ability for retailers to contract for capacity in other jurisdictions. The importance of the relative size of the merging retailer and generator was acknowledged. So when a retailer seeks to acquire a generator with capacity less than or equal to the retailer’s requirements the merger might not raise as many concerns. There is still as much spare capacity in the market as there was pre-merger. Where a retailer acquires a large generator which has capacity beyond its retail requirements it could protect its retail position by removing spare generation capacity from the market and driving up wholesale prices at the expense of its retail competitors.

211               The paper referred to contractual power station arrangements. It was said that during periods when demand is greater than expected it is possible for generators, particularly those with market power, to withhold capacity or rebid at a higher price. Although for very high prices there is some demand response the reduction is small as most loads are inelastic. AGL could therefore face very high prices for any uncontracted load. If AGL controlled, via ownership or contracting, particularly fast response, flexible intermediate and peaking generation it could ensure that it was bid in at prices to cover its position precisely for each half hour. It could also use its control of generation to influence spot prices at other times. The outcome would be an ability to have some influence in forward contract prices as well as spot prices. Dispatch hedging therefore had a strategic advantage as well as saving on under or over-hedging.

212               With respect to future trends it was pointed out that Origin, TXU and AGL had all embarked on vertical integration exercises. In all cases the plant developed or acquired had been flexible peak or intermediate generation able to influence pool price at times of highest market prices. These retailers were described as ‘the dominant parties’ in the competitive South Australian and Victorian markets. Retailers shielded from competition in New South Wales and Queensland had not adopted that approach. It was likely that the parties that emerged as successful in New South Wales and Queensland would also be integrated when government ownership was removed and over-capacity absorbed. Market cycle risk analysis showed that a pure retailer was significantly exposed compared to an integrated player. As a result, it was likely that, over time, remaining retailers would be vertically integrated, owning peak and intermediate generation. The corollary was that coal-fired base load plant would become dominated by non-retail companies willing to act as long-term price takers. This would probably include traditional generator companies but was also likely to attract long term investors. In that type of market most energy would be produced and sold by generators but integrated retailers would predominantly set price.

213               The authors of the paper recommended that AGL should definitely pursue vertical integration in South Australia. They distinguished the position in Victoria. The base load sector in Victoria was operating relatively efficiently with three counter-parties, any of whom could meet AGL’s needs. This market sector was regarded therefore, on first examination, as not being an area in which to vertically integrate. In practice, AGL obtained most of its base load cover in Victoria from Loy Yang A under the DPHA. The document went on:

‘Given the well-known financial stress on LYA, AGL may well consider an investment in LYA as part of a process to improve its credit worthiness. This would not however produce significant integration benefits as the market is already operating relatively efficiently in this region/sector. Investing in Loy Yang A may however be an effective use of AGL’s capital and should be appraised as an investment and any effects that it may have on this integration strategy should be considered as part of the investment case.’

And again, in relation to Loy Yang A, the document said:

‘It is well known that the LYA owners have had considerable difficulties in servicing debt. As a result CMS have written off their holding and Horizon’s share price implies a total enterprise value of only $112 million. AGL has discussed the option of taking a minority stake in a consortium to acquire LYA. Vertical integration in Victorian base load generation has not however been identified as a necessary step. Purchase of part of Loy Yang A may however be an attractive investment opportunity for AGL, separate from this vertical integration process.’

214               This document, in my opinion, provides a useful indicator of the informed opinions of people operating within AGL as a major market participant in the NEM. Like AGL South Australia’s submission to NECA on the rebidding rules, referred to later in these reasons, they are views expressed outside the framework of any adversarial debate with the ACCC which, at the time of production of the document, had not emerged. I regard it as evidence of a natural tendency on the part of major retailers in the NEM, which I find exists, to undertake some degree of vertical integration at the level of peaking or intermediate plant. In my opinion practical considerations which generate a commercial pressure in that direction will operate independently of any acquisition by AGL or opposition to any such acquisition by the ACCC. Indeed, Dr Small who was called by the ACCC to give evidence of the effect of vertical integration in New Zealand, and whose evidence is discussed later in these reasons, agreed with the proposition that there is a natural tendency to vertical integration in the electricity markets which did not depend upon any initiating event. This is because of the price risk.

215               Moreover, there are more ways of vertically integrating than by acquisition of the assets of, or shares in, another corporation. The construction of new generation capacity undertaken by AGL in respect of the Somerton and Hallett plants is an example. The vertical integration by acquisition of control of a generator is just one part of a spectrum of possible responses to the pressures that lead in that direction. These include a partial acquisition with constraints on control as adopted in the present case. They may include joint ventures or contractual relationships or non-contractual ‘arrangements’ which might attract the designation ‘virtual’ vertical integration. It is not to be assumed that such arrangements would necessarily contravene other provisions of Pt IV of the Act.

216               In light of the preceding it is a point of particular interest that as early as January 2003 officers of AGL were treating the acquisition of an interest in Loy Yang as an investment strategy rather than an aspect of vertical integration. In my opinion, that rationale bests reflects its commercial character and is the rationale which informed the decision by AGL’s Board to proceed.

AGL’s Development of an Acquisition Proposal

217               The acquisition of a partial interest in Loy Yang was first considered by AGL’s Corporate Development division as part of its business strategy role. Mr Fraser was Chairman of the Internal Loy Yang Due Diligence Committee in the early stages of discussion of the proposed transaction. He was not directly involved in the negotiations. In November 2002, AGL and CBA executed a Commercial Term Sheet to govern their relationship in participating in the consortium to bid for Loy Yang. That Term Sheet was amended when TEPCo joined the consortium. Mr Fraser was of the view that the DPHA and other arrangements executed in January 2003 would satisfy CBA and TEPCo that AGL had an existing hedging arrangement with Loy Yang which covered a significant volume of its capacity so that a further power purchase agreement would not be necessary. This appears to have been reflected in the Commercial Term Sheet referred to earlier which was executed on 9 January 2003.

218               The sequence of events leading to execution of the transaction documents was outlined by Mr Hayes, the Chief Financial Officer of AGL, in his affidavit of 9 October 2003.

219               Great Energy Alliance Corporation Pty Ltd (GEAC), the proposed vehicle for the acquisition, engaged JP Morgan in January 2003 to undertake a valuation of the Loy Yang Business based on certain modelling assumptions. JP Morgan also prepared a valuation for AGL based on AGL’s modelling assumptions which differed from GEAC’s. Both valuations were prepared on the common assumption that 75% of Loy Yang A’s output would be the subject of hedge contracts. Revenue projections for the valuations were based on three categories of electricity revenue:

. sales of electricity into the NEM according to forward spot pool price projections

. sales of electricity to AGL under the DPHA

. sales of electricity to multiple counter-parties under existing and new contracts.

220               At a meeting held on 5 February 2003 the AGL Board considered a proposal for the lodgment of a non-binding indicative bid for the Loy Yang power station by the proposed consortium. A Board paper prepared for the meeting explained the rationale for the acquisition thus:

. AGL is part of a strong consortium that consists of one of Australia’s strongest financial institutions and one of the world’s largest generating utilities which will bring financial substance and operational stability to a key asset which is the major electricity supplier to AGL’s retail market position in Victoria;

. an investment is necessary to protect AGL’s ongoing position around the DPHA;

. Loy Yang is the lowest cost producer in the entire NEM and this transaction will provide a further opportunity for AGL to secure attractive off take arrangements into the long term;

. AGL will be well placed to take advantage of future brownfields expansions at the Loy Yang power station site (either through gas or modern brown coal technologies) and thereby secure access to the most competitive supplies of electricity to support its retail electricity business well into the future;

. AGL will also be able to influence the timing of such expansions more readily than if it were not a part owner of the plant and this is envisaged to be a significant competitive advantage in years to come;

. through part ownership of Loy Yang, AGL would be well placed to introduce Agility Management Pty Ltd (Agility) to offer services to the Loy Yang power station to permit Agility to gain access to Loy Yang’s non-labour operating cost pool and power station capital expenditure pool and thereby permit Agility to widen its skills into servicing the power generation sector;

. AGL will take a leadership role in working with the consortium owners to proactively manage future greenhouse emissions from the Loy Yang power station and take advantage of any carbon emission trading benefits that result from such initiatives; and

. AGL would secure a high yield equity return through an investment in tangible assets which will provide a long term and growth in earnings for AGL.

The paper recommended that AGL in a consortium with CBA and TEPCo lodge a non-binding indicative bid for Loy Yang Power aimed at securing the investment parameters and target returns set out in it. The Board endorsed that recommendation.

221               A further meeting of the Board was held on 3 April 2003. A paper was presented setting out the status of the bid. The strategic rationale for the acquisition was revisited. Reference was made to the future growth opportunities offered by investment in large-scale power generation. It was noted that with TEPCo as a partner, AGL’s consortium had the experience to understand both the market issues and technical issues surrounding the asset which should assist in minimising the risk of entering into this new sector. It was also stated in the paper:

‘While AGL will be restricted from day to day trading, hedging and dispatch operations in Loy Yang, AGL will benefit from a natural earnings hedge from an investment in Loy Yang. Volatility in earnings from AGL’s retail activities should be partially offset through increased access to earnings from the generation sector.’

At this meeting the Board endorsed continuing investigation of the investment.

222               On 22 May 2003, the GEA consortium executed a non-binding Letter of Intent with the Loy Yang partners following completion of the second stage of a due diligence process which was commenced by the consortium in January 2003. The Letter of Intent set out the key commercial terms and conditions upon which the acquisition would proceed and its key conditions.

223               At a meeting of the Board held on 2 July 2003 a further paper was submitted with a recommendation that the Board authorise the company to acquire a minority ownership stake of up to 35% in the new Loy Yang partnership and the business carried on by that partnership. There were recommendations for Board resolutions authorising the company to enter into a shareholders agreement, called the GEAC Shareholders Agreement and a subscription deed, called the GEAC Subscription Deed and all other agreements and arrangements necessary to give effect to the transaction. In the Strategic Rationale section of the Board paper reference was again made to the benefit accruing to AGL from a natural earnings hedge resulting from an investment in Loy Yang. The Board accepted the recommendations and the relevant transaction documents were executed on 3 July 2003.

Overview of Transaction Documents

224               The Loy Yang partnership which owns the Loy Yang A power station and the associated coal mine, comprises Horizon Energy Holdings Ltd, CMS Generation Horizon Energy Holdings Ltd, Horizon Energy Investments (No 2) Pty Ltd and NRGenerating Holdings (No 4) BV. The vehicles for the proposed acquisition are Great Energy Alliance Corporation Ltd (GEAC) and its subsidiary, GEAC Operations Pty Ltd (GEAC OpCo), which were incorporated by AGL together with Tokyo Electric Power Company International BV (TEPCo), the Commonwealth Bank of Australia (CBA), the Motor Trades Association of Australia Superannuation Fund Pty Ltd (MTAA) and Westscheme Pty Ltd (Westscheme).

225               On 3 July 2003, GEAC OpCo agreed with the members of the Loy Yang Partnership to acquire all of the shares in those companies. That agreement is called the Loy Yang Share Sale Agreement. The Share Sale Agreement which has been amended twice, on 2 September 2003 and 15 September 2003, has not yet been completed and can be terminated by any party if completion does not occur by 19 December 2003 because conditions precedent are not satisfied or waived. When the Share Sale Agreement is completed GEAC OpCo is to acquire from the Loy Yang Partners 100% of the shares in each of them. Post-acquisition the structure of the Loy Yang Business is to be governed by the Shareholders Subscription Deed.

226               By another agreement, also made on 3 July and called the GEAC Subscription Deed, each of the GEAC members agreed that prior to completion of the Share Sale Agreement they will acquire additional shares in GEAC so that their shareholdings will be at the following levels:

(a) AGL – 35%

(b) TEPCo – 35%

(c) CBA – 14.48%

(d) MTAA – 11.21%

(e) Westscheme – 4.31%

227               Under a further agreement, called the GEAC Shareholders Agreement, also made on 3 July 2003, the shareholders of GEAC, other than AGL, are required before the completion of the Share Sale Agreement to establish two companies being:

1. Loy Yang Marketing Holdings Pty Ltd (MM HoldCo)

2. Loy Yang Marketing Management Company Pty Ltd (MMCo) which is to be a wholly owned subsidiary of MM HoldCo.

The shareholding of MM HoldCo, again according to the GEAC Shareholders Agreement and the MMHold Co Shareholders Deed is to be held in the following proportions:

1. TEPCo – 49.97%

2. CBA – 26.32%

3. MTAA – 17.11%

4. Westscheme – 6.60%

Under the Shareholders Agreement (cl 5.1) AGL is not to hold any Economic Interest (a defined term) in MM HoldCo or MMCo. MMCo is to be the exclusive agent of the Loy Yang partnership to carry out dispatch and marketing activities currently carried out by the partnership. The agreement effectively imports the terms of an undertaking which AGL has offered to the ACCC under which it could not be involved in the dispatch and marketing activities of MMCo, nor have access to confidential generator or customer information. The terms of that undertaking and the way it is bought into the agreement are referred to in more detail below.

228               There is, in addition to the abovementioned agreements, a proposed MMCo Agency Deed under which the Loy Yang partners, post-acquisition, appoint MMCo as exclusive agent of the partnership to undertake dispatch and marketing activities and to take responsibility for and manage the existing hedge book. There is also an MM HoldCo Shareholders Deed. The parties to this Deed are the proposed shareholders of MM HoldCo, being TEPCo, CBA, MTAA, Westscheme, MM HoldCo and MMCo.

229               AGL contends that by virtue of the preceding structure and associated arrangements for the operation of the Loy Yang power plant it will not have any control over the way in which electricity from the plant is made available, bid, dispatched or contracted on a day- to-day basis. Nor, it submits, will it have any control over the entering into, administering or enforcing of electricity derivative contracts on behalf of the Loy Yang partners. Procedures for amending the structure require the agreement of 75% of non-AGL shareholders. AGL argues that those shareholders regard the present arrangements as according with their commercial interests and that there is no reason to think that a change in those arrangements is likely to be made. Any change in shareholding arrangements, if in contravention of s 50 of the Trade Practices Act, would fall for determination at the time of that change.

Details of Transaction Documents – The Share Sale Agreement

230               The parties to the Share Sale Agreement are the Loy Yang partners as vendors, GEAC OpCo as purchaser, GEAC as purchaser guarantor and CMS Generation Corporation as GMS guarantor.

231               It is not necessary for present purposes to set out the rather complex details of that agreement beyond referring to the conditions precedent in cl 2. In particular, cl 2.1(k) defines a condition precedent in the following terms:

‘(k) Both the Purchaser (acting reasonably) and the Vendors (acting reasonably form the view, and advise the other parties in writing of having the view, that:

(i) the ACCC has no reasonable prospects to, or will not be able to, restrain the acquisition of the Sale Interests by the Purchaser or the acquisition by AGL of securities in GEAC pursuant to the GEAC Subscription Deed; and

(ii) the ACCC has no reasonable prospects to, or will not following Completion be able to, set aside the acquisition of the Sale Interests by the Purchaser or the acquisition by AGL of securities in GEAC (including by obtaining any divestiture order against the Purchaser or AGL in relation to the foregoing).’

Details of Transaction Documents – The GEAC Shareholders Agreement

232               The parties to the Shareholders Agreement are GEAC, AGL, TEPCo, CBA, MTAA and Westscheme. The first operative clause provides for appointments to the board of directors of GEAC. Clause 2.2 defines the entitlement of each shareholder to appoint a number of directors according to the proportion of shares held. The effect of that table is that AGL and TEPCo, with 35% interests respectively, will each be entitled to appoint three directors. CBA with a 14.48% interest will be entitled to one director, as will MTAA with an 11.21% interest. Westscheme, with 4.31% is not entitled to appoint a director. Out of a board of eight directors therefore, AGL and TEPCo will each have a 37.5% representation and 12.5% representation will be held by each of CBA and MTAA.

233               Small shareholders in GEAC can combine their interest to appoint a number of directors by reference to the combined total of their shareholding proportions. So CBA and MTAA, together with Westscheme, have 30% of the shares in GEAC and so could collectively appoint three directors (cl 2.3). On that basis the size of the board would be expanded to nine members. It may be noted that a director can be an employee of the appointing shareholder (cl 2.2(e)).

234               Prior to the commencement of these proceedings, AGL offered an undertaking to the ACCC in order to facilitate the acquisition. The undertaking was intended to address the ACCC’s concerns that the proposed acquisition would involve a contravention by AGL of s 50 of the Trade Practices Act. The substance of that undertaking is that AGL will limit its interest in the Loy Yang Business to 35%. It would also undertake to take no economic interest in MMCo and not to participate in the appointment or supervision of the executive management of that company. AGL would also undertake not to be involved in the dispatch and marketing activities of the Loy Yang Business or any board or management decision-making in respect of such activities. The terms of the undertaking are set out as Annexure 8 to these reasons for judgment.

235               The Shareholders Agreement defines the ACCC undertaking in cl 1.1 as an undertaking substantially in the form attached as an annexure to that agreement which it has said AGL has offered to provide to:

(a) the ACCC; or

(b) if required, a court or other regulatory body

and which, it is said, may in future become binding on AGL.

236               Section 4 of the Agreement deals with the governance and decision-making of GEAC. It provides that AGL must use its best endeavours to comply with the ACCC undertaking and any subsequent or varied ACCC undertaking (cl 4.3(b)). The shareholders are reasonably to co-operate with AGL to set up policies, processes and protocols between MMCo and ServCo in order to assist AGL in complying with the undertaking (cl 4.3(a)). None of the shareholders is to do or omit to do anything which would cause AGL to be in breach of the undertaking (cl 4.3(c)). In par 4.3(d) it is provided:

‘Directors appointed by AGL must not participate in, or be present at any part of a Board meeting at which any matter is discussed which would cause AGL to breach cl 3.3(b) of the ACCC Undertaking.’

Clause 3.3(b) of the Undertaking would prevent AGL from being involved in any board or management decision-making in respect of dispatch and marketing activities.

237               Directors appointed by AGL are entitled to be present at and to take part in any board discussions which relate to the protocols, procedures and framework of such matters but not any discussions which relate to the operational or transactional elements of such matters (cl 4.3(d)).

238               The board is authorised under the agreement to take such action as it reasonably considers to be necessary or appropriate ‘… to ensure, review or monitor the adequacy of controls, staffing and processes as regards energy trading and compliance with the Board-approved hedging policy’ (cl 4.3(e)). MMCo is to be solely responsible for dispatch and marketing activities. Clause 4.3(g) provides:

‘The parties will conduct their affairs in respect of the Group and the Business as if the ACCC Undertaking was binding on AGL, and they acknowledge that the ACCC Undertaking may, in future, become binding on AGL. However, if at any stage the Board reasonably considers that:

(1) the ACCC Undertaking is not at that time, and is not in the future likely to become, legally binding on AGL; and

(2) the management and governance of the Company, the Group and the Business should no longer necessarily be conducted in a manner consistent with the ACCC Undertaking,

then the Board may resolve, by Board Special Majority that:

(3) the operations of the Business ( in whole or in part) should no longer be conducted in a manner consistent with the ACCC Undertaking as contemplated by this agreement; and

(4) the terms of this agreement, including but not limited to clauses 1.1, 3.6, 4.3, 5.1, 5.4, 5.5, 6.9, 7.3, 8.6(e), 8.11 and 11.13, and any other relevant agreements be amended to reflect that resolution.

As soon as practical after the passing of a resolution by the Board as contemplated by clause 4.3(g)(4), the Shareholders must procure that this agreement and any other relevant agreements are amended to the extent and in the manner contemplated by that Board resolution. For the avoidance of doubt, any Directors appointed by AGL or its Affiliates will not be precluded from voting on any Board resolution contemplated by this clause 4.3(g).’

239               The reference to a Board Special Majority is explained by the definition of that term in cl 1.1:

‘Board Special Majority means a majority of at least 75% of the votes cast by directors entitled to vote on a matter;’

240               Further support for the arrangements contemplated by cl 4.3 is derived from cl 3.6. That clause in par (a) prohibits a director appointed by AGL from voting on or being present at any part of a board meeting where to do so would cause AGL to breach the ACCC Undertaking.

241               Clause 5.1 provides for the shareholders, other than AGL, to establish MM HoldCo and MMCo by the time of completion of the agreement on a basis consistent with cl 3.2 of the ACCC Undertaking. The parties acknowledge that AGL and its affiliates will not have any Economic Interest in MM HoldCo or MMCo. Nor will they enter any contracts, arrangements or understandings which would give or have the effect of giving AGL such an interest (cl 5.1(b)).

242               The term ‘Economic Interest’ has the meaning given to it in the ACCC undertaking. That is defined thus:

‘Economic Interest:

(a) means interests in a company or partnership, including, shares, voting rights, rights to receive dividends, rights to receive other distributions of income or capital, rights to receive a share of proceeds on winding up; but

(b) excludes:

(i) rights to purchase the interest in the Loy Yang Business or the Loy Yang Assets of a Consortium Member seeking to divest its interest the exercise of which are subject to AGL obtaining approval (on a formal or informal basis) from the Commission or the Australian Competition Tribunal; and

(ii) any rights AGL has to prevent approval of decisions in respect of Permitted Matters.’

243               The term ‘ServCo’ used in the agreement appears to refer to the designation given to LYPM by the existing partners. Post acquisition MMCo will undertake dispatch and marketing activities as agents of the partnership, whilst LYPM will continue to conduct the business of the partnership, save for these activities.

244               Under cl 5.4, GEAC must not give AGL Confidential Customer Information or Confidential Generator Information nor permit AGL to obtain access to such information. Moreover the shareholders in MM HoldCo must not provide such information to AGL or permit AGL to have access to such information and must do all things reasonably necessary within their power to procure that MM HoldCo and MMCo do not provide that information or access to it (cl 5.4(b)). The terms ‘Confidential Customer Information’ and ‘Confidential Generator Information’ are defined in the Shareholder Agreement by reference to their definition in the ACCC undertaking. It is convenient to set out those definitions here:

‘‘Confidential Customer Information’ means the details (including the identity of each counter-party, price, term and volume) of any specific Customer Contracts with the Marketing Management Company, but excluding:

(a) information which is generally known; and

(b) AGL’s own Customer Contracts with the Marketing Management Company.

‘Confidential Generator Information’ means:

(a) details (including identity of each counter-party, price, term and volume) of Dispatch and Marketing Activities; and

(b) details (including quantities, dates and times) of any reductions or expected reductions in the availability of the Loy Yang Plant to less than the Plant’s Registered Capacity

but excluding information which is generally known.’

245               In s 6 of the Agreement, dealing with management, there is a requirement in cl 6.3 for the preparation of budgets by MMCo. Thus:

‘6.3(a)The Board must procure that management of the company and/or MMCo prepare and update (annually) a five year look ahead draft Business Plan and Budget and draft annual Budgets, for presentation to the Board in February of each year.’

246               Clause 6.5 requires GEAC to provide to the shareholders, inter alia, within 15 days after the last day of each calendar month, unaudited management accounts of the Group for that month comprising:

1. Commentary on the operational and financial position for that month, including key performance indicators, and variances from the Budget;

2. A profit and loss account and cash flow statement for that month;

3. A balance sheet as at the end of that month;

4. A forecast for the performance of the Group in the month immediately following that month; and

5. A review of actual results against the Budget and an explanation of any material variances.

247               There is a requirement that as soon as practicable after completion the Board appoint a Risk Management Sub-committee to formulate and consider the Risk Management Policy (cl 3.11). The Risk Management Policy has the meaning given in Schedule 3 to the Agreement. It is in substantially the same terms as the Risk Management Policy appended to the ACCC undertaking. The character and general content of the Risk Management Policy is set out in the Schedule to the Shareholders Agreement. It is required to set global limits and controls on the business’s exposure to specific categories of business risk. It must also establish internal governance, an internal governance structure and management philosophy for managing such risk. The Policy is required to define specific risk areas being:

1. Major asset (physical) risk, (for example risks of losses arising from poor maintenance on key productive assets (turbines));

2. Trading risks (price, volume and credit risk) (for example, risk of losses from or total contract portfolio exceeding available plant capacity or a disproportionate exposure to spot prices;

3. Operational risk (for example, breakdown in human resources, processes or technology); and

4. Legal and compliance risk (for example, poor contract management systems, absence of compliance with the Trade Practices Act, Corporations Act or National Electricity Code).

248               In par (d) it is said that the global limits and controls may be quantitative or qualitative and would operate to limit or control the total level of risk that may be incurred by a specific business activity or division. By way of example it is said:

(1) Trading risk (price and volume) may be subject to limits on the proportion of financial budget forecasts which may be put at risk as a consequence of contract portfolio exceeding available plant capacity or a disproportionate exposure to spot prices;

(2) Credit risk (counter-party default risk) would be subject to limits on the financial exposure to counter-parties of varying categories of credit worthiness; and

(3) Legal contract risk would be subject to requirements on use of ISDA pro forma contracts and ISDA optional force majeure clauses in those contracts.

249               The policy also provides, in par (f), that:

‘In relation to trading, credit and legal risk the Risk Management Policy would provide for the delegation to a Risk Management Committee of MMCo responsibility:

(1) for overseeing risk management operations and procedures (within the parameters of the Risk Management Policy) of activities involving energy market exposure; and

(2) for setting individual trader limits.’

Paragraph (g) distinguishes between a Risk Management Policy and a Risk Management Framework.

250               The elements of the Risk Management Policy just referred to have been pointed to by the ACCC in these proceedings as indicating an avenue by which AGL, through its involvement of the setting of the Risk Management Policy through the GEAC Board, could influence the trading activities of MMCo.

251               In the section relating to governance and decision-making it is provided, inter alia, that Board Special Matters must not be made unless approved by affirmative vote of the Board’s Special Majority (cl 4.2(b)). A Board Special Majority, it will be recalled, is defined in cl 1.1 to mean a majority of at least 75% of the votes cast by directors entitled to vote on a matter. The definition of ‘Board Special Matter’ includes implementing or making any changes to the Risk Management Policy (l). It also covers adopting a budget, approving the financial statements or any material deviation from a budget (n). The significance of this definition is that, when read with the definition of Board Special Majority, it confers a power of veto on AGL in respect of decisions taken on any of the defined Board Special Matters.

252               The terms of the undertaking are set out at the end of these reasons and it is not necessary to repeat them in detail here. It should be noted that the undertaking in cl 3.4 defines ‘Permitted Matters’ relating to the conduct of the Loy Yang Business which, under the agreements, require the agreement and/or participation of AGL. They include relevantly:

‘(a) the financing and capital structure of the Loy Yang Business;

…

(c) expansion of the Loy Yang Business and acquisitions, construction and disposals of substantial assets;

…

(g) annual budgets (including capital and operating expenditure) and approvals of expenditures and liabilities outside of budget of the Loy Yang Business;

(h) the Risk Management Policy of the Loy Yang Business.’

Details of the Transaction Documents – The MMCo Agency Agreement

253               Attached to the GEAC Shareholders Agreement as Attachment 2 is a proposed term sheet for an agency agreement between MMCo and the Partnership. A Deed called the MMCo Agency Deed has been prepared between the Loy Yang Partners and MMCo and MM HoldCo. Loy Yang Power Management Pty Ltd is also a party to the agreement although, according to recital G that is solely to enable it to transfer to MMCo responsibility for, and management of, the Existing Hedge Book of the Partnership from Completion.

254               Under cl 2 of the Deed the Loy Yang Partners appoint MMCo as the exclusive agent of the partnership to:

1. Undertake the dispatch and marketing activities; and

2. To manage the Existing Hedge Book of the Partnership

in accordance with the terms of the deed. The Existing Hedge Book is defined in cl 1.1 of the deed as:

‘… the Electricity Derivative Contracts the Partnership entered into prior to the commencement of this Deed and in respect of which the Partnership has (or may have) any unperformed obligations as at Completion.’ (CB 579)

255               MMCo’s powers are set out in cl 3.1 of the Deed. They include:

(a) make dispatch offers, dispatch bids and market ancillary service offers (as those terms are defined in the Code);

(b) negotiate, enter into, administer and enforce Electricity Derivative contracts (including the Existing Hedge Contracts), or any other contracts required to undertake the dispatch and marketing activities within the scope of the Risk Management Policy. MMCo may enter into contracts for and on behalf of the partnership as its disclosed agent (or pursuant to powers of attorney from the Partners in favour of MMCo;

In cl 3.2(a) it is provided:

‘MMCo must not do any act, matter, or thing that:

(1) is outside the scope of the Risk Management Policy (as amended from time to time);

(2) is outside the scope of the Operating Budget; or

(3) is not consistent with the ACCC undertaking,

and each of MMCo and MM HoldCo must ensure that any contract or arrangement entered into under this Deed (other than any contract or arrangement with GEAC, GEAC Op Co, the Partners or LYPM) is on arms length terms.’

256               There is also an obligation on MMCo, no later than 1 February of each year, to submit to the partnership for its approval an Operating Budget for the dispatch and marketing activities of the partnership (including in relation to the Existing Hedge Book). The budget must also cover other costs and expenses MMCo will incur in carrying out its obligations under the deed and another deed entitled the MMCo Operational Deed with respect to the next following financial year.

257               The Operating Budget is required to describe operating expenses to be incurred by MMCo and revenue expected to be earned by the Loy Yang business in connection with MMCo’s dispatch and marketing activities. There is a proviso that no Confidential Generator Information and Confidential Customer Information can be included in the budget. It is provided that for the avoidance of doubt aggregate information can be provided in the Operating Budget (cl 6.1). The deed annexes the ACCC undertaking and also the Risk Management Policy which is an appendix to that undertaking.

Details of the Transaction Documents – The MMCo Operational Deed

258               The MMCo Operational Deed divides the role and responsibilities of Loy Yang Power Management and MMCo. Under cl 2.1 Loy Yang Power Management is to conduct the business of the partnership, other than dispatch and marketing activities, as agent of the partners subject to the terms of the deed and pursuant to the Deed for Appointment of Partnership Representative. The business includes the generation of electricity from the power station, monitoring and reporting on the need for maintenance or capital improvements, undertaking maintenance of the Facilities, operating and maintaining the mine and associated infrastructure and regulatory compliance with occupational health and safety and environmental laws. MMCo on the other hand is required to ‘… carry out the dispatch and marketing activities as agent of the partnership subject to the terms of this Deed and the MMCo Agency Deed.’

Details of the Transaction Documents – The MM HoldCo Shareholders Deed

259               A further relevant transaction document is the proposed MM HoldCo Shareholders Deed. The parties are the Loy Yang Partners (other than AGL) together with MMCo and MM Hold Co. The deed recites that MM HoldCo has been established to be the 100% owner of shares in MMCo and that MMCo has been established to undertake the dispatch and marketing activities as agent for the partners. The shareholders in MM HoldCo recite that they wish to enter into the deed to set out their ‘… arrangements in respect of the ownership, management and business of the company and MMCo.’ Under cl 4.2 of the deed each shareholder is entitled to appoint to the board a number of directors according to the portion of the total shareholding which it owns. The minimum shareholding for appointment of one director is 15%.

260               Under cl 7.1 of the Deed the parties acknowledge that AGL and its affiliates will not have any Economic Interest in MM HoldCo or MMCo. Moreover the parties agree that they will not enter any contracts, arrangements or understandings which would give or would have the effect of giving AGL an economic interest in the company or MMCo. The term ‘Economic Interest’ is defined by reference to its meaning in the ACCC undertaking which is exhibited to these reasons for judgment.

261               The parties agree that if the ACCC undertaking were to cease to prevent AGL from holding shares they will review the structure for the operation of the company, its subsidiaries and its business.

AGL Seeks Informal Clearance From the ACCC

262               AGL lodged a submission with the ACCC, on or about 20 March 2003, seeking an informal clearance in respect of the proposed acquisition. The AGL submission indicated that it proposed to acquire a 30% economic interest in Loy Yang Power as part of a consortium. The submission referred to AGL’s limited generation assets, Somerton in Victoria and Hallett in South Australia.

263               AGL understood that the ACCC would need to be satisfied that no substantial lessening of competition would arise as a consequence of an acquisition of an interest in Loy Yang. It submitted that an acquisition would not have such an effect and that in any event its interest in Loy Yang would be managed at arms length from its other business interests. Its submission described the proposed structural arrangements under which it would have no control or substantial influence over the way in which the Loy Yang capacity was made available, bid, dispatched and contracted on a day-to-day basis. It argued that the arrangements would also prevent it from being privy to commercially sensitive information about retail competitors and their dealings with Loy Yang or about Loy Yang’s bidding dispatch and contracting strategies. AGL offered to give appropriate undertakings to the ACCC under s 87B of the Trade Practices Act in order that the ACCC could have confidence as to the certainty of the proposed structural arrangements. It maintained that, with the structural undertakings proposed, its acquisition would not raise competitive concerns because:

. the future availability of Loy Yang derivatives and the competitiveness of the terms upon which they might be offered would be unaffected by AGL’s proposed 30% interest in Loy Yang

. the conditions for competition in the retail sector would be unaffected by the acquisition

. there would be no economic aggregation of the competitive positions of Loy Yang and generation control by AGL.

On or about 5 May 2003, AGL provided a draft undertaking to the ACCC.

264               On 12 June 2003, the ACCC wrote to the solicitors for AGL stating its view that the proposed acquisition raised substantial competition concerns pursuant to s 50 of the Act. It rejected the proposed undertaking. The letter from the ACCC said:

‘In reaching this view, the Commission conducted a thorough investigation so as to determine the likely effect of the transaction on both the generation and retail sectors of the electricity supply chain. This process involved contacting a wide range of market participants.

As you are aware, the proposed acquisition raises the prospect of re-aggregation of AGL’s substantial electricity retail and distribution businesses in Victoria with the largest generator in that State.

Such integration raises the following two main vertical competition concerns:

. Increased barriers to entry/expansion for other electricity retailers to compete in Victoria, through reduced scope for such retailers to secure competitively priced contracts required to hedge Victorian loads; and

. Increased barriers to entry for embedded generation projects within the AGL distribution network.’

265               The ACCC elaborated upon its concerns by observing that in mid 2002 AGL had acquired Pulse Energy and Origin Energy had acquired Citipower. The number of incumbent electricity retailers in Victoria had therefore decreased from 5 to 3. The ACCC had not opposed those acquisitions because it was likely that non-incumbent retailers would continue to compete to supply commercial and industrial customers and would leverage off their activities in that sector together with their positions in New South Wales and Queensland to increasingly compete for the ‘mass market’. The ACCC’s decision in respect of those acquisitions was based on the structure of the electricity supply industry in Victoria being one whereby such parties could secure the hedge contracts requisite to compete. The ACCC was also of the view that hedge contracts with generators in other States were substitutable only at the margin for those with in-State generators. It was therefore concerned that cross-ownership between Loy Yang Power and a large Incumbent Retailer such as AGL was likely to have the effect of substantially reducing the ability of other retailers to secure competitively priced contracts required to hedge load in Victoria.

266               By the time of the June letter the quantum of AGL’s proposed interest in Loy Yang had risen to 35% and the ACCC considered the acquisition on that basis. It was not satisfied that the proposed undertaking would sufficiently address the competitive concerns raised by the acquisition because:

1. The competition concerns were structural.

2. The ACCC was reluctant to accept undertakings of a behavioural character which would raise issues in terms of ongoing monitoring and enforcement.

3. The proposed undertaking did not provide the ACCC with sufficient certainty that all dealings between Loy Yang Power and AGL would be on an arms length basis and that confidential and market sensitive information would be fully ring-fenced from AGL.

There was a reference then to ‘embedded generators’ and their important potential in the NEM as a substitute for remote generation and network augmentation. Embedded generators are generators with capacity of less than 160MW located within distribution networks close to customers. The ACCC was concerned that AGL would have both the incentive and ability to frustrate the development of embedded generation projects that would otherwise diminish the value of its interest in Loy Yang Power.

267               Further discussion with the ACCC ensued and there was a meeting between it and representatives of AGL on 1 September 2003. The Chairman of the ACCC, Mr Samuel, then wrote a letter dated 5 September 2003, which is referred to in the pleadings. In that letter he said:

‘The Commission has now considered the section 87B Undertaking proffered by AGL and the GEA consortium on 2 September 2003, and has reached the view that the Undertaking does not adequately address the competitive concerns raised by this acquisition. In particular, and as I believe was discussed at length at the 1 September 2003 meeting, the Commission has significant reservations in relation to the following aspects of the Undertaking:

. the inadequacy of the proposed remedy to overcome the inherent shortcomings of the “reasonable endeavours” clause requiring that Loy Yang Power achieve a yearly net average contract position of 75% of its availability capacity

. the level of the proposed cap on contract between AGL and Loy Yang Power; and

. the lack of acceptable duration and sunset clauses

In view of these reservations the Commission declines to accept the Undertaking.

Further the Commission has asked me to inform you that should the GEA consortium decide to proceed and complete the acquisition the Commission will, for the present, reserve its position as to what course of action it will take. This may include, but is not limited to, taking action for contravention of section 45 and/or section 50 of the Act and seeking the full range of remedies available for contravention of these sections, including pecuniary penalties and divestiture.’

268               On 8 September 2003, the ACCC published a press release announcing that, to protect Victorian electricity consumers and customers, it would maintain its opposition to the proposed acquisition. The press release quoted statements from Mr Samuel who said, inter alia:

‘The ACCC remains firmly of the view that the proposed acquisition creates substantial competition concerns which are potentially in breach of section 50 of the Trade Practices Act 1974.

It would lead to a less competitive and less efficient market structure in Victoria and, potentially, in the National Electricity Market.

This is likely to result in higher prices, increased barriers to entry and a resulting substantial lessening of competition.

Therefore, the ACCC will oppose AGL acquiring an interest in Loy Yang Power.’

269               He also said:

‘The ACCC remains strongly opposed to this transaction and will continue to build its enforcement case should AGL, the Commonwealth Bank and TEPCO decide to close the transaction without providing undertakings satisfactory to the ACCC. The ACCC would seek appropriate remedies from the Federal Court, including divestment.’

 

The Commencement of the Proceedings

270               On 15 September 2003, one week after the publication of the ACCC press release, AGL filed an application in this Court commencing these proceedings. Programming directions were made. The trial of the action was listed for three weeks commencing 18 November 2003 on the basis that judgment would be delivered on or before 19 December 2003 when a major payment was due in respect of the proposed acquisitions. The outcome of the case may affect the fulfilment of a condition precedent to the proposed acquisitions and that payment.

271               AGL filed its defence on 6 October 2003. In that defence it raised an objection to jurisdiction. On 8 October 2003 I ordered that the question of jurisdiction be determined as a separate question, on the papers, by reference to the pleadings and written submissions. Directions were given for the filing of written submissions by the ACCC and by AGL. On 31 October 2003, I found, on the preliminary question, that the Court does have jurisdiction to entertain the proceedings – Australian Gas Light Company v Australian Competition and Consumer Commission (No 2) [2003] FCA 1229. The pre-trial preparation involved a number of case management conferences. The hearing of the action itself occupied twelve and a half days, for the most part sitting extended hours to ensure that all the evidence was received within the allocated hearing time. It concluded on 5 December 2003. The presentation of their cases by counsel and solicitors for both parties was conducted with notable efficiency.

The Pleadings – Ownership and Acquisition

272               The ownership of the Loy Yang A power station and the identities of its principal acquirers pleaded in pars 1 to 6 and 8 to 10 is not in dispute. Nor is it in dispute, as pleaded in par 10 that AGL, TEPCo and CBA hold two shares in the capital of GEAC and that MTAA and Westscheme do not. It is not admitted that MTAA and Westscheme will acquire shares in GEAC prior to completion and that AGL, TEPCo and CBA will increase their shareholdings to the specified levels at about that time. There is a position put at a number of points in the ACCC defence in answer to pleadings about the agreements dealing with the post-acquisition structures and arrangements between AGL and other consortium members and the provisions designed to keep AGL at arms length from the operations of Loy Yang A Power. This position is exemplified in par 11 of the ACCC defence. It is admitted in par 11(a) that the GEAC members entered into the Shareholding Agreement on 3 July 2003. However in par 11(b) the ACCC pleads:

‘The ACCC:

says further that, in the absence of binding undertakings given to the Court by each of the parties to the GEAC Shareholders Agreement to give effect to the GEAC Shareholders Agreement consistently with its terms (as is the fact), it does not otherwise admit the allegations contained in paragraph 11 of the Statement of Claim.’

The pleading in par 11 of the statement of claim was simply that on 3 July 2003 the GEAC members entered into a Shareholders Agreement which governs the ownership, management and business of GEAC. The defence set out in par 11(b) appears designed to support the contention that, absent legally binding undertakings, there is no certainty about the post-acquisition structures or arrangements which will enable the Court to conclude that the acquisition is not likely to substantially lessen competition in a relevant market. This defence may be designated the ‘future uncertainty defence’. It is raised in respect of the GEAC Shareholders Agreement and the Subscription Deed. The Loy Yang Share Sale Agreement itself is admitted without that qualification.

273               The statement of claim sets out the operation of the Loy Yang Partnership following the GEAC acquisition. It refers to the creation of MM HoldCo and MMCo and the appointment of MMCo to carry out Dispatch and Marketing Activities (as defined). It refers to the shareholding in MM HoldCo which excludes AGL and the continuing role of LYPM in operating the power station. Alteration of the MMCo and LYPM roles would require, it is said, a majority of at least 90% of the votes cast by GEAC shareholders entitled to vote on the matter (pars 17 and 20). The provisions preventing AGL from having access to confidential generator information and confidential customer information, the requirement that all transactions between GEAC and its shareholders, including AGL, be at arms length and the 90% voting threshold to alter those arrangements is also pleaded (sc 22-25). At par 26 it is said:

‘By reason of the matters set out at paragraphs 13, 17, 20, 23 and 25 above, AGL can not on its own alter the arrangements set out at paragraphs 16, 19, 22 and 24 above.’

The ACCC raises the future uncertainty defence in respect of par 26 thus:

‘16. In the absence of binding undertakings given to the Court by each of the parties to the GEAC Shareholders Agreement to give effect to the GEAC Shareholders Agreement consistently with its terms (as is the fact), the ACCC does not admit any of the allegations contained in any of paragraphs 16 to 26 inclusive of the Statement of Claim.’

274               The statement of claim then turns to the issue of ACCC opposition to the proposed acquisitions. AGL pleads that the ACCC has contended that the proposed acquisitions are likely to substantially lessen competition in contravention of s 50 of the Trade Practices Act (SC 27). The ACCC does not admit that allegation and says that its letter, dated 5 September 2003, to AGL’s solicitors stated its position in relation to the Proposed Acquisitions which position has not changed (D 17). Then in par 28, AGL says that if it and the other GEAC members proceed to complete the proposed acquisitions the ACCC may commence proceedings against it and the others seeking remedies arising from an alleged contravention of s 50, including possible divestiture of AGL’s shareholding in GEAC or GEAC OpCo’s shareholdings in the Loy Yang Partners. The ACCC does not admit that allegation and repeats that its position is as set out in its letter of 5 September 2003.

275               AGL then pleads:

‘29. By reason of the matters set out at paragraphs 30 to 116 below the Proposed Acquisitions do not contravene s 50 of the Act.’

In answer to this, the ACCC says in its defence at par 19:

‘19. The ACCC denies the allegations contained in paragraph 29 of the Statement of Claim.’

The Pleadings – The Electricity Industry in Victoria

276               The statement of claim pleads extensive facts about the electricity industry in Victoria. The first section of that pleading is headed ‘Retail Supply of Electricity in Victoria’ (SC 31-37). It is all admitted save for par 37 dealing with AGL’s share of the total number of Victorian retail customers and the total volume of electricity sales to those customers. In substance this section of the statement of claim alleges that, within Victoria, there is a material demand for electricity by residential, small business, commercial and industrial customers referred to as ‘Victorian Retail Customers’. They can be divided into two broad categories namely, residential and small business customers who typically consume up to 160MWh of electricity per annum and industrial and commercial customers who typically consume in excess of 160MWh of electricity per annum. The level of demand at any point of time for electricity from Victorian Retail Customers averaged over the twelve month period from 1 July 2002 to 30 June 2003 was approximately 5,008MW. Long run growth in average Victorian retail demand is approximately 1.8% per annum and in the entire NEM is approximately 2.4% per annum. The peak level of demand for electricity from Victorian Retail Customers from 1 July 2002 to 30 June 2003 was approximately 8,202MW. This is called ‘Peak Victorian Retail Demand’. Long run growth in Peak Victorian Retail Demand is approximately 2.9% per annum and in the entire NEM is approximately 3.0% per annum.

277               There is a number of existing and potential suppliers of electricity to Victorian Retail Customers which are admitted. These are:

(a) AGL, which operates its electricity distribution network through a subsidiary, AGL Electricity Ltd, and which comprises 7,041 kilometres of electricity distribution in the north-west greater Melbourne region. AGL owns and operates two electricity generation plants namely:

A. Somerton, in Victoria, with a generating capacity of 160MW and

B. Hallett, in South Australia, with a generating capacity of 220MW.

(b) Origin Energy Ltd, which owns and operates four gas-fired electricity generation plants at:

A. Roma (Queensland) – 80MW

B. Ladbroke Grove (SA) – 80MW

C. Quarantine (SA) – 96MW

D. Mt Stuart (Qld) – 304MW (Origin’s ownership of this plant is not admitted by the ACC.)

Origin also owns a 50% interest in two further gas-fired generation plants at:

A. Osborne Plant (SA) – 180MW

B. Bulwer Island (Qld) – 32MW

(c) TXU Australia Pty Ltd, which operates an electricity distribution network in Victoria providing services to over 500,000 customers in the eastern suburbs of Melbourne and the eastern side of Victoria generally. Through a subsidiary, TXU (South Australia) Pty Ltd, it owns and operates the Torrens A and B steam turbine electricity generators in South Australia with a total generating capacity of 1280MW. TXU also has in place a 966MW agreement with EcoGen which owns and operates electricity generation plants in the La Trobe Valley and at Newport in Victoria which have a generating capacity of 932MW registered with NEMMCO by which it has the right to determine when and how much electricity EcoGen will generate and the price at which it will bid for the supply of that electricity. This agreement, which is a Master Hedge Agreement, was not admitted by the ACCC on the pleadings. At trial however a statement of agreed facts was tendered concerning this agreement known as the Master Hedge Agreement and its effect as pleaded was agreed. The statement of agreed facts was a confidential exhibit.

(d) Yallourn Energy Pty Ltd trading as AusPower, which supplies electricity to Victorian Retail Customers and owns and operates an electricity generation plant in the La Trobe Valley with a capacity of 1450MW. It also owns and operates a generator at Longford in Victoria with a capacity of 31.8MW.

(e) Energex Ltd, owned by the Queensland Government, which is registered with NEMMCO as an intermediary in respect of generating capacity of 8.65MW located in the Sunshine Energy Park (Victoria), owned by Beak Industries Pty Ltd.

278               Three New South Wales government owned electricity retailers and distributors supply electricity to Victorian Retail Customers (SC 36(f)-(h)). They are Energy Australia Pty Ltd, Country Energy and Integral Energy. Aurora Energy, a Tasmanian based company, is licensed to supply electricity to Victorian Retail Customer (SC 36(i)). Ergon Energy Pty Ltd, which is owned by the Queensland Government, also has a retail license, as does Australian Energy Services Pty Ltd trading as Powerdirect.

279              It is not admitted, in relation to the Victorian Retail Customers generally, that AGL’s share of the total number of those customers measured as a percentage of the number of registered meters that it services in Victoria to the total number of registered meters in Victoria is about 35%. Nor is it admitted that AGL supplies to Victorian Retail Customers about 25% of the total electricity consumed in Victoria.

The Pleading – The Relevant Markets

280               The statement of claim pleads the relevant markets in a somewhat indirect way. In par 38 it is said that the ACCC alleges that there are certain retail markets for the supply of electricity in Victoria. The ACCC partly admits and partly denies the allegation. It is sufficient for present purposes and common ground, as emerged from case management conferences prior to the trial, that this is to be treated as a substantive pleading of the retail markets and that it is common ground that there are separate retail markets:

1. for the supply of electricity to residential and small business customers in Victoria; and

2. for the supply of electricity to industrial and commercial customers in Victoria.

 

281               The wholesale markets are also pleaded in the same somewhat indirect way. The parties are at odds on their definition. AGL alleges that there is a single market in which all electricity generators participating in the NEM compete for the supply of electricity and electricity derivative contracts to Victorian electricity retailers and other wholesale purchasers of electricity and electricity derivative contracts in Victoria. This is called the Single Wholesale Electricity and Derivatives contract market. The ACCC on the other hand asserts separate wholesale markets comprising:

1. A market in Victoria for the generation and supply of electricity by wholesale to NEMMCO which market comprises electricity generators located in Victoria and electricity supplied through interconnectors into Victoria; and

2. A market or markets for the supply of electricity derivative contracts referenced to the Victorian Regional Reference Node.

The essential point of difference is that AGL contends for a single market for the supply of electricity and electricity derivative contracts, whereas the ACCC identifies separate product markets.

 

The Pleading – Competition in the Retail Markets

282               A number of claims are made about competition in the Victorian Electricity Retail Markets. Various competitive strategies are said to be adopted by Victorian Electricity Retailers to acquire and/or retain Victorian Retail Customers. The first of these is the efficient management of risks arising from the wholesale acquisition of electricity. Others are efficient retail operations, competitive retail pricing, the provision of related customer services such as maintenance and repair, and other services such as green energy, one point of contact for the supply of both electricity and gas, direct debit and bill smoothing, and the operation of retail stores selling electrical appliances. The promotion of awareness of retailers’ services, a reputation for reliability and the close monitoring of competitive activities of other retailers are also relied upon. All are admitted save for the assertion that the Victorian retailers offer electricity at competitive retail prices. That is not admitted. It is undisputed that Victorian Retail Customers, particularly residential and small business customers, can transfer from one Victorian Electricity Retailer to another. Although AGL says this is facilitated by uniform billing and customer account identification systems, that allegation is denied. It is admitted however that there are no material regulatory restrictions on Victorian Retail Customers transferring from one retailer to another.

283               AGL says that there are many entities with the necessary expertise, systems and financial resources to market and sell electricity to Victorian Retail Customers, including each of the Victorian Electricity Retailers. The ACCC while admitting this in respect of Victorian Electricity Retailers, denies that there are other entities which have that expertise and capacity. It is common ground that there are no significant regulatory impediments to preclude Victorian Electricity Retailers from increasing their shares of total sales of electricity to retail customers. However the ACCC denies AGL’s claims that there are no significant commercial impediments which have that effect.

284               AGL says that in the period from 1 January 2002 to 2 July 2003 it has had a significant number of its Victorian Retail Customers transfer to other Victorian Electricity Retailers and vice versa. The ACCC admits such transfers but does not admit their numbers are significant. AGL says that if retail prices charged by a Victorian Electricity Retailer were to exceed those of other retailers by any material margin, a significant number of customers would transfer. This is not admitted. The ACCC also says that the retail supply of electricity to customers who consume not more than 160MWh of electricity per annum is subject to price regulation under the Essential Services Commission Act 2001 (Vic) and the Electricity Industry Act 2000 (Vic).

285              The ACCC denies AGL’s contention that it is competitively constrained in the supply of electricity to Victorian Retail Customers.

The Pleadings – Generation and Transmission

286               The statement of claim deals with electricity transmission and generation. It is largely descriptive of the national gridgeneration and transmission systems and in many respects the description is not in dispute, although there are some areas of contention. The national grid is first generally described and existing and planned interconnection capacity between transmission systems in each State and the Snowy Mountains Hydro-Electric Scheme is set out. The connection of almost all electricity generators in the ACT, New South Wales, Victoria, Queensland and South Australia to the national grid is pleaded and admitted, as is the prospective connection to the grid of the Tasmanian transmission and distribution system when the construction of Basslink is completed. It is not admitted, however, that the physical characteristics of the transmission of electricity are such that electricity supplied into the national grid at one point to a consumer at another point will affect the volume of electricity available at all points in the grid. The ACCC says that there is a range of factors that affect the volume of electricity available at all points in the national grid.

287               AGL sets out a list of Victorian-based electricity generators which supply electricity into the national grid and specifies their generation capacities. It also sets out significant electricity generators in States other than Victoria which supply electricity into the national grid and lists their capacities. The capacity of each of these generators is admitted. However, according to the ACCC, certain of the generation capacity listed is not scheduled within the meaning of the NEC and so does not participate in NEMMCO’s co-ordinated central dispatch process. It also refers to the various factors which can affect the available capacity of a generator such as ambient temperature, fuel quality, outages, and the capacity of a generator at peak summer temperatures.

288               The ability of Hydro Tasmania, with a generation capacity of approximately 2,540MW, to supply electricity into the national grid from 2005 with the construction of Basslink is admitted. AGL sets out what it says are the respective shares of electricity generation available in Victoria. This is not admitted and the ACCC specifies, in Schedule A to its defence, what it alleges is a more accurate assessment of the respective shares of electricity generation capacity available in Victoria. Schedule A actually refers to the summer rating generation capacity of Victorian scheduled generators. It displays a total capacity in 2003/04 of 8,105MW whereas AGL’s pleading alleges a total generation capacity of 10,867.65MW. The ACCC takes its table from the NEMMCO 2003 Statement of Opportunities.

289               AGL contends that the total cost of establishing a new electricity generation plant in the NEM and operating that plant for its expected life is about $35/MWh to $45/MWh depending on the nature of the plant and its likely utilisation. This is not admitted. It is admitted, as pleaded by AGL, that there are substantial proposals for the expansion of existing electricity generation capacity and the construction of new generation capacity in Victoria and other States. AGL particularises this allegation by reference to the NEMMCO 2003 Statement of Opportunities. The ACCC refers to the description of proposals in that Statement and otherwise does not admit the plea.

The Pleading – Operation of the NEM – Inter-Regional Pricing

290               The statement of claim then deals with the operation of the NEM. Much of the description of the operation of the NEM in this section is not in dispute and there is no need to repeat its elements here. They have been described earlier in these reasons.

291               AGL describes the activities undertaken by NEMMCO in operating the NEM. Those activities are given slightly different descriptions in the defence but their general outline is the same, covering as they do the receipt, adjustment, ranking and control of dispatch offers, the determination of a NEM spot price, the settlement of NEM transactions and the creation of the inter-regional settlements residues account.

292               The ACCC does not admit AGL’s contention that, where there is no interconnect constraint, dispatch offers of electricity generators in any region will affect the price of electricity at the Victorian Regional Reference Node and that, other than for inter-regional transmission losses, the Regional Reference Node prices will be the same. This non-admission is linked to its own description of NEMMCO’s activities in operating the NEM and its contention that there is a range of factors that affect the volume of electricity available at all points in the national grid.

293               There is no real contest about the amount of time in 2002 in which there was an interconnect constraint relevant to the transmission of electricity into Victoria from other regions. The relevant percentages are set out thus:

‘(a) Snowy into Victoria 0.46% of the time

(b) South Australia into Victoria 6.6% of the time

(c) New South Wales into Snowy 0% of the time; and

(d) Queensland into New South Wales 2.81% of the time’ (SC 67)

The ACCC says that the figure for Snowy into Victoria should be 0.51% and the figure for Queensland into New South Wales should be 2.87%. Otherwise the plea is admitted. However the allegation of the average NEM spot price difference between regions in 2002 where there was an interconnect constraint is not admitted.

294               AGL contends that by reason of the matters pleaded in this section of the statement of claim, interstate electricity generators compete with and constrain the pricing of Victorian electricity generators. This is not admitted and the ACCC goes on to allege that between 1 July 1999 and 30 June 2003 the average NEM spot prices in each of the regions differed. The time weighted average annual NEM spot prices in each of the regions for each financial year between 1 July 1999 and 30 June 2003 as recorded by NEMMCO are set out in Schedule B to the defence.

 

The Pleading – Supply, Demand and Pricing in the Wholesale Market

295               The average NEM spot price at the Victorian Regional Reference Node for the period from 1 July 2001 to June 2002 was $30.97/MWh and in the period from 1 July 2002 to 30 June 2003 was $27.54/MWh. That is pleaded by AGL and admitted by the ACCC. The ACCC however makes additional reference to the time weighted average NEM spot prices in the financial years ending 30 June 2000 and 30 June 2001 which it says were $26.11/MWh and $45.39/MWh. The ACCC and AGL are in contention on AGL’s pleading that the average NEM spot price at the Victorian Regional Reference Node for the financial years 30 June 2002 and 30 June 2003 was less than the costs of constructing and operating a brown-coal electricity generation plant equivalent to Loy Yang A on an MWh basis and less than the cost of constructing and operating a new electricity generation plant in the NEM on that basis. These claims are not admitted by the ACCC which says that the average NEM spot price is not the sole source of revenue earned or capable of being earned by an electricity generator equivalent to Loy Yang A.

296               In respect of electricity generation capacity available to meet Victorian demand, AGL says it substantially exceeds the level of average and peak demand in Victoria and is likely to continue to do so when considered with likely expansions in generation capacity within Victoria and interconnection capacity into Victoria. This is not admitted. The ACCC says that the reserve margin as calculated by NEMMCO in Victoria, is less than the minimum reserve level for Victoria, that being the capacity of the single largest generating unit in the region. The ACCC also denies that the capacity of other electricity generators and interconnector capacity is sufficient to meet Victorian peak demand and NEM peak demand such that it would not require to dispatch Loy Yang A in order to meet that demand. It is denied that the Loy Yang Business would forego substantial revenues necessary to offset its fixed costs if it did not generate electricity for dispatch into the NEM at or near its full capacity almost all of the time.

297               AGL says that dispatch offers made by the Loy Yang Business in relation to Loy Yang A in the financial years ended 30 June 2002 and 30 June 2003 set the Victorian NEM spot price approximately 5% and 10% of the time respectively and for 98.7% and 99.5% of the time referred to in each case the price was under $35/MWh. The important plea is made that in the event that the average NEM spot price at the Victorian Regional Reference Node were expected by participants or potential participants to exceed the long run average costs of the production of such electricity, then existing power generators could readily expand their generation capacity and new generators could readily build new plants. In addition existing interconnection transmission capacity could readily be augmented. All these claims are denied by the ACCC.

298               AGL says that the Loy Yang business is unable to sustainably and profitably price electricity above the average long run costs of generating it which represents the competitive level of such prices and is unable to sustainably and profitably withhold capacity from being available for generation at the competitive level. It is also said that the Loy Yang Business is competitively constrained in the supply of electricity into the NEM. This allegation is denied by the ACCC.

The Pleadings – Electricity Derivative Contracts and Spot Price Management

299               The next section of the statement of claim deals with the management of spot price volatility largely by reference to electricity derivative contracts. That spot price varies in the NEM and that its variability is called the ‘spot price volatility’ and is a substantial risk to Victorian Electricity Retailers and electricity generators is common ground. The ACCC says however that spot price volatility also represents a business opportunity for retailers and generators. It is common ground that retailers enter into electricity derivative contracts with generators or other parties or through trading exchanges to provide a hedge against their obligations to settle at the NEM spot price. It is also admitted that retailers agree with customers to adopt part of the wholesale price risk or otherwise influence customer demand patterns. It is not admitted that retailers can and do seek to manage their exposure by acquiring or developing electricity generators or entering into contracts with other owners or operators of such generators whereby the retailer can determine when and how much electricity that generator will dispatch.

300               While admitting that electricity generators enter into derivative contracts to offset their exposure to the NEM spot price, the ACCC denies that they have strong commercial incentives to make available competitively priced electricity derivative contracts so as to hedge their exposure. It admits that characteristically they sell electricity into the NEM and enter into electricity derivative contracts. The allegation that generators outside Victoria enter into electricity derivative contracts at the Regional Reference Node is admitted, but it is said that generators do not enter into such contracts to a significant or material extent. The same is said to be true of traders and intermediaries. The plea that generators can hedge against inter-regional price differences through inter-regional settlement residue auctions conducted by NEMMCO, is admitted but again it is said that they do not do so to any significant or material extent. AGL claims that Victorian Electricity Retailers can and do source electricity derivative contracts referenced to the Victorian Regional Reference Node and an interstate Regional Reference Node. The ACCC says that the latter is not done to any significant or material extent. The same is said to be true of resort by retailers to inter-regional settlement residue auctions and other inter-regional hedge products made available by other counterparties.

301               While the ACCC admits that, before the completion of the proposed acquisitions, there are available counterparties with the capacity to enter into electricity derivative contracts with Victorian Electricity Retailers, it denies AGL’s plea that there is a ready availability of counterparties including both Victorian and interstate electricity generators, with the capacity to enter into electricity derivative contracts with the retailers. It admits that Victorian generators compete to enter into electricity derivative contracts with Victorian Electricity Retailers as alleged by AGL. It admits also that generators and other counterparties have regard to their forecast of average NEM spot prices in determining the prices and conditions upon which they will enter into electricity derivative contracts. However the ACCC says that is not the only matter to which generators and counterparties have regard. And although it also admits that Victorian Electricity Retailers have regard to their forecast of average NEM spot prices in determining the prices and conditions upon which they will enter into electricity derivative contracts, these are not the only matters to which they have regard. It does not admit that average NEM spot prices and the prices, terms and conditions of electricity derivative contracts are closely correlated. It denies that the Loy Yang Business is competitively constrained by other suppliers of electricity derivative contracts. (D 88)

302               The statement of claim deals with the definition of electricity generator markets. That matter has been referred to earlier in the section of this summary dealing with the market definitions adopted by the parties.

The Pleadings – AGL’s Post-acquisition Control of Loy Yang

303               There is then a section of the statement of claim headed ‘AGL has no ability or incentive to control the operations of Loy Yang A to benefit its retail operations’. This encompasses pars 104 to 107.

304               AGL says that because of the competitive position in the Victorian retail markets it has no ability to control the operations of the Loy Yang Business to benefit its retail operations and/or disadvantage the retail operations of its competitors and no incentive to do so. The ACCC admits that this is the case now but denies that it will be the case after the acquisitions. AGL pleads that by reason of the competitive constraints on electricity generators it has, and following the proposed acquisitions will have, no ability or incentive to raise the dispatch offers of Loy Yang A or to withhold any of its generation capacity or raise the price of electricity derivative contracts entered into by the Loy Yang Business. Nor will it have the ability or the incentive to withhold any of the notional quantity of electricity hedged under electricity derivative contracts otherwise entered into by the Loy Yang Business or to control the operations of the business to benefit its retail operations and disadvantage the retail operations of its competitors. The ACCC admits this plea in so far as it relates to the position in the absence of the proposed acquisitions but otherwise denies the allegations.

305               AGL invokes the ownership structure of GEAC and the GEAC interests in the Loy Yang Business to support the claim that it lacks, and following the proposed acquisitions, will lack, the ability or incentive to do the various things already referred to. It separately relies upon the arrangements for the operation of the Loy Yang Business by GEAC OpCo to support the want of such ability or incentive. In each case the ACCC denies the allegations. It says that, in the absence of binding undertakings given to the Court by each of the parties to the proposed acquisitions, that they will give effect to the GEAC Shareholders Agreement and the GEAC Subscription Deed consistently with their terms, the allegations are irrelevant.

The Pleadings – The Section 50(3) Factors

306               The statement of claim then turns to the various factors to be taken into account in s 50(3) of the Act in determining whether or not the proposed acquisition is likely to have the effect of substantially lessening competition in the relevant market. It is not necessary to set out these contentions here save to say that the ACCC takes issue either by non-admission or denial of most of AGL’s pleadings in relation to the operation of those factors.

307               The statement of claim concludes by asserting:

‘By reason of the matters set out herein, and having regard to the factors in s 50(3) of the Act, the Proposed Acquisitions would not have the effect, and would not be likely to have the effect, of substantially lessening competition in:

(a) any of the Victorian Retail Markets; and

(b) any of the Electricity Generator Markets

in contravention of s 50 of the Act.’ (SC 117)

It is to be noted that the reference to the electricity generator markets includes the distinct electricity and derivative contracts markets for which the ACCC contends. The ACCC does not admit this allegation.

 

The Relief Sought

308               AGL claims the following declarations:

‘1. A declaration that the acquisition by The Australian Gas Light Company (AGL) of shares in Great Energy Alliance Corporation Pty Limited (GEAC) pursuant to the GEAC Subscription Deed dated 3 July 2003 would not have the effect, or would not be likely to have the effect, of substantially lessening competition in a market in contravention of section 50 of the Trade Practices Act 1974.

2. A declaration that the acquisition by GEAC Operations Pty Limited (GEAC OpCo), a wholly owned subsidiary of GEAC, of the Loy Yang Sale Shares (as that term is defined in the Statement of Claim) would not have the effect, or would not be likely to have the effect, of substantially lessening competition in a market in contravention of section 50 of the Trade Practices Act 1974.

3. A declaration that the acquisition by AGL of shares in GEAC pursuant to the GEAC Subscription Deed dated 3 July 2003 in combination with the acquisition by GEAC OpCo of the Loy Yang Sale Shares would not have the effect, or would not be likely to have the effect, of substantially lessening competition in the market in contravention of section 50 of the Trade Practices Act 1974.

4. Such further or other orders as the Court may consider appropriate.

5. The First Respondent pay the Applicant’s costs of these proceedings.’

309               As to this relief, the ACCC pleads in par 111 of the Defence:

‘AGL is not entitled to the relief sought in the Application because:

(a) by reason of the matters alleged in paragraphs 10 to 26 inclusive of the Statement of Claim and the absence of the undertakings referred to in paragraphs 11, 13, 16, 96 and 97 above, the proceeding is not a matter within the meaning of:

(i) Chapter III of the Constitution;

(ii) section 39B(1A)(c) of the Judiciary Act 1903;

(iii) section 21 of the Federal Court of Australia Act 1976; or

(iv) section 163A of the Trade Practices Act 1974;

(b) the Court should not, in the exercise of its discretion, make such declarations.’

310               The jurisdictional objection taken in par 111(a) has already been dealt with as a preliminary question and dismissed – Australian Gas Light Company (ACN 052 167 405) v Australian Competition & Consumer Commission (No 2) [2003] FCA 1229.

The ACCC’S Contentions About the Ways in Which The Proposed Acquisition Will Cause or be Likely to Cause a Substantial Lessening of Competition.

311               The ACCC was directed at a case management conference prior to trial to specify its competition case by filing an outline of the ways in which it says the proposed acquisition will have the effect or be likely to have the effect of a substantial lessening of competition in a market. It is convenient to set out that statement:

‘1 It is assumed that AGL will not have any direct managerial influence or control over the day-to-day operations of Loy Yang. If that assumption is incorrect, the likelihood that the acquisition creates a substantial lessening of competition is greatly increased. If the assumption is true, then:

1.1 The acquisition of Loy Yang by AGL creates a ‘natural hedge’ against electricity spot price risk for AGL;

1.2 This natural hedge will cause AGL to reduce its demand for arms-length electricity derivative contracts;

1.3 This reduction in AGL’s demand will be approximately equal to the amount of the natural hedge;

1.4 The demand for these contracts by retailers is relatively insensitive so that the reduction in AGL’s demand will most likely lead to an equivalent reduction in the volume of hedge contracts traded;

1.5 This will cause generators in the NEM to have increased unhedged generation capacity, the amount of the increase being equivalent to the amount of AGL’s natural hedge;

1.6 Because AGL’s natural hedge is with Loy Yang and hedge contracts with alternative counterparty generators are imperfect substitutes, the most likely generators to experience a reduction in their hedged capacity will be Loy Yang and possibly other Victorian brown-coal base-load generators;

1.7 AGL’s own vertical integration business strategy is to maximise its profits by creating pool [ie spot] price volatility. Therefore, AGL can reasonably be expected to use the reduction in hedge contract demand, created by the natural hedge, to increase pool price volatility by increasing the total unhedged capacity of Loy Yang, or alternatively Hazelwood and/or Yallourn;

1.8 An increase in unhedged generation capacity increases the incentive for the relevant generators to raise the spot price of electricity;

1.9 The empirical analysis undertaken by Professor Wolak shows that the brown-coal, base-load generators in Victoria have the ability to influence and materially raise the average spot prices of electricity in Victoria and the NEM;

1.10 It is likely that the reduction in the volume of hedge contracts as a result of the acquisition will lead to a material increase in the average wholesale spot price of electricity in Victoria and the NEM and a resultant increase in forward hedge prices;

1.11 The increase in the average spot price of electricity in Victoria and the NEM predicted by Professor Wolak is direct empirical evidence that the acquisition of 35% of Loy Yang by AGL is likely to cause a substantial lessening of competition in the NEM.

2 In addition (again under the “passive investment” assumption) AGL’s passive interest in Loy Yang will lead to the following outcomes:

2.1 Increase AGL’s relative demand for contracts from Loy Yang and decrease its relative demand for contracts from other generators. AGL receives an implicit discount on premiums on hedge contracts it has with Loy Yang, due to its 35% equity interest. Therefore, if the strike price offered by Loy Yang and another generator is the same, AGL will find it more profitable to enter into additional hedge contracts with Loy Yang than with the other generator.

2.2 The follow on effects from this are as follows:

2.2.1 Loy Yang will contract relatively more with AGL than other retailers (as AGL faces an implicit discount in contracting with Loy Yang) and so may offer more favourable terms to deal with Loy Yang.

2.2.2 Other retailers and generators will contract relatively more with one another rather than with AGL and Loy Yang. Other retailers will be reluctant to enter into hedge contracts with Loy Yang as they will be concerned about the confidentiality of their hedging arrangements where Loy Yang is the counter party, as Loy Yang is partially owned by one of their major retail competitors in Victoria. Conversely generators will be reluctant to enter into hedge contracts with AGL as they will be concerned about the confidentiality of their hedging arrangements where AGL is the counter-party, as AGL is the owner of an interest in one of their largest competitors in Victoria. Therefore third parties will favour contracting with other generators and retailers.

2.2.3 To the extent that, having a portfolio of contracts with a diverse number of firms is optimal for risk management, the costs of achieving this optimal portfolio will rise as the costs of hedging with Loy Yang and AGL may rise because AGL and Loy Yang prefer post acquisition to contract with each other and therefore will require a premium to contract with a third party.

3 The points in paragraphs 1 and 2 above will create an incentive for other generators and retailers to merge. That is, there be bandwagon mergers.

4 As a result new entrants into generation and retailing will have a reduced number of effective contracting partners. That is, the contract market will become thinner.

 

5 This could forestall entry as entry is required at both the generation and retail functional levels. Indeed, this is more likely the more vertical integration there is amongst major players in the industry.

6 The management of a subsidiary will consider the impacts of its decisions upon the business and profit of its shareholders. It is highly unlikely that a subsidiary will take action that would adversely affect the business of a shareholder, if only for fear of dismissal or adverse determination of compensation arrangements. Whilst not requiring actual control or influence by AGL of Loy Yang post merger, Loy Yang management are more likely to consider the impacts upon AGL’s business and profits when setting their longer-term strategic options. If Loy Yang, even in part, takes into account the impact on AGL’s business and profits of its decisions following the acquisition, all of the above will still occur but with greater strength. That is,

6.1 The internal contracting incentives of Loy Yang and AGL will be stronger.

6.2 The costs of contracting with Loy Yang and AGL will be higher.

6.3 Loy Yang will have an incentive to increase its spot market bids as the internal contracting volume becomes higher.

6.4 The impact on other generators and retailers will be greater and the bandwagon motivation stronger.

6.5 The effective contract market available to new entrants will become even thinner.

6.6 This increases the chance that entry into retailing or generation will be forestalled.

7 The above competition issues are based upon the assumption that AGL will not have direct influence or control over the day-to-day operations of Loy Yang (totally passive investment). This assumption is based upon contractual arrangements between the GEAC members. However, these contractual arrangements can be amended or waived by the parties. Therefore, without AGL and relevant parties providing binding undertakings to the Court to ensure that AGL’s interest in Loy Yang remains passive this assumption is questionable.

8 Even if AGL does not have direct influence or control over the day-to-day operations of Loy Yang, the acquisition increases the likelihood that Loy Yang and AGL would enter into arrangements that change the incentive of Loy Yang to exercise its market power. For example, AGL and Loy Yang may enter into a risk-sharing mechanism that would provide an incentive for Loy Yang to reduce its contract cover, given that the potential strategy may result in reduced profits and/or increased risk for Loy Yang as a stand-alone entity and AGL only has a partial interest in Loy Yang and so no ability to control its behaviour. Such a risk-sharing mechanism may insure Loy Yang against the adverse profit realisations from Loy Yang’s lower forward contract holdings in exchange for Loy Yang sharing the profits with AGL during the favourable profit realisations from Loy Yang’s lower forward contract holdings. Such a strategy would be expected to take the form of a wholesale electricity agreement between AGL and Loy Yang (ie a hedge contract). As a result:

8.1 An increase in the unhedged generation capacity of Loy Yang increases the ability of Loy Yang to raise the spot price of electricity;

8.2 A risk sharing agreement between AGL and Loy Yang creates the incentive for Loy Yang to exercise its market power;

8.3 The empirical results of Professor Wolak show that Loy Yang has the ability to influence the average spot price of electricity in Victoria and the NEM;

8.4 The empirical results of Professor Wolak demonstrate that even a small reduction of Loy Yang’s level of contracting of approximately 140MW will cause an increase the average Victorian spot price by 20% over the relevant period;

8.5 As stated by Henry Ergas in his report dated 10 October 2003 at para 230 “a change in competitive conditions which affected pool prices would in all likelihood significantly affect contract prices, and the relevant impact on the community would occur through this joint impact”; and

8.6 As a result AGL’s retail competitor’s costs for the acquisition of electricity both from the spot market and through the hedge market would substantially increase.

9 To address such concerns GEAC, AGL and Loy Yang would need to provide binding behavioural undertakings to:

9.1 limit the ability and incentive of Loy Yang post-merger to exercise market power; and

9.2 limit or restrict the level and nature of hedging between Loy Yang and AGL to ensure there is no foreclosure of AGL’s retail competitors.

10 If AGL does have some influence or control over the day-to-day operations of Loy Yang then, because of the significant degree of market power held by Loy Yang (as found by Professor Wolak), it is likely that the acquisition will increase Loy Yang’s incentive to exercise its market power to raise the spot price. An increase in the spot price of electricity will substantially lessen competition in the Victorian electricity retail market. AGL’s competitors will face increased costs for spot market electricity purchases and hedge contracts, which AGL is indifferent to due to its natural hedge and the near perfect coverage provided to it by the Deemed Profile Hedge Agreement with Loy Yang.

11 The acquisition is likely to decrease the incentive for AGL and Loy Yang to contract with other retailers thereby foreclosing competing retailers from hedge contracts with Victorian base load generators.’

 

The Statutory Framework

312               Section 50 of the Trade Practices Act, in the relevant parts, provides:

‘50(1) A corporation must not directly or indirectly:

(a) acquire shares in the capital of a body corporate; or

(b) acquire any assets of a person;

if the acquisition would have the effect, or be likely to have the effect, of substantially lessening competition in a market.

…

(3) Without limiting the matters that may be taken into account for the purposes of subsections (1) and (2) in determining whether the acquisition would have the effect, or be likely to have the effect, of substantially lessening competition in a market, the following matters must be taken into account:

(a) the actual and potential level of import competition in the market;

(b) the height of barriers to entry to the market;

(c) the level of concentration in the market;

(d) the degree of countervailing power in the market;

(e) the likelihood that the acquisition would result in the acquirer being able to significantly and sustainably increase prices or profit margins;

(f) the extent to which substitutes are available in the market or are likely to be available in the market;

(g) the dynamic characteristics of the market, including growth, innovation and product differentiation;

(h) the likelihood that the acquisition would result in the removal from the market of a vigorous and effective competitor;

(i) the nature and extent of vertical integration in the market.

…

(4) Where:

(a) a person has entered into a contract to acquire shares in the capital of a body corporate or assets of a person;

(b) the contract is subject to a condition that the provisions of the contract relating to the acquisition will not come into force unless and until the person has been granted an authorization to acquire the shares or assets; and

(c) the person applied for the grant of such an authorization before the expiration of 14 days after the contract was entered into;

the acquisition of the shares or assets shall not be regarded for the purposes of this Act as having taken place in pursuance of the contract before:

(d) the application for the authorization is disposed of; or

(e) the contract ceases to be subject to the condition:

whichever first happens.

…

(6) In this section:

market means a substantial market for goods or services in:

(a) Australia; or

(b) a State; or

(c) a Territory; or

(d) a region of Australia.’

313               Some of the terms used in s 50 are the subject of definition provisions in the Act. Relevantly, s 4(4) provides, with respect to the acquisition of shares:

‘4(4) In this Act:

(a) a reference to the acquisition of shares in the capital of a body corporate shall be construed as a reference to an acquisition, whether alone or jointly with another person, of any legal or equitable interest in such shares; …’

314               The concept of ‘market’ is defined non-exhaustively in s 4E:

‘4E. For the purposes of this Act, unless the contrary intention appears, market means a market in Australia and, when used in relation to any goods or services, includes a market for those goods or services and other goods or services that are substitutable for, or otherwise competitive with, the first-mentioned goods or services.’

315               The lessening of competition is referred to in s 4G:

‘4G. For the purposes of this Act, references to the lessening of competition shall be read as including references to preventing or hindering competition.’

316               There is provision in the Act for the acceptance by the ACCC of written undertakings. This is to be found in s 87B:

‘(1) The Commission may accept a written undertaking given by a person for the purposes of this section in connection with a matter in relation to which the Commission has a power or function under this Act (other than Part X).

(2) The person may withdraw or vary the undertaking at any time, but only with the consent of the Commission.

(3) If the Commission considers that the person who gave the undertaking has breached any of its terms, the Commission may apply to the Court for an order under subsection (4).

(4) If the Court is satisfied that the person has breached a term of the undertaking, the Court may make all or any of the following orders:

(a) an order directing the person to comply with that term of the undertaking;

(b) an order directing the person to pay to the Commonwealth an amount up to the amount of any financial benefit that the person has obtained directly or indirectly and that is reasonably attributable to the breach;

(c) any order that the Court considers appropriate directing the person to compensate any other person who has suffered loss or damage as a result of the breach;

(d) any other order that the Court considers appropriate.’

317               In the present case a declaration is sought under s 163A which relevantly provides:

‘(1) Subject to this section, a person may, in relation to a matter arising under this Act, institute a proceeding in a court having jurisdiction to hear and determine proceedings under this section seeking the making of:

(a) a declaration in relation to the operation or effect of any provision of this Act other than the following provisions:

(i) Division 2, 2A or 3 of Part V;

(ia) Part VB;

(ii) Part XIB;

(iii) Part XIC; or

(aa) a declaration in relation to the validity of any act or thing done, proposed to be done or purporting to have been done under this Act; or

(b) an order by way of, or in the nature of, prohibition, certiorari or mandamus;

or both such a declaration and such an order.

(2) Subject to subsection (2A), the Minister may institute a proceeding under this section and may intervene in any proceeding instituted under this section or in a proceeding instituted otherwise than under this section in which a party is seeking the making of a declaration of a kind mentioned in paragraph (1)(a) or (aa) or an order of a kind mentioned in paragraph (1)(b).

(2A) Subsections (1) and (2) do not permit the Minister:

(a) to institute a proceeding seeking a declaration, or an order described in paragraph (1)(b), that relates to Part IV; or

(b) to intervene in a proceeding so far as it relates to a matter that arises under Part IV.

(3) The Commission may institute a proceeding in the Court seeking, in relation to a matter arising under this Act, the making of a declaration of the kind that may be made under paragraph (1)(a).

(3A) In so far as this section has effect as a law of the Commonwealth, the Federal Court has jurisdiction to hear and determine proceedings under this section.’

Subsections (4) and (5) are not relevant for present purposes.

318               In addition to the jurisdiction conferred upon the Court by s 163A the Court has a global jurisdiction in matters arising under laws of the Commonwealth conferred by s 39B of the Judiciary Act 1903 (Cth) and, in particular, s 39B(1A) which provides:

‘The original jurisdiction of the Federal Court of Australia also includes jurisdiction in any matter:

(a) in which the Commonwealth is seeking an injunction or a declaration; or

(b) arising under the Constitution, or involving its interpretation; or

(c) arising under any laws made by the Parliament, other than a matter in respect of which a criminal prosecution is instituted or any other criminal matter.’

319               The powers of the Court in matters in which it has jurisdiction are defined in part by ss 21 and 23 of the Federal Court of Australia Act 1976 (Cth) which provide:

‘21(1) The Court may, in relation to a matter in which it has original jurisdiction, make binding declarations of right, whether or not any consequential relief is or could be claimed.

(2) A suit is not open to objection on the ground that a declaratory order only is sought.

…

23. The Court has power, in relation to matters in which it has jurisdiction, to make orders of such kinds, including interlocutory orders, and to issue or direct the issue of, writs of such kinds, as the Court thinks appropriate.’

A Brief History of Section 50

320               Section 50(1) of the Trade Practices Act 1974 in its original form provided:

‘50(1) A corporation shall not acquire, directly or indirectly, any shares in the capital, or any assets, of a body corporate where the acquisition is likely to have the effect of substantially lessening competition in a market for goods or services.’

Subsection (2) provided that the section did not apply to an acquisition of the assets of a body corporate in the ordinary course of business. Subsections (3) and (4) related to acquisitions conditional upon authorisation under the provisions of the Act. The history of its development is well known. It is helpfully set out by Professor Brian Johns in Threshold Tests for the Control of Mergers: The Australian Experience (1994) 9 (5) Review of Industrial Organisation 649. The following outline is based in part on that paper.

321               Section 50 was the first Commonwealth competition law directed to the control of mergers and acquisitions. Although, as Attorney-General, Sir Garfield Barwick, made proposals in 1962 for the control of anti-competitive mergers his proposals were not reflected in the Trade Practices Act 1965 nor in the Trade Practices Act 1971. In the Second Reading Speech for the 1965 Act the Attorney-General of the day, Mr Snedden, described the control of mergers and takeovers as a ‘problem … of great complexity’ – H of R Parl Deb 19/5/65.

322               In his Second Reading Speech for the Trade Practices Act 1974, the Attorney-General, Senator Murphy, referred to cl 50 of the Bill which was to become s 50. He said:

‘Legislation controlling anti-competitive practices is incomplete if it does not also provide for the control of anti-competitive mergers.’ Sen Parl Deb 30/7/74 p 545

 

He added:

‘Mergers are prohibited by clause 50 where a likely effect would be to substantially lessen competition in a market. Questions will of course arise as to whether particular mergers are likely to have such an effect on competition. In this regard the provisions for clearances and for authorisations will be available and will enable businesses to resolve their uncertainties.’ Sen Parl Deb 30/7/74 p 545

 

323               Although there was a number of authorisations and clearances granted in respect of mergers during the early life of the Trade Practices Commission, there was no judicial consideration of the first s 50. The seminal decision of the Trade Practices Tribunal in Re Queensland Co-operative Milling Association; Re Defiance Holdings Ltd (1976) 25 FLR 169 (QCMA), concerned an application for review of the Commission’s refusal to authorise alternative proposed mergers of flour milling companies in Queensland. The Tribunal, which refused authorisation, had much of lasting value to say about markets and competition but little about the proper construction of s 50. This was understandable for it was concerned with the authorisation process under s 90(5) of the Act. It did not believe ‘… that the mere act of application for authorization should carry with it any presumption as to liability under s 50 or, more generally as to the presence of ‘“significant” or “substantial” anti-competitive effects’ – at 508. A similar approach was taken by the Tribunal in Re Howard Smith Industries Pty Ltd (1977) 15 ALR 645.

324               In the Howard Smith case, the Tribunal observed that it was prohibited by s 90(5) of the Act from granting authorisation unless satisfied as to enumerated matters. It then said at p 673:

‘This involves the Tribunal in a consideration of what is likely to result, in this case a consideration of what is likely to result from the proposed merger if authorised. This involves a consideration of commercial likelihoods. As part of that consideration, it is necessary to consider what is likely to result if the proposed merger is not authorised. This does not mean that the likely effects must be more probable than not, but rather there must be a tendency or real possibility of a particular result following the refusal of an authorisation. This is an aspect of the main matter for consideration, namely commercial likelihoods resulting from the proposed merger if authorised.’

325               With the change of government in 1975 the Trade Practices Review Committee chaired by Mr TB Swanson, was established to review the operation of the Act. The Swanson Committee Report, published in August 1976, concluded that a merger law was necessary but that its application should not be as sweeping as the law then in force (par 8.7). It proposed the introduction of a monetary threshold which would exclude, from the application of the section, acquisitions involving small companies. The Committee was of the view that there was a distinction to be drawn between threshold tests for restrictive agreements and practices generally and threshold tests for anti-competitive mergers. In the Committee’s view different considerations apply to each:

‘Whereas a merger occurs at a particular point of time and a threshold test is applicable at that time, in the case of a restriction of competition which continues in time, changing conditions would require a regular review of the operation of the threshold.’ (4.25)

 

326               In relation to the importance of market structure in assessing the anti-competitive effects of a merger, the Committee said:

‘The committee is of the view that it is proper for the Commission and others called upon to judge competitive effects, to assess those effects by reference to conduct considerations as well as structural considerations because market structure of itself should not be used to impute conduct.’(8.41)

327               Following the Swanson Report the Act was amended in July 1977. The threshold for the prohibition was raised by the introduction of a post-acquisition control and dominant position test. The section after the amendment provided:

‘50(1) A corporation shall not acquire, directly or indirectly, any shares in the capital, or any assets, of a body corporate if-

(a) as a result of the acquisition, the corporation would be, or be likely to be, in a position to control or dominate a market for goods or services; or

(b) in a case where the corporation is in a position to control or dominate a market for goods or services-

(i) the body corporate or another body corporate that is related to that body corporate is, or is likely to be, a competitor of the corporation or of a body corporate that is related to the corporation; and

(ii) the acquisition would, or would be likely to, substantially strengthen the power of the corporation to control or dominate that market.’

328               Reference to a market for goods or services was to be construed as a reference to ‘a substantial market for goods or services in Australia, or in a State’ (s 50(3)(a)). The reference to controlling or dominating a market for goods or services was to be construed as a reference to controlling or dominating such a market either as the supplier or as an acquirer of goods or services in that market (s 50(3)(b)). It was intended that the categories of mergers to be subject to the Act should be ‘quite limited’ – Parl Deb H of R 3/5/77 at p 1478 – Second Reading Speech. In its Annual Report of 1977-78 the Trade Practices Commission observed that merger activity had increased after the amendments and that few companies now applied for authorisation because generally they felt they were not at risk of breaching s 50.

329               In 1986, the Labor Government released a Green Paper entitled ‘The Trade Practices Act – Proposals for Change’. One of the proposals included in the paper proposed reverting the threshold test for mergers to that applying between 1974 and 1977. In the event the 1986 amendments to the Trade Practices Act only made a minor change to s 50. The reference to ‘control’ of a market was deleted leaving dominance or increased dominance of a market as the sole criterion for determining whether or not a proposed acquisition would breach the law.

330               Following the 1986 amendment, the House of Representatives Standing Committee on Legal and Constitutional Affairs found insufficient evidence to justify a change in the existing dominance test. This was the Report of the Griffiths Committee – Mergers, Takeovers and Monopolies: Profiting from Competition, House of Representatives Standing Committee on Legal and Constitutional Affairs(1989) Canberra AGPS. A dissenting report by two of the Committee’s members supported reintroduction of the substantial lessening of competition test. They saw the fundamental problem with the existing s 50 as a failure to recognise that a corporation could be in a position to engage in anti-competitive conduct without dominating a market.

331               The question of the appropriate test for mergers and acquisitions was referred to the Senate Standing Committee on Legal and Constitutional Affairs then chaired by Senator Cooney. That Committee reported in December 1991 and recommended by majority that s 50 be amended to reintroduce the threshold test of substantially lessening of competition in a substantial market – Mergers Monopolies and Acquisitions: Adequacy of Existing Legislative Controls, Senate Standing Committee on Legal and Constitutional Affairs (1991) Canberra AGPS.

332               Section 50 was amended to its present form by the Trade Practices Legislation Amendment Act 1992 (Cth). In the Second Reading Speech the then Attorney-General, Mr Michael Duffy, referred to merger control as an important element of any law aiming to preserve levels of competition. He observed that mergers can lessen competition, potentially providing increased scope for price rises or collusive behaviour and lessening dynamic factors such as the rate of innovation. These detriments underpinned government intervention in the area of mergers. He acknowledged that mergers could be a valuable source of increased efficiency or other public benefits.

333               Explaining the change in the threshold for the operation of the s 50 prohibition he said:

‘After much consideration the Government has decided to amend section 50 to prohibit mergers or acquisitions which are likely to substantially lessen competition and which have not been authorised by the Commission. In an Act which seeks to preserve competition it is appropriate that the merger test should focus on the effect on competition in a market rather than on the dominance of a particular firm. The effect of the amendment will be to broaden the range of transactions which can be examined under section 50. This can only be procompetitive.’ Parl Deb H of R, 3/11/92 p 2406

 

He also referred to the factors mentioned in s 50(3) as assisting businesses to understand the new test and generally adding ‘… certainty as to the policy intent underlying the changed merger test’ – Parl Deb H of R 3/11/92 p 2406.

334               The Explanatory Memorandum circulated with the Bill contrasted the previous market dominance test with the new test of substantially lessening competition. It said, at par 11:

‘The previous test of market dominance has been interpreted by the court as a situation where one firm has a commanding influence in the market. It is a test which focuses largely on changes to the structure of a market that would be affected by the acquisition but it also takes some account of the likely effect on the competitive process of such an acquisition. The substantial lessening test focuses on changes to the state of competition in the relevant market. As the Trade Practices Act is about competition, a test which concentrates on competition and whether there is a lessening of that competition is more consistent with the policy underlying the legislation.’

The Explanatory Memorandum discussed the use of the term ‘substantially lessening competition’ throughout the principal Act and said:

‘It is here intended to mean an effect on competition which is real or of substance, not one which must be large or weighty.’

The Explanatory Memorandum emphasised that under the new merger test it was the effect on competition which would be important rather than the particular position of the acquiring firm. As to the operation of subs 50(3) that subsection was described as providing a non-exhaustive list of matters to be taken into account in determining whether an acquisition would have the effect or be likely to have the effect of substantially lessening competition. Those matters were said to consist of ‘well-understood economic concepts which are considered in determining whether competition would be, or is likely to be, substantially lessened’. The Explanatory Memorandum cited the observation of Bowen CJ and Fisher J in Outboard Marine Australia Pty Ltd v Hecar Investments (No 6) Pty Ltd (1982) 44 ALR 667 that:

‘[t]he economic meaning must be applied in a practical way to accommodate the concern of the Act with business and commerce.’

The list of factors was not intended to affect the interpretation of the phrase ‘substantially lessening competition’ in other provisions of the principal Act.

335               The Cooney Committee, in considering the threshold at which the prohibition in s 50 would apply, was presented with argument on one side that a high threshold was necessary to allow mergers to occur so that local firms could grow to a size at which they would be internationally competitive. On the other side, it was put that mergers reduce incentive to develop industry efficiency, reduce international competitiveness and facilitate the concentration of economic power. One of the arguments for reintroducing the substantial lessening of competition test was that it would create consistency in the statutory scheme of Pt IV – Cooney 3.94.

336               The Cooney Committee concluded, inter alia, that the existence of a dominance test in the area of merger regulation was ‘… difficult to reconcile with the essential thrust of the Act which is directed to preventing anticompetitive conduct’ (par 3.110). The alternative threshold of substantially lessening competition had operated throughout Pt IV of the Act since 1974 and had accumulated a body of interpretive law (par 3.118). There is no doubt that the Cooney Committee contemplated that the meaning of substantial lessening of competition in s 50 would be the same as its meaning in the other provisions of Pt IV in which it appears.

The Construction of Section 50

337               The construction of s 50 of the Trade Practices Act begins, as does all statutory construction, with the words of the section. The section takes the form of a general prohibition subject to a specific condition. The relevant elements of the prohibition are as follows:

1. A corporation

2. shall not directly or indirectly acquire shares

3. in the capital of a body corporate.

The condition upon which the prohibition operates is that:

1. The acquisition would have the effect;

2. or be likely to have the effect;

3. of substantially lessening competition in a market.

Turning first to the prohibition. It is directed to corporations. The definition of ‘corporation’ in s 4 includes ‘… a trading corporation formed within the limits of Australia’. The case law on what constitutes a trading corporation is well established. There is no dispute that AGL is a trading corporation and therefore a corporation to which the prohibition applies.

338               The second element requires that the corporation not directly or indirectly acquire shares. Acquisitions of shares are defined by what amounts to a deeming provision in s 4(4)(a) of the Act. A reference to such an acquisition ‘shall be construed as a reference to an acquisition, whether alone or jointly with another person, of any legal or equitable interest in such shares’. On that definition, one transaction may give rise to successive acquisitions for the purposes of s 50. A corporation which enters into a contract to purchase the shares of a body corporate may acquire an equitable interest prior to settlement and thereby acquire the shares pursuant to s 4(4). This will not necessarily attract the prohibition in s 50. Where the acquisition of the legal interest has not been completed and no right subsists in the acquirer as a shareholder in the target body corporate then forging any link to a substantial lessening of competition would be problematic.

339               In the case of a contract subject to a condition precedent, as in the present case, no interest is conveyed until satisfaction of the condition. Broken Hill Pty Co Ltd v Trade Practices Tribunal (1980) 31 ALR 401 concerned an acquisition of shares conditional upon authorisation by the Trade Practices Commission. It conveyed no direct beneficial interest in the shares. A direct beneficial interest was acquired only when the contract became specifically enforceable by an order to convey or transfer. Prior to the condition being satisfied, the purchaser could seek an order to require the vendor to do what it must under the contract to secure fulfilment of the condition. Although this case was decided after the 1977 amendments and the introduction of the control and dominant position tests, it is relevant to the question of acquisition today.

340               The transaction documents provide for acquisition, by the GEA consortium, of shares in GEAC and for GEAC OpCo’s acquisition, in turn, of the shares in the Loy Yang partners. At this point no legally relevant acquisition has occurred because of the conditions precedent to be satisfied, including that in the Share Sale Agreement relating to the possibility of s 50 proceedings. It may be that the acquisition, once completed, will constitute a direct acquisition of shares in GEAC or an indirect acquisition of shares in the Loy Yang partners. No point is made of the distinction for present purposes and in my opinion no point need be made of it. The case was conducted on the common assumption that, for the purposes of s 50, completion of the transaction will constitute an acquisition of shares by AGL representing a 35% interest in the Loy Yang partners. There is no dispute that GEAC and the Loy Yang partners are bodies corporate for the purposes of s 50 although that term is not defined in the Act.

341              Constructional questions of importance arise in relation to the condition that the acquisition would have the effect, or be likely to have effect, of substantially lessening competition in a market. The case was conducted upon the basis that the relevant question is whether the acquisition would be likely to have the effect of substantially lessening competition in a market. That is not surprising given that the certainty required by the words ‘would have the effect’ is at the upper limit of the range of probabilities conveyed by the words ‘likely to have the effect’.

342              The collocation ‘likely to have the effect’ is capable of bearing two meanings. One is that the proposed acquisition will ‘more probably than not’ have the requisite effect. The other is that there is a sufficiently high finite probability that the acquisition will have that effect. The latter construction is sometimes expressed by saying that there is a ‘real chance’ of a substantial lessening of competition resulting from the acquisition. The first construction renders the prohibition imposed by s 50 narrower than the second. In the Second Reading Speech for the 1992 amendments the Attorney-General said that mergers can lessen competition and on the other hand can be a valuable source of increased efficiency or other public benefits. He said:

‘Such possible benefits require that a line be drawn between those mergers which are likely to be beneficial and those which are likely to be detrimental to the community as a whole.’

Although it seems plausible that the word ‘likely’ in that passage was being used in the sense of ‘more probable than not’, such speculation is not a useful guide to construction.

343               In my opinion, having regard to the statutory context provided by the other sections of Pt IV the correct construction is that ‘likely’ refers to a significant finite probability or ‘a real chance’ rather than ‘more probable than not’. Although there has been some divergence in the construction of ‘likely’ in various provisions of the Act, the weight of authority supports the wider view. In Tilmanns Butcheries Pty Ltd v Australasian Meat Industry Employees Union (1979) 42 FLR 331, Deane J thought the word ‘likely’ in the secondary boycott provision, s 45D of the Act, referred to ‘a real chance or possibility’ that the conduct if pursued would cause damage. Bowen CJ, with whom Evatt J agreed, did not think it necessary to decide the question. But in Global Sportsman Pty Ltd v Mirror Newspapers Pty Ltd (1984) 2 FCR 82 at 87, the Full Court, dealing with the construction of s 52 of the Act, cited Deane J and said:

‘Conduct is likely to mislead or deceive if that is a “real or not remote chance or possibility regardless of whether it is less or more than 50%.”’

In News Limited v Australian Rugby League Football Ltd (1996) 64 FCR 410, the Full Court referred to the use of the word ‘likely’ in s 4D(1) and adopted the construction used by Deane J. In the Full Court in Munro Topple & Associates Pty Ltd v Institute of Chartered Accountants in Australia (2002) 122 FCR 110, Heerey J, with whom Black CJ and Tamberlin J agreed, observed at [111], with respect to s 47(10):

‘“Likely” does not mean “more likely than not”.’

His Honour referred to Deane J in Tilmanns Butcheries stating that his reasoning with respect to s 45D was equally applicable to the concept of likely effect in s 47(10).

344               Different views have been expressed from time to time in connection with s 50. In Trade Practices Commission v Ansett Transport Industries (Operations) Pty Ltd (1978) 32 FLR 305, Northrop J was concerned with the construction of the section as it stood after the 1977 amendments. He considered that ‘likely’ meant ‘more probable than not’. A similar view was expressed in connection with s 45 of the Act by Lockhart J in Radio 2UE Sydney Pty Ltd v Stereo FM Pty Ltd (1982) 62 FLR 437 at 445. His Honour acknowledged that the word ‘likely’ is susceptible of various meanings:

‘It may mean “probable” in the sense of more likely than not or more than a fifty per cent chance. It may mean a real or not remote possibility. There are other possible meanings.’

Lockhart J also referred to Tilmanns Butcheries. He said he did not find it necessary to determine the question. The conclusion he reached would be the same whichever construction of the word ‘likely’ was adopted but he went on to say:

‘… I reject the view that, in the context of s 45(2), it means a mere possibility, whether real or not.’

345               In Trade Practices Commission v Australian Iron and Steel Pty Ltd (1990) 22 FCR 305 at 321, Lockhart J, in the context of an application to strike out a statement of claim, considered the nature of the nexus that must exist between the acquisition prohibited by s 50(1) and the position of market dominance or likelihood thereof as a result of the acquisition. His Honour observed that the question had not been fully argued and expressed only tentative views on it. He did not think it useful to attempt to put a gloss upon the words of the Act by substituting other words or by saying that the relationship must be ‘direct’ or ‘immediate’ or that it connoted a ‘sole’ or ‘dominant’ cause. He said:

‘For the purposes of s 50(1) it is not enough that the acquisition is the enabling circumstance or causa sine qua non of the effect on the market to which pars (a) and (b) of that subsection are directed. The acquisition must be either a sufficient cause of the existence or likely existence of the state of dominance or substantial strengthening of the power of dominance in the relevant market or one of a number of causes which together lead to or would be likely to lead to that state.’

His Honour was not satisfied that the nature of the nexus that must exist in s 50(1), between the acquisition and the position of market dominance, was the same as that required by s 90(9) of the Act requiring, in an authorisation context, that public benefit would result or be likely to result from the proposed acquisition.

346               In Heydon, Trade Practices Law (LBC) at 9-380, it is suggested that although the language of s 50 to which those remarks were directed was different from its current language they remain sound, mutatis mutandis. In my opinion the passages referred to and cited from the judgment of Lockhart J do not, in the end, define the concept of ‘likely’ except in the context of a causal nexus between acquisition and the outcome which in that case related to market dominance.

347               The collocation ‘… would have the effect, or be likely to have the effect, of substantially lessening competition’ appears in similar and identical versions in other provisions of Pt IV. It appears in ss 45, 45A, 45B, 45C, 47(10) and 50A. In my opinion that formulation is intended to have the same construction throughout Pt IV. Neither language nor policy mandates a variation in its construction from section to section. In any event as a matter of construction if ‘likely’ simply meant more probable than not, it would be difficult to distinguish the application of that limb of the formula from the application of the first limb which, having regard to the onus of proof applicable in proceedings under Pt IV, could be established on the balance of probabilities.

348               The meaning of ‘likely’ reflecting a ‘real chance or possibility’ does not encompass a mere possibility. The word can offer no quantitative guidance but requires a qualitative judgment about the effects of an acquisition or proposed acquisition. The judgment it requires must not set the bar so high as effectively to expose acquiring corporations to a finding of contravention simply on the basis of possibilities, however plausible they may seem, generated by economic theory alone. On the other hand it must not set the bar so low as effectively to allow all acquisitions to proceed save those with the most obvious, direct and dramatic effects upon competition. By the language it adopts and the function thereby cast upon the Court and the regulator in their consideration of acquisitions s 50 gives effect to a kind of competition risk management policy. The application of that policy, reflected in judgments about the application of the section, must operate in the real world. The assessment of the risk or real chance of a substantial lessening of competition cannot rest upon speculation or theory. To borrow the words of the Tribunal in the Howard Smith case, the Court is concerned with ‘commercial likelihoods relevant to the proposed merger’. The word ‘likely’ has to be applied at a level which is commercially relevant or meaningful as must be the assessment of the substantial lessening of competition under consideration – Rural Press Limited v Australian Competition and Consumer Commission [2003] HCA 75 at [41].

349               The term ‘competition’ was the subject of consideration in the decision of the Trade Practices Tribunal in QCMA at 188-189. The Tribunal regarded ‘rivalrous market behaviour’ as the expression of competition. It is a ‘process rather than a situation’. It requires ‘both that prices should be flexible, reflecting the forces of demand and supply, and that there should be independent rivalry in all dimensions of the price-product-service packages offered to consumers and customers’.

350               Competition in a market is not assessed by a snapshot view of participant behaviour at a particular time. The theatre of competition is a theatre of real actors and shadow actors. The shadows are cast by the potential for new entry. The competitive process is informed by the rivalry of the participants and the potential rivalry of potential participants. Competition so understood is conceptually distinct from the idea of the market and the elements of market structure which may constrain or facilitate it. Those structural elements are referred to, inter alia, in the factors set out in s 50(3) of the Act.

351               The word ‘substantial’ in ‘substantial lessening of competition’ is another term requiring qualitative judgment which suggests that the use of analogues such as ‘large’ or ‘weighty’ would misdirect. It applies to ‘lessening’ which encompasses hindering or preventing competition (s 4G). As I said in Stirling Harbour Services Pty Ltd v Bunbury Port Authority (2000) ATPR 41-752 there is only limited assistance to be gained by replacing the words used in the Act with other words. A description of the kind of judgment required by the word ‘substantially’, which appears recently to have been approved in the High Court, is that the effect of the acquisition be ‘meaningful or relevant to the competitive process’ – Rural Press Ltd at [41]. It is also relevant to the present case to bear in mind that in determining whether the acquisition is likely to have the effect of a substantial lessening of competition, the Court will give little if any weight to short term effects readily corrected by market processes – Universal Music Australia Pty Ltd v Australian Competition and Consumer Commission (2003) 201 ALR 636 at [242].

352               In determining whether it could be said that there is likely to be a substantial lessening of competition in a market it is necessary to consider the future state of the relevant market with and without the proposed acquisition – Dandy Power Equipment Pty Ltd v Mercury Marine Pty Ltd (1982) 64 FLR 238 at 259; Outboard Marine Australia Pty Ltd v Hecar Investments (No 6) Pty Ltd (1982) 44 ALR 667 at 669-70. The parties are ad idem on the appropriateness of that test in principle. The closing submissions for AGL indicated that it was content to assume that the future state of competition would fall to be judged on the basis that LYP would continue to operate and that it would be controlled either by its existing owners, the syndicate of banks who have financed them or a new purchaser.

353               The competitive process under scrutiny with and without the acquisition is competition in a market. That means a substantial market for goods or services in Australia or a State or a Territory or a region of Australia (s 50(6)). The definition in s 50(6) does not exclude the operation of the definition in s 4E which will pick up, in a product market, goods and services substitutable for or otherwise competitive with each other. However the definition in s 50(6) introduces the qualifying term ‘substantial’ before ‘market’. It is suggested in Heydon, Trade Practices Law (LBC) at [9.570] that the aim of the qualification is to exclude from the Act cases where a merger occurs in a very small market. The learned author there observes:

‘Section 50(6) involves sacrificing the interests of those in small markets to the interests of the parties to the merger. If a small merger in a small market were to be unlawful on the ground that it led to the acquiring corporation obtaining market control, though this result may be harsh for the acquiring corporation, the merger would be likely to cause as much damage to competition in that market as would be caused to competition by like events in a much larger market.’

It does not seem likely that the relativity implied by the term ‘substantial’ in s 50(6) relates to the size of other markets in whichever of the geographical areas mentioned in the definition the market is to be found. For there is no lower bound on the size of ‘a region of Australia’. It may be that having regard to s 4E the substantiality of the market in question, even if it be geographically limited to a State or a Territory or a region, is to be judged by reference to Australia as a whole. I express no concluded view on that difficult constructional issue because the present case does not appear to throw up any dispute between the parties that, whichever of their propounded markets is in issue, it is a ‘substantial market’ for the purposes of s 50(6).

354               It is necessary next to consider the matters that, pursuant to s 50(3), must be taken into account in determining whether the acquisition would be likely to have the effect of substantially lessening competition in the market. Brief observations can be made with respect to each of these:

(a) The actual and potential level of import competition in the market. As the Explanatory Memorandum for the Trade Practices Legislation Amendment Bill 1992 set out, this matter refers to import competition from outside Australia. It did not figure explicitly in the present case although it may have application indirectly by reference to potential market participants from outside Australia interested in establishing generation capacity within the NEM. It is notable, in that context, that a very substantial international energy company is a major participant in the proposed consortium and that the alternative acquirer of LYP may be a Malaysian consortium.

(b) The height of barriers to entry to the market. This is a well-known economic concept. The term is referred to in the Explanatory Memorandum as ‘any feature of a market that places an efficient prospective entrant at a significant disadvantage compared with incumbent firms, including, for example, the presence of economies of scale or scope, control over essential inputs or government regulations which restrict entry into the market. The term encompasses barriers to exit, such as high ‘sunk’ costs. It represents the ease with which new firms can enter or leave the market now or in the future. It is a metaphor for economic disincentives for new entrants into an existing market.

(c) The level of concentration in the market. In assessing whether or not a proposed acquisition is likely to substantially lessen competition in the market, its existing concentration and any increase in that concentration, by reduction of the number of competitors or the accrual of significant additional market share to one of them, will be relevant.

(d) The degree of countervailing power in the market. As set out in the Explanatory Memorandum the concept of countervailing power ‘… refers to the extent to which market power held by the merged firm could be offset by market power held by customers or suppliers’. As appears later in these reasons that power may not be limited to direct economic power but may be exercised through price sensitive regulators or even political mechanisms, particularly where essential facilities are concerned.

(e) The likelihood that the acquisition would result in the acquirer being able to significantly and sustainably increase prices or profit margins. This factor, according to the Explanatory Memorandum, may be an indicator of the extent to which the merged firm would acquire market power sufficient to allow it to raise prices significantly above costs which would not be neutralised by the competitive responses of competitors, new entrants or imports. There is a slight dissonance between the explanation and the language which looks to the post-acquisition power of the acquirer rather than the merged firm. It may be that this factor is drawn upon the assumption of a complete acquisition of one firm by another resulting in an enhancement of market power for the acquiring firm. The competition argument advanced by ACCC rests upon the premise that LYP already has the requisite ability and that its incentive to exercise that power would be increased by the acquisition.

(f) The extent to which substitutes are available in the market or are likely to be available in the market. The Explanatory Memorandum, in respect of this matter, observes that the availability of substitute products in a market where a merger takes place allows consumers to purchase alternative products if the merged firm seeks to raise its price. There is no relevant substitute suggested, in the context of this case, for electricity in the wholesale market. To the extent that there is a product market for derivative contracts, the availability of different forms of hedge cover may be relevant.

(g) The dynamic characteristics of the market, including growth, innovation and product differentiation. This factor requires a consideration of change in the market place. The Explanatory Memorandum refers to markets being dynamic in the sense that demand for products may increase or decrease over time with changes in taste, quality and incomes. The language is not so limited. The dynamic characteristics of the Electricity Markets may extend to the evolving rules under which they operate and the increasing sophistication and experience of their participants.

(h) The likelihood that the acquisition would result in the removal from the market of a vigorous and effective competitor. As stated by the Explanatory Memorandum the removal of a vigorous and effective competitor, even one with a small market share, may have a significant impact on the level of competition in the market. This may not be of significance and concern where barriers to entry are low or where meaningful import competition exists.

(i) The nature and extent of vertical integration in the market. The Explanatory Memorandum states that vertical mergers can lessen competition where, prior to the merger, one of the firms has substantial market power at one level which can be exploited in the relevant upstream or downstream market as a result of the merger, eg by denying downstream competitors access to essential inputs. This matter is relevant to one of the two primary competition issues raised by the ACCC in this case, namely the apprehended increase in the extent of vertical integration following from the proposed acquisition.

The Onus of Proof

355               AGL seeks declarations in the present case that the relevant acquisitions would not have the effect or would not be likely to have the effect of substantially lessening competition in a market in contravention of s 50. Those being the declarations that AGL seeks it is necessary for it to establish, on the balance of probability, what it seeks to have declared. It will succeed in that task if it establishes, on the balance of probability, that each of the relevant acquisitions would not be likely to have the effect of substantially lessening competition in the relevant markets. That is to say, AGL must negative the existence of any real chance, in the sense discussed earlier, of a commercially relevant or meaningful lessening of competition flowing from the acquisition.

356               The question of onus gives rise to the interaction between the standard of persuasion which is on the balance of probabilities, and the facts required to be proved, which is the absence of a real chance of a substantial lessening of competition. The interaction is analogous to that which arises between standard of proof and the nature of the finding required in determining whether damages can be awarded for loss of a future commercial benefit or opportunity. Issues of that kind arose in Sellars v Adelaide Petroleum NL (1994) 179 CLR 332 and Norwest Refrigeration Services Pty Ltd v Bain Dawes (WA) Pty Ltd (1984) 157 CLR 149. The ACCC submits that these cases decided that the loss of a future commercial benefit or opportunity as a result of alleged wrongful conduct must be proved on the balance of probabilities. More specifically, a plaintiff must show that the hypothesis in favour of causation of such loss is more probable than any competing hypothesis denying causation. It submits that similar principles should be applied to the question raised in this proceeding while recognising the differences in its nature. The ACCC submits that AGL must satisfy the Court that its hypothesis against any likely substantial lessening of competition in any relevant market is more probable than the competing hypotheses which are advanced to suggest a real chance of competition being substantially lessened in any such market. I accept that formulation of the approach which should be taken in this case. I accept also the proposition advanced by the ACCC that AGL is not entitled to relief if:

(a) the Court is left in a position of uncertainty about the competing hypotheses; or

(b) the Court concludes that the hypotheses suggesting a real chance of competition being substantially lessened are more probable than the opposing hypotheses.

I should add that in so describing the ACCC’s submission, which I accept, I have substituted the term ‘real chance’ for ‘real possibility’ which in my opinion may set the bar too high.

The ACCC Case Summarised

357               The ACCC’s outline of ways in which the proposed acquisition would lessen competition may be reduced, by way of summary, to the following sequence of propositions:

1. Post-acquisition AGL, having acquired a natural hedge, will reduce its contract cover leaving LYP or other base load generators less hedged and with greater incentive to exercise their market power by raising spot prices in the wholesale market for electricity. These spot prices will occur, reflecting a substantial lessening of competition.

2. Post-acquisition AGL will contract more with LYP than with other generators because it gets an implied discount via its equity. Other generators and retailers will contract inter se rather than with AGL and LYP for confidentiality reasons. There will be increasing vertical mergers between generators and retailers thinning the contract market, raising barriers to entry and so substantially lessening competition.

3. The preceding effects will be exacerbated by the natural tendency of LYP’s management to consider AGL’s interests in the conduct of LYP.

4. The preceding effects will be further exacerbated if AGL exercises influence or control over the operations of LYP.

5. AGL, notwithstanding its contractual arrangements, will exercise control or influence over LYP.

6. The acquisition will increase the probability of a risk sharing agreement between AGL and LYP which will increase the incentive of LYP to exercise its market power.

Whether AGL Will Control or Influence LYP

358               It is convenient first to consider whether AGL will, as a result of the acquisition of its interest in LYP, control or influence the operation of that business in its own interests and in a way which gives rise to the likelihood of a substantial lessening in competition. AGL’s position, as stated in its closing submissions, is that there is no credible evidence, even at the level of a realistic possibility, that the proposed acquisition will result in it being in a position to control the activities of LYP. It points to the specific provisions in the transaction agreements for separation between AGL and LYP. These matters have already been referred to in the description of the transaction documents and the ACCC undertaking. AGL submits that in particular:

(a) AGL will have no control over the way in which LYP through MMCo bids electricity into the NEM or the derivative contract position of LYP with other parties; and

(b) the structure of ownership of LYP means that AGL will be constrained by the shareholdings of the other GEAC members. There is no incentive for those other members who control MMCo to engage in a policy which would favour AGL (a single customer of LYP) at the expense of its other customers as this would not maximise the profitability of LYP.

359               The ACCC on the other hand submits that the Court should disregard the various contractual provisions on which AGL seeks to rely. It is said to be clear as a matter of law that those provisions can be altered by agreement of the parties thereto and indeed can be altered by parties representing 90% of the voting rights in GEAC. They can be altered at any time. AGL, it is said, has adduced no evidence to suggest that they would not be altered in the future. The proffered undertaking has not been accepted by the ACCC and so is of no legal effect in the context of these proceedings. The ACCC said, in its closing written submission, that AGL has not offered a similar undertaking to the Court. However, in his oral closing submission, counsel for AGL did so.

360               Assuming the subsistence of the undertaking, the ACCC pointed to a number of respects in which it said AGL would be able to take ‘an active and influential role in the business affairs of Loy Yang’. These were:

(a) the formulation of a risk management policy defined in cl 1.1 of the Shareholders Agreement (to have the meaning given it by Schedule 3 to the Shareholders Agreement);

(b) the power of veto in respect of Board Special Matters (defined in cl 1.1 of the Shareholders Agreement);

(c) the requirement that the GEAC Board consider and, where appropriate, approve business plans and budgets for the business;

(d) the role played by the initial partners (and now GEAC which has effectively taken over their role) regarding operating budgets and regular trading reports by MMCo under the MMCo Agency Agreement.

The ACCC pointed out that AGL is entitled to participate in the preparation and approval of the risk management policy which will set limits and controls on the exposure of the business to specific categories of business risk, including trading risks. Those limits and controls may be quantitative or qualitative and could operate to limit or control the total level of risk which a specific business activity or division may incur. So trading risk, by reference to price and volume, may be subject to limits on the proportion of financial budget forecasts which can be put at risk as a consequence of the contract portfolio exceeding available plant capacity or a disproportionate exposure to spot prices. Credit risk would be subject to limits on the financial exposure to counter-parties of various categories of credit worthiness.

361               In respect of AGL’s effective power of veto, arising from the requirement of a 75% majority vote of directors in relation to Board Special Matters, the ACCC referred to the definition of those matters which includes:

1. changing the risk management policy;

2. adopting a budget approving the financial statements or any material deviation from a budget;

3. making a capital expenditure or other budgetary variation exceeding by $10 million per item, the amount proposed in a Board approved budget or business plan.

These, it was said, would confer upon AGL a significant practical role in determining and approving budgets and policies for LYP. Moreover, it was submitted, the GEAC Board will receive from the management of GEAC and/or MMCo an annual updated five year business plan budget and annual budgets, and will set the five year and annual business plans and budgets. AGL is able to actively participate in decision-making in respect of those matters affecting the operation of LYP, and has a veto right over their approval.

362               Finally it was said that AGL has practical influence over the operations of LYP through the MMCo Agency Deed. MMCo is required to give the partnership monthly reports within seven business days at the end of each month, quarterly reports within seven days at the end of each quarter and detailed annual reports within one month of each of the financial years, in each case setting out the details of the trading activity of those elements of the business administered by MMCo as are reasonably required by the partnership. MMCo is required to submit to the partnership, for its approval by 1 February each year, an operating budget for the dispatch and marketing activities of the partnership and any other costs and expenses it will incur in carrying out its obligations under the Deed and the MMCo Operational Deed with respect to the immediately following financial year. The final form of the operating budget is to be approved in writing by the partnership.

363               The ACCC contended that AGL will be the only shareholder in LYP that has experience and expertise in the Australian electricity industry. The CBA, it said, is purely a financial investor with no intention of remaining in the longer term. TEPCo is investing in order to gain further expertise about the industry. The remaining investors will be passive financial investors in the true sense of that expression.

364               Mr Fraser agreed, as the ACCC pointed out, that the risk management policy for LYP would involve such issues as counter-party risk, the percentage allocation between contracted and uncontracted generation and the level to which the generator had hedging contracts. Mr Fraser’s evidence that AGL would not know LYP’s actual contract position with any reasonable degree of accuracy on a daily, monthly or annual basis, was seen to be inconsistent with the evidence in which he recognised that AGL would be able to participate in the formulation of the risk management policies governing MMCo’s activities. He considered, however, that in respect of any given half-hour period, AGL would not be in a position to know what LYP’s uncontracted capacity was.

365               The ACCC referred to the evidence of Mr Hayes who agreed in evidence that the risk management policy to be approved by the GEAC Board could include global limits on the percentage LYP was to be contracted on a monthly basis. He did not know whether approvals for LYP budgets would be six-monthly or yearly and was not aware of a provision of the Shareholders Agreement requiring GEAC to provide shareholders monthly with unaudited management accounts of the group for that month. He agreed that a profit and loss account for GEAC would include a revenue line principally comprising revenue from LYP. This would comprise spot market revenue, contract revenue and coal sales.

366               Mr Kolhatkar of the CBA had said that as a director on the GEAC Board he would expect LYP revenue to be broken down at least into spot sales and total contracts. It would be important information for him as a GEAC director in assessing financial performance of LYP over the previous month to know in terms of GWs the precise breakdown between the contracted and uncontracted quantities of electricity dispatched by it. Mr Kudama had agreed that AGL would be involved in setting the risk management policy by GEAC in respect of the trading and bidding strategies of LYP. He agreed, upon examination of the risk management policy criteria annexed to the Shareholders Agreement, that the policy had to deal with trading risks in terms of price, volume and credit risk and could set limits on the financial exposure to counter-parties. He also agreed that LYP would derive income from spot market revenue and contract revenue but was not sure whether the GEAC Board would make a specific decision about the proportion of uncontracted generation dispatch at Loy Yang. He felt it was a good idea to set some approximate proportion.

367               On the basis of these considerations, the ACCC argued that it was clear that AGL would not be a passive investor in Loy Yang but would play an important if not pivotal role in the formulation of risk management policies and financial decisions.

368               AGL made the point that while it retains a Victorian electricity retail business it is limited by Victorian legislation to a minority shareholding in one generator. So much is provided by the Electricity Industry (Prohibited Interests) Regulations 2003 to which reference was made earlier. It also argued that any additional acquisition of shares in GEAC, MMCo or any other holding company or subsidiary of either would be susceptible to testing against s 50 at the time of such acquisition. I accept those propositions and do not consider that it is necessary, for the purpose of the present proceedings, to have regard to the possibility that AGL might increase its interest in LYP at some later time.

369               AGL then submitted that, absent any further acquisition of shares, the clear and compelling commercial motivations of all shareholders in GEAC would be to maximise the value of their shares in that company and of the underlying LYP assets. That is to say, LYP would act in its best interests in the putative future world ‘with’ the proposed acquisition as it would ‘without’ that transaction.

370               AGL submitted that despite the extensive and lengthy investigation by the ACCC of the proposed transaction no document nor any aspect of the extensive cross-examination of the witnesses could be pointed to which contradicted the compelling commercial motivation of the LYP shareholders. In any event it was submitted AGL directors would be obliged, even if they had access to LYP confidential information or were in a position to influence its contracting behaviour, only to use that information or to act in GEAC’s interests. Were GEAC’s interests and AGL’s to diverge they would be obliged to prefer GEAC’s. Counsel for AGL cited Whitehouse v Carlton Hotel Property Pty Ltd (1987) 162 CLR 285. In that case Mason, Deane and Dawson JJ at 290 referred to the distinction between the indirect proprietorship and ultimate control of the shareholders on the one hand and the powers of management of the directors on the other. In the factual context of that case, they observed that it was no part of the function of the directors as such to favour one shareholder or group of shareholders by the discriminatory exercise of their fiduciary power to allot shares.

371               In reply submissions filed after the hearing, the ACCC referred to cl 3.6(b) of the GEAC Shareholders Agreement which provides:

‘To the extent not precluded by law, a Director may make a decision in the interests of the Shareholder which appointed the Director, without being required to have regard to the interests of the other Shareholders individually or collectively.’

The ACCC also pointed out that GEAC is a proprietary company, not a listed public company. It was said that it is very rare for the directors of proprietary companies to come under any judicial or other external scrutiny in connection with the discharge of their fiduciary obligations other than in oppression proceedings, which are highly unlikely to occur in the affairs of GEAC. Reference was also made to Levin v Clark [1962] NSWR 686 at 700 where Jacobs J said that it was possible for a director who was ‘particularly appointed for the purpose of representing the interests of a third party’ to ‘lawfully act solely in the interests of that third party, where the breadth of the fiduciary duty has been narrowed, by agreement amongst the body of the shareholders’ – see also Re Broadcasting Station 2GB Pty Ltd [1964-5] NSWR 1648, Berlei Hestia (NZ) Ltd v Fernyhough [1980] 2 NZLR 150 and Japan Abrasive Materials Pty Ltd v Australian Fused Materials Pty Ltd (1998) 16 ACLC 1172. So it was said that resort to the ordinary rules concerning the fiduciary duties of directors does not assist AGL’s case.

372               I accept the ACCC’s submission in this respect that the AGL directors on GEAC would not be legally constrained, on account of their duties as directors, to disregard the interests of AGL in board decisions. In my opinion however, a powerful constraint against AGL directors acting in their own interests and contrary to those of other shareholders arises from the presence of the representatives of those other shareholders on the Board and the nature of those other shareholders. I refer to my earlier observations concerning the role likely to be taken by TEPCo BV in LYP. Nor is the CBA to be discounted on the basis that its investment is relatively short-term. A further constraint upon the hypothesised actions by AGL are the supervisory limits likely to be placed by LYP’s external financiers on its operations and, as I have found, likely to be specifically concerned with both hedging and risk management policies generally.

373               In accepting that AGL potentially has the role underlined in the ACCC’s submissions, I do not accept that the negative aspects of such power as it has through a veto will be capable of being effectively deployed in such a way as to advantage it in respect of the detailed decision-making that must be taken in the day-to-day bidding and pricing functions which will be in the hands of MMCo. Even absent the undertaking, I am inclined to regard the hypothesis that LYP somehow becomes hostage to AGL’s interests as unlikely. I am reinforced by the undertaking in the conclusion that it will not. The directors of AGL would have to be aware that any procuring by them of a breach of the undertaking by AGL would be capable of amounting to a contempt of Court on their part. In my opinion, such influence as AGL has in relation to the LYP operations will not extend to allow it to control or influence the detailed marketing decision-making of LYP nor provide it with access to confidential information about that detailed conduct or about its retail competitors.

374               A fortiori, I do not accept the so called ‘internalisation’ argument concerning the post-acquisition behaviour of LYP management indifference to AGL’s interest is more than speculative. It does not form a basis for hypothesising that LYP management would accommodate AGL’s perceived interests so as to effect some anti-competitive outcome beneficial to AGL. There was no evidence of the likelihood of such a response or of its consequence for competition in the wholesale electricity and derivatives market.

375               The ACCC advanced other somewhat speculative submissions about the post-acquisition conduct of AGL and LYP. One suggestion was that the acquisition would increase the likelihood of AGL entering into a risk strategy arrangement with LYP to provide an incentive for LYP to reduce its contract cover. This, it was proposed, would again have the effect of increasing the incentive for LYP to raise the spot price of electricity. But as AGL submitted, the acquisition is not a prerequisite for AGL to enter into such an agreement with LYP. It already has in place the DPHA. There is nothing about the acquisition in my opinion to make such an arrangement or consequence more likely.

376               The ACCC also submitted that the acquisition would increase the likelihood that AGL would want to coordinate the bidding and dispatch of the Somerton plant at price levels above the LY power station to enable LYP to increase spot prices for its uncontracted capacity. There was no evidence to support this speculation and it was rejected by Mr Fraser. Most of the time the shortrun costs of Somerton are so much greater than the spot price that it would not be able to meet capacity not met by LYP. For the short periods in which Somerton is running it is able to meet AGL’s exposure to spot price. Somerton is therefore unlikely to be able to take advantage of high spot prices over the market hedge it provided to AGL. In my opinion there is no ‘real chance’ of any substantially lessening of competition based upon the hypothesis about the uses to which the Somerton power station would be put post-acquisition.

Market Definition

377               Definition of the relevant markets for the purposes of s 50 is logically anterior to the analysis of competition in those markets with and without the acquisition. That does not require a bright line separation, in consideration of the evidence, between such definition and the consideration of issues of structure and power distribution. As I observed in Singapore Airlines Ltd v Taprobane Tours WA Pty Ltd (1991) 33 FCR 158 at 178:

‘It is a focusing process and the court must select what emerges as the clearest picture of the relevant competitive process in the light of commercial reality and the purposes of the law. There is a feedback between any proposed market and the structure and power distribution which that proposal throws up.’

This simply reflects the observation of Mason CJ and Wilson J in Queensland Wire Industries Pty Ltd v The Broken Hill Pty Co Ltd (1988-89) 167 CLR 177 at 187 that:

‘Defining the market and evaluating the degree of power in that market are part of the same process, and it is for the sake of simplicity of analysis that the two are separated.’

Those observations were made in the context of s 46 but as Spender J said in QIW Retailers Ltd v Davids Holdings Pty Ltd (No 3) (1993) 42 FCR 255 at 258, similar considerations arose under s 50, as it stood prior to the 1992 amendments, in determining the potential for dominance of the market.

378               The concept of market describes, in a metaphorical way, an area or space of economic activity whose dimensions are function, product and geography. A market may be defined functionally by reference to wholesale or retail activities or a combination of both. The concept of product encompasses goods and services and, having regard to the definition of ‘market’ in s 4E, includes the range of goods or services which are substitutable for or competitive with each other.

379               The process of market definition was expounded in QCMA where the Tribunal defined ‘market’ as the area of close competition between firms and observed that substitution occurs within a market between one product and another, and between one source of supply and another in response to changing prices:

‘So a market is the field of actual and potential transactions between buyers and sellers amongst whom there can be strong substitution, at least in the long run, if given a sufficient price incentive.’ (at 190)

In Re Tooth Co Ltd and Tooheys Ltd (1979) 39 FLR 1, the Tribunal identified the task of market analysis as involving:

(1) Identification of the relevant area or areas of close competition.

(2) Application of the principle that competition may proceed through substitution of supply source as well as product;

(3) Delineation of a market which comprehends the maximum range of business activities and the widest geographic area within which, if given a sufficient economic incentive, buyers can switch to a substantial extent from one source of supply to another and sellers can switch to a substantial extent from one production plan to another.

(4) Consideration of longrun substitution possibilities rather than shortrun and transitory situations recognising that the market is the field of actual or potential rivalry between firms.

(5) Selection of market boundaries as a matter of degree by identification of such a break in substitution possibilities that firms within the boundary would collectively possess substantial market power so that if operating as a cartel they could raise prices or offer lesser terms without being substantially undermined by the incursions of rivals.

(6) Acceptance of the proposition that the field of substitution is not necessarily homogeneous but may contain sub-markets in which competition is especially close or especially immediate. This is subject to the qualification that competitive relationships in key sub-markets may have a wide effect upon the functioning of the market as a whole.

(7) Identification of the market as multi-dimensional involving product, functional level, space and time.

380               In the present case, the parties have a common position in relation to the definition of the relevant retail markets. There are separate retail markets for:

(a) the supply of electricity to residential and small business customers in Victoria; and

(b) the supply of electricity to industrial and commercial customers in Victoria.

The parties are divided however, as to the definition of the market related to the production and sale of electricity by generators. This is the market characterised by transactions between suppliers and retailers. The ACCC contends for separate markets in respect of the sale of electricity in the spot market conducted by NEMMCO which it calls the ‘wholesale electricity market’ and the forward contract market which it calls the ‘electricity derivatives contract market’. In each case the ACCC says that these markets are regional and that the relevant market in each case is confined to Victoria.

381               There is a degree of artificiality in my opinion in defining the relevant markets as though electricity were a product sold by wholesalers to retailers and onsold by retailers to end users. The ACCC argued, in its closing written submissions in reply, that retailers do acquire ownership of electricity purchased in the spot market and on-supply that electricity to their customers. The submission relied upon cl 2.3.1 of the NEC which refers to customers as persons registered with NEMMCO who engage in ‘… the activity of purchasing electricity supplied through a transmission system or a distribution system to a connection point’. The ACCC also referred to s 16 of the Electricity Industry Act 2000 (Vic) which provides, inter alia, that ‘[a] person must not engage in the generation of electricity for supply or sale or the transmission, distribution, supply or sale of electricity unless that person is the holder of a licence.’ So it was said under the Code only a market customer may purchase electricity from the national electricity pool and under the ElectricityIndustry Act only a licensed retailer can sell electricity to consumers.

382               The words ‘purchase’ and ‘sell’ in the Code and the Act will no doubt take their meaning from the kinds of transactions which are commonly so described in the operation of the NEM. They do not mean that in any intelligible sense the retailer acquires ownership of electricity as one would own a piece of personal property. As I commented earlier in these reasons, electricity delivered into the common transmission system has the character of a fluid joining at common stream. Once within the transmission system it cannot be subdivided by reference to its origins. It flows through transmission and distribution networks to the end user. The generators and retailers operate in a kind of ‘virtual reality’ of sale and purchase whose rules are defined by the bidding, spot pricing and dispatching mechanisms and the derivative contract arrangements which are an essential aspect of the relationship between the participants. There is also a degree of unreality involved in separating out and identifying separate markets for the sale of electricity and the provision of derivative contracts. Although there are some loose, but not entirely appropriate, analogies between the derivative contract and a form of insurance in my opinion, for present purposes, the derivative contracts ought to be regarded as an integral part of the pricing and payment arrangements between generators and retailers in relation to the underlying product, which is electrical energy, and which they deal with ‘as if’ it had been sold from supplier to retailer. In so saying, I have regard to the fact that the great bulk of derivative contracts are entered into on an over-the-counter basis. In the end, even if one were to define the separate markets for which the ACCC contends it would not seem materially to affect the outcome of this case having regard to the way in which the ACCC contends that the proposed acquisition would be likely to have the effect of substantially lessening competition in the market.

383               As to the geographic bounds of the market, there is no doubt that from time to time, according to interconnector constraints or supply/demand imbalances within a particular region, there may be price separation between regions. The ACCC, in respect of its separately defined wholesale electricity market, relies upon the incidence of price separation and references in the internal documents of AGL and other participants in the electricity industry to different regional markets with different commercial characteristics.

384               With respect to the electricity derivatives contract market asserted by the ACCC, it was contended that the evidence establishes that both generators and retailers wish to hedge their supply and acquisition of electricity at the Regional Reference Node where that supply or acquisition occurs. So a generator located in Victoria will wish to offer a contract referenced to the spot price at the Victorian Regional Reference Node. Similarly, a retailer acquiring electricity in Victoria would wish to acquire a contract referenced to the Victorian Regional Reference Node. Both generators and retailers, it was submitted, are far less willing to offer contracts referenced to a Regional Reference Node at which they are not buying or acquiring electricity. This is because of the basis risk. The inter-regional settlement auctions established by NEMMCO to mitigate those risks were said to be recognised as not entirely effective because they are unable to fully hedge the risk. For that reason, as the ACCC pointed out, generators will also seek to offset inter-regional risk with a hedge contract with a generator in the other region which has the effect that the hedge contract is converted from an inter-regional contract to an intra-regional contract. In this respect they relied upon the evidence of Mr Nethercote.

385               On the other hand, AGL pointed to evidence that generators can and do offer hedge products denominated against nodes other than their home nodes. Mr Flukes gave evidence that Southern Hydro trades hedges against all nodes except the Snowy node and that up to half its contract turnover is in regions in which it does not have physical generation. Similarly, although NRG Flinders plant is wholly located in South Australia, Mr Williamson gave evidence that NRG had entered into a substantial number of such swaps. Mr Willis foreshadowed that Hydro Tasmania would be likely to sell hedge products referenced against the Victorian Node even though it has no Victorian plant. Mr Nethercote said that LYPM’s Board has permitted the business to enter into inter-regional contracts of up to approximately one quarter of its total sent out electricity generation capacity and could presumably chose to increase the limits, and that the Trading Team does in fact chose to undertake such contract sales. Mr Thompson provided an example in the summer of 2001 when LYPM traded substantial hedge contract coverage in New South Wales because it expected that the New South Wales and Victorian nodes would be generally aligned.

386               As Dr Price demonstrated, the incidence of interconnecting constraints between the regions is low. In relation to interconnections between Victoria, South Australia and Snowy in 2002, the incidence of constraint into Victoria was about 0.5%. And when there were constraints the value of divergence in regional prices was generally less than $10. Referring to basis risk, AGL pointed to Mr Denton’s statement that:

‘By purchasing an exposure to this value, participants can effectively transfer Spot Market Price exposure from one regional price to another. In my experience the Settlements Residue Auction is now a well understood product, and is used to varying degrees by most participants.’

AGL also referred to the evidence of Mr Ergas who undertook tests based on pool price data which suggested that generators in different regions were being substituted with each other to meet demand in all regions. These were the so-called variance decomposition same-market test, the Granger causality same market test and a correlation test which tended to confirm those results. He concluded, on the basis of his analysis, that there was a market for wholesale supply and demand for electricity that encompassed the NEM regions. Alternatively, there was a market for the wholesale supply and demand for electricity that encompassed Victoria, South Australia and New South Wales.

387               The geographic market is not to be determined by a view frozen in time or by observations based on shortrun time scales. The NEM is an evolving market which is intended and designed to operate as a single market for electricity throughout the regions which it covers. Transient price separations between those regions may define temporally limited sub-markets which can be referred to for the purposes of competition analysis. And they may well attract the appellation ‘market’ in the ordinary parlance of suppliers and retailers operating within them. In my opinion, however, having regard to the structure of the market and the extent to which its major participants operate across regional boundaries, I am satisfied that there is one NEM-wide geographic market for the supply of electricity, and associated with that, entry into electricity derivative contracts.

Matters to be Considered Under Section 50(3)

388               AGL and the ACCC both made submissions on the market characteristics which must be considered under s 50(3).

389               In respect of import competition s 50(3)(a) addresses imports from outside Australia. As was observed earlier in these reasons that appears to have little or no application in the present case other than the possibility of international corporations entering or acquiring interests in the wholesale market, as exemplified in the present case. If s 50(3)(a) has a wider application there was in any event no question of imports of electricity in to the NEM-wide wholesale market. Importation was discussed solely in the context of the transmission of electricity between regions within the NEM.

390               The next matter under s 50(3)(b) concerns barriers to entry. It was submitted for the ACCC that the construction of major new generating plant involves high capital costs and a significant lead-time. Capital costs are sunk costs as the plant and its component parts cannot be redeployed to any use other than generating electricity from the fuel source available at the relevant site. This, as the ACCC said, is particularly true of the Victorian brown-coal burning base load plants like Loy Yang A. Entry barriers to the market however are not to be judged by reference to base load brown-coal burning generators alone although there was evidence from Dr Price that Victorian brown-coal plant has the lowest fuel cost of any of the new generation options in different regions of the NEM.

391               AGL referred to a number of matters to demonstrate that entry barriers are low. The matters to which it pointed were:

(a) evidence from Dr Price that gas turbines are able to be commissioned in under two years;

(b) the fact that since the start of the NEM 552MW of gas-fired generation operating as peaking plant has been introduced into Victoria;

(c) Dr Price’s evidence of the commissioning of 4,780MW of capacity into the NEM since its commencement;

(d) Dr Price’s evidence of 5,000MW of generation capacity currently proposed for development in the NEM;

(e) the existence of new projects indicating a range of parties are considering investment options in various regions of the NEM;

(f) the significant increase of new entrants into the market by generators and potential generators in 2001 following the upward movement in spot prices. This is discussed subsequently in consideration of the market response to spot price increases in the summer of 2000/2001.

(g) the existence of a very significant amount of commercial information available to generators which allows them to make informed decisions about how and when to enter the wholesale electricity market;

(h) Mr Fraser’s evidence that:

. AGL has been able to commission generating plants being the Somerton plant and the Hallett plant;

. if AGL or any other market participant were having difficulties in obtaining hedge contracts this would ultimately trigger participants adding new capacity to the grid;

(i) the evidence of Mr Lee of Energex that a retailer would consider investing in peaking generation as an alternative to investing in hedge products and that the demand for hedge products is related to the cost of investing in peaking plants.

In my opinion, having regard to the above matters and the response of potential new entrants to price signals in the summer of 2000/2001, it cannot be said that barriers to entry into the NEM-wide wholesale market are such as significantly support or contribute to market power on the part of any of the market participants.

392               The next matter, which is identified by s 50(3)(c), relates to the level of concentration in the wholesale market. On the basis that the relevant geographic wholesale market covers the entire NEM, Loy Yang A has 5.3% of total NEM capacity. That is 2,000MW out of a total of 37,769MW. Moreover, it has less than half the generating capacity of the largest generator, which is Macquarie Generation, which has 12.4% of NEM capacity. In Victoria, Loy Yang A has a share of generation capacity which is 18.8% which, as AGL points out, is less than the capacity which can come into Victoria from other regions in the NEM at any given time. Although Loy Yang is the third biggest generator in the NEM, it generates 8.6% of the electricity.

393               The ACCC relied upon evidence that the three base load generators, Loy Yang A, Hazelwood and Yallourn accounted for 75% of actual generation in Victoria in 2002 and that Loy Yang A accounted for 30%. On that basis it was submitted that there is a higher degree of concentration in the Victorian market. AGL responded that the figure is not meaningful because it does not have regard to the electricity imported into and exported from Victoria across interconnectors nor the fact that there is a net export from Victoria of electricity to other regions.

394               Having regard to my view of the definition of the wholesale market there is not, in the NEM-wide wholesale market as earlier defined, a degree of concentration of suppliers that could be said to support the subsistence of market power on the part of Loy Yang A or any of the major generators.

395               Section 50(3)(d) requires consideration of the degree of countervailing power in the market. The ACCC argued that the absence of countervailing power of consumers and retailers was reflected in the inelasticity of the demand for electricity. It was accepted by all witnesses that demand is almost completely price inelastic. The consequence of this is that a generator such Loy Yang has the opportunity during periods in which supply and demand balance is reasonably tight to increase prices substantially with a potential increase of up to 300 times average spot prices.

396               It is probably correct to say that in an economic sense there is not a high degree of countervailing consumer power which can be brought to bear upon the pricing practices of generators. There is however some scope for demand side management and perhaps of particular significance in the NEM is the reality that high electricity prices can very quickly become a political or regulatory issue. While countervailing power may not be exercised economically it can be exercised politically or by the regulator as proxy for consumers. LYP’s own sensitivity, in devising its Summer Strategy for 2001 to regulatory and political responses is an indicator of the reality of those considerations. The Security Trust Deed held by LYP’s current bankers and referred to earlier requires risk management practices that have regard to the possibility of regulatory intervention. The statutory frameworks within which the various regions of the market operate and the bidding rules themselves provide mechanisms for the exercise of that countervailing power.

397               Section 50(3)(e) requires consideration of the likelihood that the acquisition would result in the acquirer being able to significantly and sustainably increase prices or profit margins. It is not clear how this factor has application in the present case. The acquirer is AGL. The focus of the market power debate has been on LYP’s ability to increase prices and profit levels. Moreover, it does not appear to be the ACCC’s case that LYP would increase its market power as a result of its acquisition by AGL and the other consortium members. Rather it is that LYP would have a greater incentive to use the market power that it presently has. That incentive, on the ACCC case, will arise by virtue of AGL reducing its contract cover on account of the natural hedge conferred on it through its equity in LYP. For reasons, which are set out later, I do not accept that scenario as a likely one.

398               The matter to be considered under s 50(3)(f) is the extent to which substitutes are available in the market or are likely to be available in the market. In the present proceedings there are no substitutes in the wholesale market for electricity. There is however considerable flexibility in the hedging arrangements into which parties may enter, albeit the most popular form of hedge is the swap contract. Consideration of that flexibility has some analogy to substitution analysis. A party unable to obtain one form of hedge contract may be able to obtain protection in another form. Overall, however, substitution analysis did not play a significant part in the case.

399               In relation to the dynamic characteristics of the market referred to in s 50(3)(g), it is plain that the wholesale market for electricity and derivative contracts is dynamic. The rules that regulate bidding, dispatch and pricing change as the market evolves. It is a characteristic of electricity markets, identified in the writings of the ACCC expert Professor Wolak, that as flaws in market design emerge the rules can be changed. The introduction of a new rule requiring good faith in the lodgment of bids and rebids, referred to later in these reasons, exemplifies this evolutionary character. The increasing sophistication of market participants is another relevant factor. By way of example there was evidence that market participants negotiating forward contract strike prices will not readily be misled by transient price spikes designed only to endeavour to increase the forward contract price. They are also showing flexibility in the way in which they deal with spot price volatility, not only by the use of hedge contracts but also by the acquisition or construction of peaking generators.

400               As to the question, raised by s 50(3)(h), whether the acquisition would result in the removal from the market of a vigorous and effective competitor, there was no evidence to suggest that that would be an outcome of this acquisition.

401               The nature and extent of vertical integration falls for consideration under s 50(3)(i). There is already some degree of vertical integration, either virtual or actual, in the wholesale market. Origin Energy and Energex, both substantial retailers, have integrated generators and retail businesses. Yallourn, a Victorian generator, has a retail business conducted through AusPower. AGL already has control of two peaking plants, one at Somerton and one at Hallett. Stanwell Corporation is a Queensland generator with a retail licence in Queensland. Delta Electricity, a New South Wales generator, has a retail licence in New South Wales and Tarong Energy is a Queensland generator with retail licences in Queensland and South Australia. In addition, TXU, which is a substantial retailer, has a ‘virtual’ vertical integration with the supplier Ecogen by way of a dispatch hedge arrangement. NRG Flinders, a South Australian generator, has a retail licence in South Australia. The question whether vertical integration will increase as a result of the acquisition is considered in other parts of these reasons.

402               None of the characteristics of the market, identified by reference to s 50(3) of the Act, in my opinion supports the conclusion that LYP has and exercises substantial market power or that the proposed acquisition is likely to cause a substantial lessening of competition in the wholesale market. Those matters of course are not exhaustive of the matters that have to be considered. The evidence required closer consideration of the question whether LYP has market power and would, as a result of the acquisition, acquire an incentive to use it, which it would not otherwise have.

Projections for Supply and Demand in the NEM

403               It is alleged by AGL in its statement of claim that electricity generation capacity available to meet the demand for electricity in Victoria substantially exceeds the level of average and peak demand for electricity in that State. AGL contends that, generator capacity overall, when assessed by reference to likely expansions of generator capacity within Victoria and of interconnection capacity into Victoria, is likely to continue to exceed the level of peak and average demand. This is contested by the ACCC, which says that the reserve margin in Victoria is less than the minimum reserve level for that State. The reserve margin is the amount by which generation capacity in a region, including capacity supplied by interconnectors, exceeds peak demand in the region measured at 10% POE, as calculated by NEMMCO. The POE is the probability of exceedence. A 10% POE refers to a level of demand that would only be expected to occur once every ten years. The minimum reserve level is the capacity of the single largest generating unit in the region. The ACCC relies upon the NEMMCO Statement of Opportunities 2003 Ch 6 at p 8. It denies AGL’s contention in par 75 of its statement of claim that the capacity of generators, other than Loy Yang, taken along with interconnection capacity, is sufficient to meet Victorian peak demand and NEM peak demand to such an extent that Loy Yang A is not necessary to meet those demands.

404               It is useful to begin consideration of this issue by reference to the NEMMCO Statement of Opportunities. Each year, before 31 July, NEMMCO publishes a Statement of Opportunities based on information available as at 31 May 2003 in order to provide technical and marketing data and information regarding opportunities in the NEM. The stated object of the Statement is to provide information to assist market participants and others assess the future need for electricity generating capacity, demand management capacity or augmentation of the power system to support the operation of the NEM. The NEMMCO Statement of Opportunities for 2003 is the fifth statement that it has published since the commencement of the NEM in December 1998.

405               In the Executive Summary to its Statement of Opportunities 2003 NEMMCO says that the generation capacity installed across the NEM is greater than the sum of all peak regional demands for the next five years. It then goes on:

‘However taking into account the requirement to maintain generation reserves, Victoria and South Australia do not have adequate reserves from (sic) the coming summer and the summer peak of 2005/06 is the first projected instance of a NEM-wide reserve deficit.’

406               And further:

‘Another important observation is that the reserve deficits in 2005/06 and onwards are not affected by interconnector capacity. Assuming the present levels of generation, additional interconnector capacity anywhere in the NEM will not act to improve reserves for this period. For example, in the analysis for the 2005/06 summer, there are no interconnectors operating at their technical limits.’

The forecasts on the demand side assume extreme levels of electricity demand. That is to say they assume demand conditions that would only be expected to occur once every ten years, the 10% POE mentioned earlier. Extreme demands are driven by severe climatic conditions. They would only be realised in the event of sustained temperatures in excess of 35 degrees Celsius. The Summary continues:

‘Importantly, the supply/demand balance also assumes that the extreme demand conditions occur in unison across all of the NEM States. This is a conservative assessment in that it implies that Queensland, New South Wales, Victoria and South Australia would be concurrently exposed to similar extreme weather patterns.’

The Executive Summary observes that NEMMCO’s assessment of capacity is based only on existing generation and transmission and ‘… committed new projects’. NEMMCO is nevertheless aware of numerous projects which have not achieved committed status and whose commitment would add capacity to the NEM and so increase reserves. The reserve margin calculated by NEMMCO also allows for an outage of a large generating unit in each region at the same time.

407               In its supply/demand balance information for Victoria and South Australia NEMMCO presents a combined assessment. The supply/demand balance for these two regions is dependent on the network capability into Victoria from Snowy and they are subject to similar weather patterns. Generally they experience peak demands at the same time. NEMMCO’s analysis of supply and demand in Victoria and South Australia indicates that although reserves for winter will be adequate until 2011, summer reserves fall below the minimum requirement in the coming summer. The forecast reserve deficit is 69MW for summer 2003/04. In summer 2008/09 demand will exceed supply side capacity on this projection.

408               The supply/demand balance projection is set out in more detail in Ch 6 of the Statement of Opportunities. The part of the NEMMCO statement relied upon by the ACCC in its Defence appears at p 6-8. At par 6.5.1 NEMMCO sets out the assumptions upon which the base-case for the calculation of the Victorian reserve margin rest. They are as follows:

. Reserve trigger level of 530MW for the first year then increasing to 560 MW for the rest of the outlook period;

. Hume generation (58MW) is dispatched fully to the Victorian region;

. Demand Side Participation (DSP) at time of peak loads is approximately 182MW – this is a reference to load shedding by end users eg by shutting down for a time;

. No diversity in demand between Victoria and other NEM regions;

. Sharing of generation reserves is optimised across the NEM;

. Commissioning of Basslink late 2005 provides an additional 590MW of capacity to the mainland regions of the NEM.

Under these assumptions the summer reserves are said to be below the minimum reserve level for the medium economic growth scenario for the summer of 2003/04 by 51MW. A shortfall of supply first occurs in summer 2007/08. On a low economic growth scenario the reserves fall below the minimum reserve level in summer 2004/05 and a shortfall of supply first occurs in summer 2008/09. For a high economic growth scenario the reserves fall below the minimum reserve levels for all periods and a shortfall of supply first occurs in 2006/07. The commissioning of Basslink by the summer of 2005/06 adds 590MW shared across all regions resulting in a deficit in the reserves of approximately 66MW for Victoria in the base case (medium economic growth, 10% POE demand) scenario.

409               The combined outlook for Victoria and South Australia is stated in par 6.7 and is substantially as already set out in the Executive Summary.

410               The NEMMCO Statement of Opportunities included a post-publication Addendum to take account of a proposed upgrade of the Loy Yang power station. After the Statement had been finalised for publication Loy Yang Power advised NEMMCO that an upgrade, scheduled to be completed before the winter of 2004, was to be brought forward. The upgrade is expected to increase the capacity of Loy Yang to 2110MW in winter and 2020MW in summer under extreme weather conditions with a maximum capacity of 2030 MW. This would provide an extra 70MW capacity in Victoria for the summer of 2003/04 and up to 10MW extra capacity in subsequent summers. The Addendum observes that the upgrade will not change projected winter capacities. It will impact on the supply/demand balance which is no longer projected to have a reserve deficit in Victoria and South Australia for the summer of 2003/04.

411               A witness called by the ACCC who gave evidence on this question was Greg Denton. Mr Denton is a Director of Tavis Pty Ltd trading as Tavis Consulting which provides market design and regulatory advice to stakeholders in the electricity and gas industries. He has had approximately seven years experience working in the electricity industry primarily in consulting roles but also in a commercial role with Enron Australia.

412               In his affidavit, as part of a market power analysis, Mr Denton considered critical factors affecting the ability of Loy Yang to act unilaterally in the future with respect to pricing. The factors he identified were:

(a) the mix and ownership of generation;

(b) the capacity of interconnectors allowing competition from inter-state generation;

(c) contracting capacity; and

(d) demand growth and demand-side management.

Mr Denton referred to the NEMMCO Statement of Opportunities for information about each of these factors other than contracting capacity. He compared the demand and capacity situation during the 2000/01 summer when, as appears later in these reasons, Loy Yang was able to affect spot market prices, with the situation for the 2003/04 summer. He claimed to be able to get an indication of whether or not Loy Yang would have an ongoing ability to influence spot market prices. He constructed a table showing a summary of available capacity and actual demand in each of the last four years. He said it shows that the situation for the 2003/04 summer closely reflects the 2000/01 summer when Loy Yang was able to affect the price.

413               The table which he constructed was as follows:

	2000/01 Summer	2001/02 Summer	2002/03 Summer	2003/04 Summer
	Actual	Actual	Actual	Forecast
Available capacity	9,603MW	9,576MW	10,112MW	10,326MW
Demand	8,088MW	7602MW	8,202MW	8,758MW
Surplus capacity	1,515MW	1,974MW	1,910MW	1,568MW

For 2000/01-2002/03 actual peak demands are displayed. For 2003/04 the medium growth forecast from the NEMMCO Statement of Opportunities for 2003 has been provided. Mr Denton said the situation could be quite different in the event that a hot summer eventuates resulting in the 10% POE. In that connection he referred to NEMMCO’s median demand forecast for Victoria as follows:

Victoria summer maximum demand forecast (Probability of exceedence)
90%	50%	10%
8,351MW	8,758MW	9,417MW

He observed that if a hot summer eventuates giving rise to a 10% POE the surplus capacity in Victoria will be just 909MW. In such a scenario, according to Mr Denton, price spikes are highly likely.

414               Mr Denton was cross-examined about his reliance upon the NEMMCO forecast. It was put to him that the minimum reserve level referred to in the NEMMCO Statement was designed to ensure enough power in any region to ‘keep the lights on’ in the event of the hottest day in ten years and the loss of the largest generator in the region. He accepted that as an interpretation of the term minimum reserve level. He said:

‘The idea behind the reserve margin is to give an indication that – and for the reliability panel to satisfy itself that the lights won’t go out in those sorts of events that you describe, yes.’ (T 726)

 

He agreed that when NEMMCO projected future supply and demand it considered demand in extreme conditions that might be expected to occur once every ten years. The kind of case in which that demand could occur would require sustained temperatures of about 35^o Celsius over all the NEM States. This had never occurred. He agreed that the NEMMCO projections were conservative.

415               It is also the case, and Mr Denton accepted, that in projecting future supply NEMMCO only relied upon what it described as ‘committed projects’. The term ‘committed projects’ is defined in par 1.6.2 of the NEMMCO Statement by reference to the following criteria:

1. The proponent has acquired or commenced legal proceedings to acquire land for the construction of the project.

2. Contracts for the supply and construction of the project’s major plant or equipment (including generators, turbines, boilers, transmission towers, conductors) including provisions for project cancellation payments have been executed.

3. The proponent has obtained all required planning and construction approvals and licences, including completed and approved environmental impact statements. These include planning and environmental approvals from duly authorised planning bodies at both State and Federal government levels.

4. Financing arrangements for the proposal, including debt plans, have been finalised and contracts executed.

5. Construction has commenced or a firm date has been set for construction to commence.

The definition goes on to say that for any generation or transmission project to be included in the NEMMCO forecasts, NEMMCO must have formed the view that there is a high probability of the project proceeding as well as satisfying these enumerated criteria.

416               Graphs produced by AGL and tendered in evidence during Mr Denton’s cross-examination showed actual Victorian winter and summer peaking demand for the years 1988/99 to 2002/2003 compared with the NEMMCO forecasts of peaking demand shown in its Statement of Opportunities for each of those years. The graph demonstrated actual outcomes significantly below forecast at both the 10% POE and the 50% POE for the years 2000/2001 and 2001/2002. (X 16)

417               Mr Nethercote was asked in cross-examination about the NEMMCO projection of a supply and demand imbalance for Victoria and South Australia in 2004/2005. He said:

‘That’s been seriously questioned and the subject of debate outside of NEMMCO based on the way in which they arrived at those numbers. The industry as a whole has questioned that.’

He was asked whether he accepted that if NEMMCO’s forecasts were accurate it would be as easy today and over the next two or three years to engage in the same sorts of pricing strategies that Loy Yang had engaged in in the past. He said:

‘I wouldn’t say it’s going to be as easy or follow the same pattern because I don’t think that’s true, and the NEMMCO SOO or Statement of Opportunities is certainly under a large degree of question from all participants in the industry at the moment. The publication of that document was very – not very well received.’ (T 417)

 

418               Mr Thompson of LYPM, was asked whether he rejected the underlying theme in the NEMMCO Statement of Opportunities for 2003 that there would be a reserve shortfall for Victoria and South Australia in 2004/2005. He said:

‘I would suggest that they’re overly conservative. Last year we were planning an outage in February and they were suggesting maximum demands of Victoria of over 9,000 Megawatts. We never got it within 1,000 of that.’

He agreed later in cross-examination that he had not read the whole of the Statement of Opportunities. He was aware in a general sense of what the documents said about supply/demand imbalances. It was put to him that future supply/demand imbalances would be an essential piece of information in connection with the implementation of any contracting strategy for the future. He replied by referring to new generation capacity in Queensland comprising 1,200MW of new supply and an interconnector upgrade which had been rated by NEMMCO at 1,500MW and now had a capacity of 1,900MW. It was pointed out to him that, in NEMMCO’s Statement, reserve deficits in 2005/06 and onwards were not affected by interconnector capacity. Mr Thompson accepted on that basis that the supply/demand shortfall forecast by NEMMCO for 2004/05 was not affected by interconnector capacity. (T568-569)

419               Dr Price described NEMMCO’s construction of its reserves as involving ‘an extraordinarily conservative approach. He said:

‘So what they assume is that each of the state systems achieve their maximum peak demand under the hotter summer conditions, under the highest economic growth scenarios modelled all at the same time. That has never happened, and then they establish reserves to meet that contingency, such that they assume that that plant will go out in every system at the same time under those extreme loading conditions, but that sort of undermines the whole value of what an interconnected system is about.’

420               In its 2000 Statement of Opportunities NEMMCO said it was not aware of any publicly announced proposals for new scheduled generators in Victoria. However its Statement of Opportunities for the following year showed a very substantial increase in committed new plant. It comprised a total additional capacity of 731MW to 781MW. In addition to that new capacity there was an announcement by AES Transpower Pty Ltd on 30 March 2001 of its intention to develop a 500MW peak load power station in Victoria. This was known as the Golden Plains Project. Mr Thompson observed:

‘Following these announcements the rise in forward contract prices stopped as shown in the following graph produced by LYPM which shows the 4pm daily bid/ask spreads from brokers’ prices listed on Reuters for annual contracts. The graph also shows that hedge prices for 2002 and beyond began to fall at this time and that the hedge price for each subsequent year was lower than its predecessor.’ (CB 4037)

 

It was also pointed out in AGL’s closing submissions that at p 2 of the NEMMCO Statement of Opportunities for 2003 the following statement appeared:

‘NEMMCO has a program commencing in July 2003 that will consider some aspects of reliability policy. This program may result in recommendations to the NECA Reliability Panel that, if accepted, would reduce reserve requirements initially in Victoria and South Australia.’

421               The extent of interconnector capacity and, associated with that the incidence of inter-regional constraints, are relevant to supply/demand projections within regions. Historically there has been a very small number of constraints affecting the flow of electricity into Victoria. Constraints on the Victoria/Snowy interconnection in 2002 occurred for only about 0.5% of the year. At times when the interconnector was constrained the price difference between those regions was usually less than $25 per MWh and mostly less than $10 per MWh. As between Victoria and South Australia in that year the interconnector was constrained into Victoria for about 6.6% of the year. At those times the price difference between the regions was generally less than $20 per MWh and predominantly less than $5 per MWh. Dr Price observed that the limited constraints and price differentials into Victoria imply a relatively low risk for generators in other NEM regions selling inter-regional contracts by reference to the Victorian price. I accept that observation.

422               There has been a number of new transmission investments which have been made since the NEM commenced operation. There is also a number of proposals for new interconnectors or the upgrade of existing interconnectors. One of the proposed interconnectors is the Murraylink interconnector between Victoria and South Australia which has been announced and is expected to be put in place in two to three years time. It will provide additional export capacity of 100MWs and an additional import capacity of 150 MWs. Dr Price expects operation of Murraylink as a regulated interconnector to reduce the number of constraints between Victoria and South Australia by making all capacity available at all times. In my opinion that prognosis is inherently reasonable and I accept it. The construction of the SNI interconnector from New South Wales to South Australia in 2004 will further reduce the incidents of inter-regional constraints. The amount of additional export and import capacity arising from the construction of that interconnector is 250MW. The incorporation of Tasmania into the NEM through Basslink is expected by its proponents to result in Tasmania exporting to Victoria at peak times and importing to Tasmania at off peak times, taking advantage of Victoria’s relatively cheap base load capacity. This is because the total output of hydro power stations across the year is affected by hydrological considerations such as the amount of water in the reservoirs.

423               Based on data in the NEMMCO Statement of Opportunities, and according to Dr Price’s calculations, which I accept, the average growth in sent out energy by region over the period 2003/2004 to 2012/2013 is projected to be as follows:

1. New South Wales – 2.5%

2. Victoria – 1.8%

3. Queensland – 3.2%

4. South Australia – 1.5%

5. Tasmania – 1.5%

6. NEM weighted average (excluding Tas) – 2.4%

A continued demand growth is forecast for all regions in response to increases in population, real income and the cost of substitute fuels. This may be offset to an extent by rising electricity prices, increased production from renewable sources and demand side initiatives. Growth rates are highest in Queensland because of the increasing utilisation of air conditioning in that State.

424              Peak demand will vary across the years depending on weather patterns. The peak demand forecast in the NEMMCO Statement correspond to an extreme, an average and a mild temperature at the time of peak demand. A median forecast is made based on temperatures expected to occur one year in two and a low forecast based on temperatures expected to occur nine years in ten. All regions in the NEM are currently summer peaking save for New South Wales. Under the median temperature medium economic growth forecast, New South Wales is forecast to become summer peaking by 2010/2011. Dr Price has calculated the projected average growth in peak demand by region for the period 2003/2004 to 2012/2013 from data presented in the NEMMCO Statement of Opportunities thus:

a) New South Wales – 2.4%

b) Victoria – 2.9%

c) Queensland – 3.9%

d) South Australia – 3.1%

e) Tasmania – 1.5%

f) NEM weighted average (excluding Tas) – 3.0%.

425               There is a number of new industrial loads proposed for the NEM. An 875MW aluminium smelter, near Gladstone in Queensland, the Aldoga Smelter, has received environmental approval and requires only a small number of further permits and approvals before construction commences. I have no reason to believe that additional generation capacity would not become available over time to meet the growth in demand in accordance with normal market mechanisms.

426               In my opinion the NEMMCO Statement of Opportunities provides very conservative projections which are unlikely to predict actual market or new entrant behaviour in response to material price rises underpinned by supply/demand imbalance. The extreme nature of the assumptions and the narrow criteria for committed projects adopted by NEMMCO do not render its projections particularly helpful for the purpose of these proceedings as a guide to future market behaviour in response to price signals arising out of supply/demand imbalance. This is not to criticise the NEMMCO projections. They no doubt serve a variety of purposes relevant to the development of the market and the approach of government and regulators to pricing and policy affecting its operation. In my opinion however, the industry does respond to supply/demand imbalance with increases in generation capacity and it is highly probable that it will continue to do so, particularly in the large market that is Victoria. This finding has implications for the assessment of Loy Yang’s market power in the foreseeable future. That will be separately considered later in these reasons.

Market Power – Definition and Relevance

427               The concept of market power was explained by the Trade Practices Tribunal in Re QCMA thus:

‘… the antithesis of competition is undue market power, in the sense of the power to raise prices and exclude entry. That power may or may not be exercised. Rather, where there is significant market power the firm … is sufficiently free from market pressure to ‘administer’ its own product and selling volume at its own discretion’. (at p 188)

428               In Queensland Wire Industries Pty Ltd v Broken Hill Pty Co Ltd (1989) 107 CLR 177, market power was described by Mason CJ and Wilson CJ as the ability of a firm to increase prices above supply cost without rivals taking away customers in due time (at 189). Dawson J said the term was ‘ordinarily to be taken to be a reference to the power to raise prices in a sustainable way’ allowing that it had aspects ‘other than influence upon the market price’ (200). The majority in Melway Publishing Pty Ltd v Robert Hicks Pty Ltd (2001) 205 CLR 1 said at 21:

‘The notion of market power as the capacity to act unconstrained by the conduct of competitors is reflected in the terms of s 46(3).’

See also Boral Besser Masonry Ltd v Australian Competition and Consumer Commissioner (2003) 195 ALR 609 at 635 (Gleeson CJ and Callinan J) and 664 (McHugh J). The relevance of market power in the present case has already been adverted to in the summary of the ACCC’s contentions earlier in these reasons. Although no doubt market power does have aspects other than pricing behaviour it was with respect to pricing behaviour that issue was joined in the present case.

 

Market Power – The Contentions

429               AGL contends that, absent market power on the part of at least one of the participants in a vertical merger, there is virtually never any anti-competitive effect. It is part of its case that neither Loy Yang nor AGL enjoys any significant market power in their respective markets. This is reflected in the pleadings and submissions with specific claims that:

1. AGL is competitively constrained in the supply of electricity to Victorian retail customers.

2. There are substantial proposals for the construction of new generator capacity and the expansion of existing capacity in Victoria which will provide ongoing competitive constraints for Loy Yang.

3. Loy Yang and other Victorian generators compete with and are constrained in their pricing by inter State generators.

4. Loy Yang is unable to substantially and profitably price electricity above the average long run costs of generating it and is unable to sustainably and profitably withhold its generator capacity from availability at a competitive level.

5. Loy Yang is competitively constrained by other generators in the provision of electricity derivative contracts to retailers.

430               The ACCC has raised a positive case of market power enjoyed by Loy Yang at present and after the proposed acquisition. It relies upon structural characteristics of the wholesale markets by reference to market concentration, import constraints, barriers to entry, vertical integration between wholesalers and retailers and the absence of countervailing power from consumers and retailers. It relies also upon bidding and price setting in the regional Wholesale Electricity Markets and the effect of hedge contracts on the generators’ bidding behaviour. Ultimately it contends that LYP will have a greater incentive to exercise its market power if the acquisition proceeds.

431               The ACCC rested its case substantially, although not exclusively, upon econometric evidence presented by Professor Frank Wolak to demonstrate that the Loy Yang Power Partnership had the ability to move spot market prices through unilateral bidding behaviour in the period 1 July 2000 to 30 June 2003. This evidence was contested by AGL witnesses, Professor Hogan, Dr Hieronymus, Mr Ergas, Dr Webber and Mr Smart. It was relied upon and supported by Professor King of Melbourne University. Before considering that evidence it is useful to consider evidence of bidding and pricing behaviour by LYPM in the summer of 2000/2001 and the market response to it.

The Loy Yang Power Station – Market Power and Pricing in the Summer of 2000/2001

432              The market behaviour of Loy Yang Power in the summer of 2000/2001 was relied upon by the ACCC in support of its contention that the Loy Yang Partners had market power and that they had exercised that market power to raise spot prices in that summer. There is no doubt that Loy Yang’s bidding and rebidding significantly increased spot prices in the market at that time. But while the ACCC pointed to this as an indicator of market power, AGL characterised Loy Yang’s behaviour in 2000/2001 as a high risk strategy under pressure imposed by its loan covenants and successful due to abnormally hot weather and other fortuitous events.

433               Mr Kenneth Thompson of LYPM gave evidence about the 2000/2001 pricing strategy. Although he, along with other LYPM executives, had an interest in advancing AGL’s case he impressed as a careful witness with significant practical experience in the industry. He worked for the SECV from 1975 until 1993, rising to the post of Civil Works Engineer responsible for civil works activities in the SECV across the La Trobe Valley. He was seconded to the La Trobe Regional Commission in 1994 as an Industry Development Officer. In 1995 he joined LYPM as Account Manager, rising to the position of General Manager Marketing to which he was appointed in August 1999. Between July and October 1997 he worked in the Marketing and Trading Group of CMS Energy in Detroit gaining experience in trading systems and strategies, risk management requirements and portfolio management.

434               Mr Thompson said that from 1995 to 1999 pool prices were low and negotiated contract prices, very low. He referred to a fall in annual average pool prices from approximately $40/MWh to below $15/MWh. The latter figure accords with the low pool prices reported by the ACCC witness, Professor Wolak in a study which he published in 2000 which examined bidding and pricing behaviour of generators in the early stages of the NEM. Mr Thompson’s evidence included a graph prepared by LYPM which compared forecast pool prices from modelling made available to potential purchasers of the Loy Yang Business in 1997 and actual Vesting Contract and pool prices from 1995 to 30 October 2003. The graph demonstrated, and I accept, that pool prices were significantly below those forecast during the period 1997 to 1999.

435               LYPM’s loan covenants require it to meet revenue related benchmarks. The spot price and contract price levels affect its ability to do so. A confidential table of key performance statistics for LYPM in the financial year 1999 showed what could only be described as a low average rate of MWh earned in that year. LYPM was facing termination of the Vesting Contracts at the end of 2000. If it were to replace them with contracts at prices being requested by retailers and other parties it was likely to default on its Senior Debt Service Ratio in the loan covenants. Mr Thompson established a task force to plan for this contingency.

436               The formation of the task force led to the formulation of the LYPM 2001 Pool and Contracting Strategy, also known as the 2001 Summer Strategy. Its essential elements as described by Mr Thompson were:

1. LYPM would ensure it had a high degree of exposure to the pool for the period between January and March 2001; and

2. LYPM would, during January and March 2001, bid and re-bid in ways designed to take advantage of any tight demand/supply periods that might arise over the summer months.

The company decided to implement the strategy. The relevant Board paper noted there was a high degree of certainty that a ‘business as usual’ scenario would lead to business failure, not just in terms of a technical default but a real default under the debt agreement.

437               The proposed strategy was set out in the Executive Summary of the Strategy Document in the following terms:

‘Loy Yang Power (LYP) will contract at current market rates for all periods where customers could contract their total requirements without any supply from LYP. At other times where LYP capacity is required, LYP will retain spot market exposure unless contract prices in this period rise to levels which would provide annual average contracts in excess of LYP business requirements. Simply translated, for the year 2001, LYP would retain total spot exposure for Q1 peak periods (ie no contracts) and contract all remaining periods at market rates for volumes in keeping with normal practice…. Consequently, LYP would have the ability to set price at times during the Q1 peak and secure revenue opportunities that it would not be able to achieve if highly contracted.’

Q1 is a reference to the first quarter of 2001.

438               The key risk strategy was said in the paper to be:

‘… not so much in terms of market behaviour or price risk, but any intervention that might occur from regulators or the State government.’

The paper nevertheless maintained that the strategy did not rely upon illegal or unethical action nor was it designed to withhold physical supply that would increase the likelihood of government intervention as in the preceding summer when electricity restrictions were mandated. The impact of increased spot prices on forward contract prices was expressly recognised in the Executive Summary:

‘The Proposed Strategy provides a significant opportunity to achieve business goals in 2001. The consequences should be evident in following years with a significant increase in terms of forward prices due to such a quantum change; even if the strategy did not deliver in 2001, the fact that Loy Yang Power had taken a position in Q1 peak would have a lasting legacy given that there would always be the future threat of repeated action, thereby pushing up forward prices significantly.’

439               One section in the body of the document dealing with contract levels observed that:

‘2.3 Tailored Contract Level.

In assessing its target contract level Loy Yang Power must review its ability to impact price. Loy Yang Power is the largest generator in the Victorian Market and as such is able to influence the market price significantly at times. However the Victorian market is characterised by over supply. Therefore there are only limited periods where Loy Yang Power can exert sufficient influence for the move in market price to have a benefit in terms of revenue, eg in the spot market we may remove half our capacity but not double the spot price.’

440               The paper recognised that once LYPM was contracted above a certain level its ability positively to affect its revenue through the pool reduced dramatically. If there were no contracts in place and 50% of capacity was withdrawn prices would have to double to be revenue neutral. If 25% were contracted, prices would have to quadruple. The higher the contract coverage the less volume to be removed from the pool and the greater the impact on the spot price required to remain revenue neutral. The aim of the strategy was to withhold contract volume during calendar year 2001 in those periods where LYPM had sufficient market influence to positively effect its revenue levels.

441               The paper referred to modelling work undertaken by LYPM to assess its ability to influence the market. It was focussed on the contracts market and was designed to assess whether, if LYPM held back contract volume, that volume could be supplied by other sources. The modelling showed, that in off peak or low demand periods in any month, LYPM could not influence the market and if it held back contract volume then the market would supply the retailers’ load with ease.

442               For peak demand periods in the milder months, LYPM might be able to place pressure on prices but higher contract prices would drive retailers to lower their desired contract levels and generators to raise their desired contact levels. This would limit LYPM’s ability to contract and place it in an exposed pool position. In the colder months, if it held back it could influence price significantly but would only expect to pick up contracts to cover less than half of its capacity. The position in the colder months would improve if winter were volatile as it would drive retailers to perceive winter as a higher risk period. For peak demand periods during the hot months sufficient contract cover was required by retailers to ensure that LYPM could influence price significantly if it held back and was the last generator to contract.

443               Mr Thompson was aware of significant risks attached to the strategy including the supply demand balance which would be affected by external events such as weather extremes, supply events such as generator unit trips and failures and transmission line outages and network repairs. In his experience, only a very small number of periods in a year have the demand/supply balance where a single generator’s bids have the potential to significantly influence the pool price. Such periods by their very nature are unpredictable. He was aware that for the 2001 Summer Strategy to succeed, there would have to be a number of working days when the temperature in Melbourne exceeded 35^oC in order to drive high demand. There was also the risk that interconnectors between the regions would remain unconstrained.

444               Although in Mr Thompson’s view it was possible for the 2001 Summer Strategy to be successful without other generators adopting similar bidding strategies, the likelihood of its success would be greater if other generators in the NEM were not fully contracted. However, he had no knowledge of their contracting position. He was concerned that it might be higher than normal because of LYPM’s lower contract position. Mr Thompson was also aware that some large consumers of electricity had the ability to alter their demand patterns in times of high pool price and that this could impact on the success of the Strategy. There was also a risk that regulators or the State government might intervene in response to political pressure.

445               In the event, throughout the summer period of 2000/2001 LYPM increased its exposure to the Victorian pool price although where prices offered were sufficiently high the company was willing, and preferred to reduce risk, by selling contract cover. The net result of its contract trading leading up to the commencement of 2001 was that LYPM was relatively highly exposed to pool prices so that it would benefit if they rose in the Victorian or New South Wales regions.

446               LYPM’s Strategy was successful. It avoided financial default in 2001. It was successful primarily because 2001 was one of the hottest summers on record. Another event contributing to the success of the Strategy was a bushfire which broke out on 15 January 2001 in the area of the New South Wales to Snowy interconnect. This prevented electricity flowing from Victoria and the Snowy region into New South Wales and the New South Wales price reached the then VoLL of $5,000/MWh for ten half-hour trading intervals. Prices in Victoria reached only $65.30. However LYPM earned significant revenues from the higher New South Wales pool prices because of its net contract position on the New South Wales Node.

447               A further relevant event was a mechanical fault in LYPM’s Unit 2 on 7 and 8 February. LYPM was of the view that the plant should be taken out of service. NEMMCO however issued a direction that the unit continue to be dispatched. When such a direction is made the usual method for determining and settling pool market prices is changed. The spot price was determined as if the unit had not run. LYPM received the higher of the pool price and its costs of operating unit 2. Because the unit was generating about 500MW, other generation was not dispatched even though various bids for it were lower than the pool price determined pursuant to the NEMMCO direction. Such generators received compensation payments based on pool market revenue that they would have received had they been dispatched. On 7 February the price averaged $400.35/MW and on 8 February $1,000.98/MW. Since that time the rules have changed so that a generator which is subject to a NEMMCO direction no longer receives the greater of the pool price and its costs. It receives only its costs.

448               On 20 February 2001, demand across Victoria, New South Wales and South Australia reached approximately 20,000MW and prices reached $3,681.22/MW in Victoria. Forward hedge prices rose in response to the increase in prices. Announcements were made of new generation plants by various entities. They started in March 2001. Following the announcements the rise in forward contract prices stopped. Hedge prices for 2002 and beyond began to fall and the hedge price for each subsequent year was lower than its predecessor. So much was indicative, in my opinion, of a competitive market at work.

449               In May 2001, Mr Thompson and his group considered their bidding and contracting strategy for 2002. Forward contract prices for 2002 and 2003 meant that LYPM was no longer under the same pressure that it had been in 2001 in respect of projected revenues. For 2002 it was likely to be able to achieve acceptable revenues without exposure to the various risks associated with a continuation of the 2001 Summer Strategy. According to Mr Thompson, those risks had increased considerably since the previous year because of the expected introduction of announced new peaking generation, the expected development of transmission interconnectors and the higher price levels of hedge contracts. By risking another year of significant pool exposure, LYPM could lose significantly more revenue by not contracting than it would have lost in the previous year.

450               In 2002, the average rate earned by LYPM was over $6/MWh greater than its average rate for 2001. Mr Thompson was pressed in cross-examination with the proposition that the higher prices achieved in January and February 2001 were not simply explained by the fortuitous events of bushfire and mechanical outage. His own marketing reports to the LYPM Board for January and February 2001 referred to days of abnormally high temperatures across three States generating high demand coupled with Loy Yang’s bidding strategy. He accepted, as he had stated in the marketing reports, that the major contributors to the higher prices in these months were higher demands and changed Loy Yang Power bidding strategy. He accepted that there was a legacy of higher forward contract prices, but said that within two or three months new entrants had arrived and ‘killed it’ (T 572). He agreed that the strategy adopted in summer 2001 had been successful in that it mitigated against financial failure at the time. He accepted the general proposition that although a generator could by bidding influence pool price in a very small number of periods, if there were an influence, it would have dramatic effects on the pool price.

451               Mr Thompson maintained that in 2000/2001 the company’s focus was on survival. It was very focussed upon producing enough cash to meet its debt requirements. The company was extremely lucky. There were events impacting favourably on forward contract prices which were completely outside Loy Yang’s control and fortuitously close in time to the summer period. These events were:

(a) an industrial dispute affecting Yallourn Energy which resulted in that company’s power station being shut down for over four months in 2000;

(b) failure by the Edison Mission Generator for six months through a cold winter when the average price was about $60/MWh;

(c) the California crisis in the United States in 2000 which had resulted in limited supply and very high prices – presumably the impact of this event was to induce lack of confidence in price stability and a preparedness to pay more for forward contract protection;

(d) suggestions by market critics that there would be blackouts for some of 2000/2001;

(e) a strike on 2 November 2001 in which Loy Yang, Yallourn and Hazelwood were shut down and prices went to VoLL for about six hours;

(f) the fact that the summer was one of the hottest summers on record and the hottest summer at any time since 1980;

(g) that the hot summer was experienced not only in Victoria but also in New South Wales and South Australia;

(h) that the incidence of consecutive hot days in each of the three States on 22, 23 and 24 January (all week days) where on each occasion demand was either at or exceeding 20,000MW;

(i) the bushfire which affected the interconnector;

(j) the mechanical fault in the Loy Yang unit on 7 and 8 February;

(k) demand across Victoria, New South Wales and South Australia on 20 February 2001 reaching again approximately 20,000MW;

(l) the behaviour of other generators whether voluntarily or otherwise which was not inimical to the Loy Yang strategy.

All of these factors were ‘exceptional circumstances’. Mr Thompson said in cross-examination that if Loy Yang had run the same strategy the following two summers it would have had far worse outcomes than being hedged.

452               Mr Nethercote rejected the proposition put to him in cross-examination that, if contract prices fell too low, a normal first instance trading strategy of Loy Yang would be to reduce contract exposure and try to push up the spot price (T 376). He accepted that, absent the Loy Yang debt problem, he would consider such a strategy if he perceived contract prices to be too low. That consideration would require taking into account all relevant risks. Increasing the pool price would ordinarily feed back into strike prices under the contracts and increase them. The risk of not being dispatched in adopting such a strategy would be less in times of peak demand where there might be a tightening of the supply/demand balance.

453               There is no doubt that generators can, and have, endeavoured to ‘spike’ the pool price with a view to increasing forward contract prices. Mr Fraser acknowledged as much and that, in such a situation, generators would effectively be the price setters and retailers the price takers. He agreed that volatility in the spot market is affected by the extent to which generators are covered by hedge contracts. The less contracted a generator or generators, the more likely the incidence of volatility in the spot market. Mr Nethercote agreed that from a generator’s perspective it would have an interest in spiking the spot price in respect of uncontracted electricity to maximise revenue on the amount of uncontracted generation dispatched (T 371). In that case a long-term objective would also be to increase contract prices. He agreed that the bidding strategy in the summer of 2000/2001 produced contract prices at higher levels which were sustained for the succeeding twelve to eighteen months.

454               Generally in cross-examination, Mr Nethercote appeared to accept that a bidding strategy which had been used by LYPM was to price a small percentage of its capacity at a much higher range than normal to encourage peak load generators to come on stream. The relevant question and answer in his cross-examination were as follows:

‘Q. But in terms of your bidding strategy there have been times when you’ve considered it economic for you to put in your offer for the bulk of your generation at low levels and leave some of your generation to be bid in that very high prices, leaving the peak load generators the gap in between, so to speak?

A. We’ve had that sort of a mix.’(T 408)

He rejected the claim that bidding part of a capacity at high prices amounted to economic withholding of that capacity. He agreed that LYPM would opportunistically change bid patterns to very high prices, in excess of $400/MWh. This would be done if LYPM perceived that it would deliver sustainable levels.

455               An example put to Mr Nethercote assumed that LYPM had 300MW uncontracted and offered 295MW at a low price, $35/MWh, and 5MW at VoLL. On that scenario only 5MW would be at risk of not being dispatched. If that tactic raised the pool price it would set it for the full 300 uncontracted MW. He was asked:

‘So that in times of peak demand, where there are say interconnector constraints perceived, you would engage in a strategy, would you not, in terms of your uncontracted generation capacity of putting the majority of that in lower price bands, putting a small amount in a very high price band and hope that that small amount in the very high price band was the last generator dispatched?’

And answered:

‘That’s one option, but we don’t always do it that way.’

456               In my opinion the market tactics here being discussed assume the character of something that looks less like the exercise of market power than moderately well informed betting on the market. The latter characterisation is reflected in the observation of ACCC expert, Professor Frank Wolak of Stanford University, who described a competitive electricity market as ‘… an extremely complicated non-cooperative game with a very high-dimensional strategy space’ – Wolak, ‘An Empirical Analysis of the Impact of Hedge Contracts on Bidding Behaviour in a Competitive Electricity Market’ (2000) 14 International Economic Journal 1-39 at p 4. There is no doubt that LYPM did affect spot prices and forward contract prices by reason of its bidding strategy in summer 2000/2001. I am not satisfied that it has adopted that as a general strategy subsequently or that such a strategy could be relied upon, at the level of confidence necessary for commercial decision-making, to work in conditions other than those which fortuitously came together in that summer. The risk LYPM took was the lesser of two evils, the other being default under its loan covenants. No doubt, as Victoria’s largest generator, it is in a position opportunistically to respond to supply/demand imbalance in very short time intervals and if all the variables are in the right place, to affect both spot and forward contract prices. The question is whether the existence of such opportunities and the fact that it responds to them from time to time reflects the existence of market power. There is here a distinction to be drawn between what was referred to as ‘transient market power’ and ‘persistent but intermittent’ market power. It may also be that that distinction is able to be reflected in the concept of temporal sub-markets and what is elsewhere described as the inter-temporal variation of market power.

Market and Regulatory Response to Pricing in the Summer of 2000/2001

457               There was a reaction to the 2000/2001 summer spot price increases. This came in the form of a change to the rebidding rules under the Code. In April 2001, NECA published an assessment of the performance of the market in summer 2000/2001. It noted that prices in Queensland over the summer had averaged $52/MWh which was the lowest since market launch. This had been contributed to by the Queensland-New South Wales interconnector (QNI), commissioned on 18 February 2001. Average prices nevertheless rose by just over a third in New South Wales and by significantly more in Victoria compared to the previous summer. On the other hand average prices in South Australia were lower than the previous summer, but higher than the summer of 1998/99. The generally higher prices were attributed to extreme weather conditions. Average maximum temperatures were higher than long term averages throughout the summer across all regions significantly increasing the demand for electricity. In South Australia, the January average maximum temperature was 33^o, up 4.5^o on the long-term average. Increases in peak total demand in all regions were significant. They were 4% in Queensland, 7% in South Australia and Victoria and 8% in New South Wales compared to the previous summer. The total energy demand increased by 1,800GWh or 5% over the previous year. NECA observed that despite these extreme conditions and the higher demand there was no involuntary load shedding because of supply shortfall. There was also clear evidence for the first time of a direct demand-side response to prices. This was regarded as an immensely encouraging sign for the market.

458               NECA observed however that the amount of rebidding activity had increased by about a fifth compared to the previous summer. There was an emerging trend towards greater price volatility in response to relatively small changes in demand, particularly in Victoria because of some participants’ bidding strategies which had priced all capacity at either very low or very high prices significantly changing supply characteristics.

459               NECA attributed the trend to greater price volatility against relatively small changes in demand in the New Year, particularly in Victoria, to the bid structure presented by LYPM. On twenty-two days and particularly in the week of 19 February, LYPM had been dispatched to around half its capacity during the afternoon despite having presented 2,000MW to the market. On those occasions it had offered around 1,400MW at low prices and about 600MW at prices above $4,500/MWh as part of its day-ahead bid. Hazelwood Power had also introduced a two price bid structure from around 19 February with capacity presented at either zero or greater than 4,800/MWh as part of its day-ahead bit. In both cases, capacity was often shifted down to low price bands in small amounts during the day closer to dispatch. The short-term impact of the strategies saw average prices in the spot market increase as higher cost replacement capacity was dispatched.

460               In May 2001, NECA published an Issues Paper entitled ‘Bidding and Rebidding Strategies and their Effect on Prices’. The paper reported that total rebidding activity from November 2000 to April 2001 had increased by 20% over the same period in the 1999/2000 year. The paper reported that the scope for manipulation of rebidding had emerged in practice in February 1999. It referred to particular events supporting that conclusion. Subsequently the practice of significant rebids close to dispatch largely ceased following the NECA report of its investigation into those events. The paper canvassed options for additional safeguards including a prohibition or restriction on rebidding. NECCA invited submissions.

461               On 6 July 2001, AGL South Australia Ltd wrote to NECA in response to the invitation for submissions. It supported ‘immediate change’ to a position which would ‘mitigate or minimise the opportunity for price gouging when situational market power exists’. It argued that rebidding should be restricted except for bona fide technical reasons. It commented upon market structure, rules and power observing that:

‘As a market in transition from vertically integrated monopolies to competition there are many aspects where time will be required to allow for adjustment to the new conditions. Collapse of the pool price in the early days of the wholesale market and the subsequent run-up in the peak summer months are manifestations of such adjustments.’

In the longer term it expected that generation capacity and set size could be expected to adapt to the profile of the new market to avoid the excess that triggers collapse and the shortages which push prices to VoLL. AGL South Australia observed that generators were engaging in certain types of conduct with the intention of raising prices above competitive levels. The submission argued that ‘[t]he ability of suppliers to control capacity in a tight market combined with a single price auction provides the basic framework for the exercise of market power in both markets’ (ie California and the NEM). It also observed that under conditions of peak demand or a downgrade or outage of an interconnection, tightness of supply could provide generators with the opportunity to lift prices without the risk of a competitive response by other producers.

462               AGL South Australia identified the mechanism for the exercise of market power in the NEM as the practice of economic withholding of capacity and strategic rebidding. It referred to the studies carried on by Professor Wolak and applied one of his techniques, using South Australian data, to test for the presence of persistent above-cost pricing outcomes. The data was analysed to determine differences in the spot price behaviour within the day and week across peak and off-peak periods. Comparable periods in 2000 and 2001 were chosen. The figure was said to illustrate a persistent pattern of extremely high prices between half hour periods 24 and 35, Monday to Thursday, in both winter and summer. The pattern was said to be similar to that identified by Professor Wolak in an analysis of the UK market at least to the extent that Friday exhibited less variability. AGL cited Professor Wolak’s argument that this pattern of prices represented the exercise of market power by generators:

‘It should not be possible in a normally competitive market to lift prices persistently and to such extent above the efficient price level.’

463               To a degree the AGL South Australia submission to NECA in July 2001 was just that, a submission. It did, however, offer some insight into the operation of the NEM which carries weight because it came from an important market participant offering the perspective of a retailer unconcerned by any pending alliance with a generator. I am therefore inclined to accord weight to it to the extent that it offers a commercial perspective on the operation of the market.

464               On the other hand the ACCC itself made a number of important observations about the impact of the rebidding strategies and the rebidding rule change in its determination of NECCA’s application for authorisation of the changes to the bidding and rebidding rules and in particular Rule 3.8.22 which has been set out earlier in these reasons (X 32). The determination noted that the proposed rebidding code changes had been developed by NECA after criticism of price outcomes that arose during the summer of 2000/2001. It referred to NECA’s process of consultation with the release of its Issues Paper, Draft Proposals for Change which had been published in July 2001 and the release of a report by the Code Change Panel in September 2001 recommending Code changes governing bidding and rebidding rules. It referred to NECA’s concerns about the use of rebidding to generate higher prices but observed that the number of bids and rebids that give cause for genuine concern was currently ‘comparatively very small’. Most rebids were benign. It was accepted by NECA that rebidding provided essential flexibility to generators to enable them to respond to changes in physical and commercial circumstances.

465               The ACCC considered it prudent to introduce the requirements proposed that bids and rebids be made in good faith and therefore represent the true intentions of generators. The requirement to submit bids and rebids representing a generator’s true intention was intended to give them the incentive to submit initial bids that were meaningful and accurate, rather than bids that were generic and relied upon the ability to rebid. The ACCC referred to concerns about recent bidding and rebidding behaviour in the NEM which appeared to be characterised by economic withholding. However in relation to network constraints it said:

‘Behaviour that deliberately exploits constraints effectively punishes the market for under-investment or lack of development. Price spikes identify investment opportunities in transmission or generation. Without these signals the energy-only market would cease to work effectively. The fact that such constraints can be used indicates that constraints and congestion in transmission pricing in the NEM is being inadequately addressed. Investment opportunities may be more effectively signalled if there were more regions in the NEM, or if nodoal pricing was introduced. In their absence, pricing signals on the supply side should be maintained. Any muting of these signals will raise questions about the market’s design and its ability to develop into the future.’

466               The ACCC referred to evidence that the market was already addressing constraints including investment in base load generation in Queensland and peaking generation in South Australia and Victoria. Proposals to upgrade and build new interconnectors to take advantage of price differences between regions were also referred to. It was said:

‘This is the market at work.’

The further observation was made that short term price spikes are common to deregulated electricity markets and that focussing on price outcomes over a short period of time was irresponsible as it could lead to biased conclusions. The high priced period of 2001 that raised initial concerns were seen by the ACCC as ‘… largely part of the cycle of development for the NEM’. Then it was said:

‘While recent bidding behaviour lends support to the view that changes should be made to prevent contrived price spikes, the Commission is wary of short term solutions that could impact negatively on the long term development of the market. Having said this, the Commission believes that there are opportunities for the NEM framework to be refined if sufficient consultation is conducted to target specific anti-competitive behaviour.’

467               The ACCC also observed in its determination that two of the three consultants’ reports to it had concluded that market power in the NEM was not a systemic problem. In relation to the spot market there was little to indicate conclusively that there was significant abuse of market power occurring at that time. There was no evidence that contract prices were remaining above new entrant levels without enticing new entrants, as might occur if new entrants were being deterred by the threat of predatory pricing. Contract prices appeared to be at or below new entrant levels. There was, however, some evidence of the average spot market price exceeding new entrant prices which would be of concern if the condition persisted over a period long enough to allow new entrants to enter, for example two years or longer.

468               The ACCC is a party to the present litigation. However the observations contained in its Authorisation Determination were made outside the framework of the litigation and unaffected by any consideration of the potential acquisition by AGL of a 35% interest in Loy Yang Power. The Authorisation Determination represents a considered assessment of the consequences for competition of bidding and rebidding practices and the extent of market power in the relevant market place. Like the commercial perspectives reflected in the submission from AGL, also outside the framework of the present litigation, they may be taken into account in forming a balanced view of the operation of the relevant markets.

469               In my opinion, the events of the summer of 2000 and 2001 and the perspectives on those events offered by the AGL submission and the ACCC determination in the context of the rebidding rule change support the following broad conclusions:

1. The rebidding rules as they were prior to the change in 2003 provided some opportunities for manipulation of the bidding process to affect spot prices, particularly in times of high demand.

2. The opportunities afforded to generators by the previous bidding rules either conferred market power or enhanced existing market power for brief intervals when bidding strategies could affect spot prices and forward hedge contract prices.

3. The change to the bidding rules has reduced the opportunity for manipulative bidding and has increased the risk of punitive responses by NECA against attempts to lodge bids other than in good faith or rebid other than for legitimate reasons.

4. To the extent that the bidding regime does permit price spiking and economic withholding of capacity at times of high demand, it provides a mechanism for price signals upon which existing participants can act to enhance capacity or new participants can enter to relieve the demand/supply imbalance.

5. The market did respond in a competitive way to the price increases of the summer of 2000/2001 by the announcement of new generator capacity.

6. The particular conjugation of pressures from financiers and fortuitous events upon which Loy Yang Power relied in the summer of 2000/2001 does not allow an inference to be drawn from the successful application of the summer bidding strategy that it had then and continues to enjoy now, market power in terms of an ongoing ability to price consistently above its marginal costs of production. So to conclude is not to conclude that Loy Yang lacks market power. It is simply that the inference of market power is not supported by the events of the summer of 2000/2001.

Sustainable Pricing in the Wholesale Electricity Market and its Relationship to the Longrun Marginal Costs of Generation

470               In his outline of the NEM, Dr Price compared long run spot prices with the long run marginal costs of generation. The long run marginal cost is the cost of establishing generation capability incorporating a component for the cost of plant. Dr Price regarded the comparison as valid on the basis that the NEM was designed as a pool through which all electricity was to be sold and through which, together with associated contractual arrangements, all generation costs were to be recovered.

471               Dr Price referred to two methods which have been used to estimate long run marginal costs of generation in the Australian context. The first method rests upon a prediction of the costs of the next power station to be built. This is the new entrant cost approach. The costs of the new entrants are used to establish a price benchmark. The justification for this approach is that no generator can price above the new entrant’s cost level for a sustained period because that would attract an investor to build a plant to undercut the incumbent’s price, eventually bringing average prices down to the costs of the new entrant. The second approach is based on the proposition that system load will be met by a combination of plant with varying cost structures namely base load, mid merit and peaking plants. The price in a perfectly competitive market would reflect the least costly mix of those plants as distinct from the cost of a single plant type predicted to be commissioned next.

472               Two recent studies used these methods. The first by McLellan Magasanik Associates which was carried out on behalf of the ACCC found that in Victoria new entrant costs vary between $35 per MWh and $38 per MWh – McLellan Magasanik Associates, Impact of Rebidding on the National Electricity Market : A Report to the Australian Competition and Consumer Commission, May 2002 p 29.

473               A second new entrant study was carried out by ACIL Tasman for NEMMCO. This estimated the long run marginal costs of various types of plant in various locations. It found that base load coal plant costs vary between $31.42 per MWh to $34,66 per MWh, base load to mid merit gas plant varied between $41.93 per MWh and $45.33 per MWh and peaking gas plant costs varied from $126.77 per MWh to $183.47 per MWh. A realistic composition of base mid merit gas plant and peaking plant would allow overall average costs of the order of $40 per MWh, assuming relative proportions of base, mid merit and peaking plant at 75%, 20% and 5%.

474               A study based on the second approach was conducted by Cap Gemini Ernst Young on behalf of the New South Wales regulator. The Cap Gemini study estimated the range of the long run marginal cost of the generation system to be from $36 per MWh to $56 per MWh including the costs of purchasing so-called ‘green’ energy required under Federal and State legislation and market fees payable to NEMMCO. The findings of this study were consistent with those conducted by McLellan Magasanik Associates and by ACIL Tasman.

475               In comparing spot prices with costs, Dr Price presented the (demand weighted) annual (financial year) average spot prices for each State in the NEM together with the long run marginal generation cost estimates of Cap Gemini, which he believed was a more suitable basis for determining the underlying cost structure of the generation system. He produced a Figure which plotted annual average demand weighted prices across the financial years from 2000 to 2003 and also depicted in a shaded area the Cap Gemini estimate of the long run marginal costs for the generation system. A reproduction of that Figure (Figure 28 in his statement) appears as Annexure 7.

476               The average prices are generally well within the Cap Gemini estimate of the long run marginal costs for all years except for the first two years of its operation in South Australia. In the early years of the NEM generation reserves were relatively low in that State. As new interconnectors were developed and new generation capacity commissioned the price fell to levels reflecting efficient generation costs. Dr Price made a similar comment about the Queensland costs which does not appear to be borne out by his graph. Queensland appears generally to have fallen within the longrun marginal cost range shown on the Figure.

477               South Australia has experienced the highest average spot prices and a high degree of price volatility. This appears to have resulted from a tight supply and demand balance, relatively more expensive generating plant and limited interconnection to Victoria. Since the middle of 2001, South Australian prices have converged with those of other NEM regions. This is largely due to the commissioning of new generating capacity in that State, including Pelican Point and the peaking plant near Adelaide, and the additional interconnection capacity provided via Murraylink.

478               Queensland also experienced a high degree of price volatility before mid 2001 and relatively high spot prices in 1999 and 2000. This arose from:

1. Limited interconnection – where the regulated interconnector, QNI, had not been commissioned and the interconnection via Directlink was limited.

2. A tight supply and demand balance requiring new generating capacity.

Queensland now enjoys the benefit of the regulated interconnector, QNI, and additional significant low cost base load generating plant including Millmerran which has a capacity of 840MW, ‘Callide C’ which has a capacity of 840MW and Tarong North which has a capacity of 450MW.

479               The average spot prices in New South Wales rose from a low of $25/MWh in 1998/99 to about $35/MWh in 2002/03. I acceptthat this reflected a reduction in the level of excess generation capacity in that State. Growth in demand across the NEM has led to a reduction in surplus capacity and subsequent higher prices. There was price volatility in Victoria in the summer of 2000/01 and later in 2001 because of particular factors including extreme summer weather conditions and industrial relations disputes. Since then, however, there has been a reduction in the average Victorian spot price.

480               The large variations in the half hourly prices were illustrated by graphs prepared by Dr Price on a logarithmic scale for each NEM region, each season (cold, hot and mild) and each type of day (weekend, public holiday and working weekday). He attached to his outline the charts for the Victorian region. They were based on NEMMCO dispatch data for the 2002 calendar year. The figure for that year reflected the most up to date information including the effect of new generator capacity and interconnectors and the doubling of the VoLL price cap in April 2002 to $10,000 per MWh and therefore some of the most extreme pricing events likely to be seen in the NEM. The charts support the observation that prices rise at times of high demand when capacity is relatively scarce and fall at times of relatively low demand. The NEM price setting mechanism produces a pricing pattern which reflects marginal costs of production including the ‘scarcity’ value of capital. Price changes signal to generators when new capacity is required and the type of capacity required. Periodic price spikes are best managed by investment in short-term demand reduction or a highly responsive peaking plant. A sustained rise in price can most effectively and efficiently be managed by the development of a new base load plant.

481               Dr Price observed that periodic price spikes should be considered in the context of average price. If such price spikes are associated with a sustained rise in average prices above the efficient long run marginal cost of the generation system, this might be cause for concern. There is no such price trend. The opposite has occurred. Since the commencement of the NEM there has been a significant fall in average prices. In this respect, the National Grid Management Council said of the role of prices in an NEM:

‘Pool prices are intended to reflect the value of electricity that is traded at a particular point in time. If prices reflect the value of the product, then for an appropriate investment in generation, revenue from the pool will cover both its variable production costs and its fixed costs over the life of the investment.

There is no guarantee however, that in a particular period, the revenue a generator receives from the pool will cover its costs. The lumpy nature of the capital and the variability of demand, mean that there will be some periods in which pool revenues more than cover total costs, and other periods when it does not.’ – National Grid Management Council, Transition to a National Electricity Market, July 1993 p 11

482               According to the pricing evidence, the NEM works efficiently. That is to say prices rise at times of high demand and fall when demand is low. In those NEM regions where reserves are short, the price is relatively high. Where there is excess capacity, prices are relatively high and power flows from low priced to high priced regions.

483               By reference to NEMMCO half hourly price data for the 2002 calendar year at the Victorian Node, it appears that the overall average cost of generation is around $40/MWh as estimated by Dr Price. This figure is supported by Mr Ergas’s estimate of long run marginal cost for the operation of the Loy Yang power station which is discussed below. Dr Price’s analysis of the data indicates that the price at the Victorian Node during 2002 exceeded $40/MWh for only 13% of the year with the price being below average costs for the majority of the time. As a result there was an average spot price over the 2002 calendar year of $35.44/MWh. On this basis he argued that generators do not have the ability to sustain average spot prices above average costs. He also observed that Loy Yang Power was, for the most part, a price taker in the sense that it did not set the Victorian spot price for a significant percentage of the year. The actual figures were referred to in a confidential part of his report.

484               Reference should also be made to the approach to the calculation of long run marginal costs for the Loy Yang Power Station and coal mine adopted by Mr Ergas, who also gave evidence for AGL. He calculated the LRMC for Loy Yang for the period July 2002 to the end of July 2003 and the average monthly price received per MWh dispatched. This was said to show that Loy Yang’s average revenue was below its longrun marginal cost (LRMC). That conclusion was relied upon to support the proposition that Loy Yang did not have market power over the relevant period and does not have market power to day.

485               Mr Ergas referred to other studies of generator costs and in particular the ACIL Tasman Study which was conducted in April 2003 for NEMMCO and a paper presented at the ABARE Conference in 2002 by Barry Naughton entitled ‘Economic Assessment of Combined Cycle Gas Turbines in Australia’. Both papers contained estimates of LRMC for a ‘greenfields’ brown-coal base load plant which were $38.9/MWh and $33.60/MWh dispatched for ABARE and ACIL respectively. The ABARE estimate was higher than the ACIL estimate because it used higher figures for some elements of its calculation.

486               Mr Ergas identified the values he adopted for the various parameters which he required to calculated the LRMC of a brown-coal base load plant:

1. Capital cost - $1,800 per kw +/- 15% to allow for cost overruns and efficiencies in plant construction – this figure was taken from the ACIL study.

2. Economic life – 30 to 40 years respectively estimated by ACIL and ABARE. Most likely value was set at 35 years.

3. Capacity factor – a most likely rate of 92.5%.

4. Heat Rate – 12.856 to 12.906 gigajoules/gWh dispatched – most likely value at mid point.

5. Fuel price - $0.477/gigajoule to $0.38/gigajoule.

6. Operating and maintenance costs - $18 per kW. Fixed operating costs - 0.002 per kW variable costs.

7. Real pre-tax weighted average capital costs (WACC) – 10.57% to 13.8% with likely mid point 12.08%.

The LRMC was calculated with combinations of the variables for which ranges were produced and the specified value for variables with no range. In each calculation the value for variables for which a range was produced was selected randomly for the range specified.

487               From the preceding calculations Mr Ergas derived a distribution of LRMC of a brown-coal power plant. He derived a mean LRMC of $40.25/MWh in an interval of $35.55/MWh to $44.96/MWh dispatched at 95% confidence. This was a figure for a greenfields plant. To relate it to the Loy Yang Plant he calculated its depreciated optimal replacement cost of the plant and reran the analysis. He obtained a mean LRMC of $39.17/MWh with a 95% confidence interval of $34.44 /MWh to $43.89/MWh.

488               In an appendix to his statement Mr Ergas set out in detail the basis upon which he calculated the real pre-tax WACC. He was challenged in cross-examination about certain aspects of that calculation. These related to a number of components of the calculation which included:

. a risk free rate of 5.61% based on the average cost of Commonwealth bonds in the relevant period from January 2002 to July 2003

. a market risk premium (MRP) of 6.5%

. an asset beta of 0.50 to 0.70 (based on a debt deta of 0.00)

. cost of debt of between 191.5 and 229 basis points above the risk free rate

. a gamma of 0.30 to 0.50.

It is not necessary for present purposes to define the components of these calculations. It is sufficient to say that the MRP and beta figures involved evaluative assessments on which views might differ. Mr Ergas justified his assessment of the beta figure by reference to international and other practice. However what this indicates and what he acknowledged is that there is no ‘true’ value for LRMP. It appears to be used as a figure or range of figures which, because of its definition, has persuasive force as a benchmark for assessing whether a firm can be regarded as taking monopoly profits.

489               Mr Ergas also calculated Loy Yang’s average bid price per month over the period June 2002 to June 2003 and found that its weighted average bids were substantially below its LRMC. This, he said, was to be expected as Loy Yang is a base load station and would incur a substantial reduction in profits if it were not dispatched. Indeed over this period energy dispatched from Loy Yang was equivalent to 94% of its maximum generation capability. When auxiliary use is taken into account, a load factor of over 100% is implied. This in turn implies that the station was fully available during the entire period studied and must have been dispatched whenever it was available during that period.

490               The ACCC pointed out that Mr Ergas accepted in cross-examination that a firm may possess market power even though it earns profits below its LRMC and that a snapshot view of LYP’s profitability would not be enough to show whether it could earn profit in excess of LRMC over the life of its assets.

491               There was, as emerged from Mr Denton’s evidence and the cross-examination of Mr Ergas, a good deal of room for debate about how to determine LRMC for a base load generator. Mr Ergas did undertake further calculation of the LRMC after Mr Denton said his estimate was at the high end of what he would consider reasonable. In his re-analysis Mr Ergas relied upon the ACIL Tasman paper preferred by Mr Denton. He added that it is unlikely that over the periods he considered LYP had made substantial profits or had even achieved substantial returns above LRMC.

492               The LRMC estimates derived by Mr Ergas appear to fall close to or perhaps on the upper bounds of a debatable range. They are consistent with the proposition that LYP does not have market power defined by reference to pricing relative to LRMC. His evidence taken with that of Dr Price and the market response to the Summer Bidding Strategy of 2000/2001, leads me to conclude that LYP does not have market power in the sense of an ability to secure price increases free of competitive response. I might add that success at ‘gaming’ in the market during limited periods of high demand does not reflect market power even if it results in a high forward contract price.

493              The ACCC has made subsequent submissions about price spikes said to derive from economic withholding by LYP. I am prepared to accept that there are periods of high demand where a generator may opportunistically bid to increase the spot price. I do not accept that such inter-temporal market power reflects more than an intermittent phenomenon nor does it reflect a longrun phenomenon having regard to the possibilities of new entry through additional generation capacity and the upgrade of interconnections between regions. It does not amount to an ongoing ability to price without constraint from competition.

The Acquisition, the Natural Hedge and the Likelihood of Price Increases – The Economic Case

494               The burden of the ACCC’s economic case in support of the proposition that LYP has market power, and the effect of the proposed acquisition on pricing, fell upon Professor Frank Wolak. Professor Wolak is a professor at the Department of Economics at Stanford University. He is highly qualified. His research interests lie in the areas of industrial organisation, regulatory economics, econometrics and health economics. He has published a number of papers related to the Californian electricity market and has also published articles relating to the early functioning of the NEM in Australia in 1997. It is useful before referring in more detail to his evidence to reflect upon its nature, which involves the application of the technique of econometrics.

495               One of the documents relied upon by Professor Wolak and annexed to his second report was entitled ‘Structural Econometric Modelling: Rationales and Examples from Industrial Organisation’. This is a recent contribution by Professor Wolak and PC Reiss to the Handbook of Econometrics published this year. In the introduction to their chapter the authors refer to a long-standing definition of ‘econometrics’ as ‘a branch of economics in which economic theory and statistical method are fused in the analysis of numerical and institutional data’. They focus upon an aspect of econometrics which they describe as ‘structural economic modelling’ and which appears to underpin Professor Wolak’s approach to the analysis of market power in the present case. There is an absence of consensus among economists on how to build and interpret structural econometric models. Professor Wolak and Mr Reiss propose a general framework for developing and evaluating such models. They observe that economic theories deliver mathematical statements about the relationships between two variables, x and y. Such statements are often deterministic and do not speak directly to the distribution of ‘noisy economic data’. The applied researcher adds the second source of structure which is statistical sampling and other stochastic assumptions specifying how data on variables x and y were generated. This is necessary to transform deterministic models of economic behaviour into stochastic econometric models. The term ‘stochastic’ in this context refers to phenomena following a probability distribution or pattern amenable to statistical analysis but not able to be precisely predicted. So the ‘structure’ in structural models typically comes from both economics and statistics.

496               Structural econometric models can be used to estimate unobservable economic parameters or behavioural responses from non-experimental data. Examples of behavioural or structural parameters include marginal costs, returns to scale, the price elasticity of demand and the impact of a change in an exogenous variable on the amount demanded or on the amount supplied. Structural models can be used to simulate changes in equilibrium outcomes resulting from changes in the underlying economic environment. So the estimated structure can be used to predict what would happen if certain elements of the environment change. Structural models are also used to compare the predictive performance of two competing theories. The authors acknowledge that the advantages of structural models do not always favour them over descriptive models.

497               In the context of market power models Professor Wolak and Mr Reiss observe that most empirical researchers in the field of industrial organisation equate competition with price equal to marginal cost. When price is above marginal cost firms are said to have ‘market power’. Some studies are content simply to estimate price-cost margins. Many go further and attempt to infer what types of firm behaviour are associated with prices that exceed marginal costs. Absent a structural model one cannot infer the extent of competition from the joint distribution of market clearing price and quantity. Put another way, it is necessary to have an economic model which can estimate marginal costs and hence price-costs margins from the joint distribution of market clearing prices and quantities. Such a structural model will involve functional form assumptions and often distributional assumptions that cannot be tested independently of hypotheses about competition. They also observe that while price cost margins can be estimated using a structural model, it is problematic to link them to more than a few specific models of firm behaviour.

498               In considering Professor Wolak’s evidence, I do so against the background that both descriptive economics and econometric modelling, structural or otherwise, offer modes of analysis and predictive tools with respect to market behaviour that have application to policy making and market design in semi-regulated markets. They are necessarily subject to the limitations imposed by their assumptions and, related to that, the incompleteness of the data which inform them and the sometimes complex multidimensional character of commercial behaviour. It is, as Professor Wolak has observed, an unusual feature of electricity markets of the kind under consideration in this case that there is a wealth of data about pricing and bidding conduct. But in a market, which is globally regulated through a central auction based pricing and dispatch mechanism subject to differing regional regulatory regimes defined by State law and affected by a variety of external variables affecting both supply and demand, the past and future of pricing strategies by market participants are not to be judged by econometric models alone, however sophisticated. The Court is concerned to make judgments about competition in a living commercial setting whose actors operate upon conjectures and predictions that may prove to be wholly or partly incorrect and that may be, from an economist’s perspective, irrational. Even economists and econometric analysts are still coming to grips with the workings of semi-regulated electricity markets. The point is well illustrated in a paper about the Californian market published in 2000, Borenstein, Bushnell and Wolak, ‘Diagnosing Market Power in California’s Restructured Wholesale Electricity Market’ (December 2000, American Economic Review pp 1367-1405). There the authors observe that, while market power has been studied and estimated in many industries there has been little attention paid to the intertemporal variation in the ability to exercise market power. In most industries such variation is small because inventories smooth the ability of firms to exercise market power. But that is not the case with non-storable goods. The problem is exacerbated for electricity because demand is very inelastic in the short run and supply becomes inelastic as the capacity limits of the generators are approached.

499               It is also the case, as the authors observed, that because of the electricity industry’s long history of regulation there is little existing work that attempts to estimate the competitiveness of an electricity market based upon actual observed outcomes. Much work up to 2000 had relied upon market simulations based on some form of oligopoly equilibrium. I interpolate that Professor Wolak and some of his colleagues appear to have been pioneers in the attempt to develop analytical tools that address the peculiar characteristics of semi-regulated electricity markets.

500               There is another interesting caveat in the paper on the Californian markets. That is that, in the period under study, prices did not significantly differ from the authors’ estimate of marginal costs, indicating no systemic inefficiencies raising prices in all periods. It was then said:

‘Still, the estimates must be taken with the caveat that they include failure to achieve competitive market price for reasons other than market power, including bad judgment and confusion on the part of some generators or market-making institutions.’

501               The uncertainties of judging and predicting market place behaviour are particularly acute where, as in the present case, the market is relatively immature and where the rules are still changing, albeit the actors are developing experience and increased sophistication. A good example of a rule change relevant to this general overview is that relating to rebidding. As can be seen from the discussion of that rule change in the ACCC’s authorisation determination, it is by no means marginal. Certainly it enjoins caution in extrapolating from past rebidding behaviour, eg of LYPM, what an entity’s future rebidding opportunities and behaviours would be.

502               Professor Wolak placed reliance upon studies which he had published in 2000 and 2003 in developing the methodology which he applied to analyse competition in electricity markets. The first paper, Wolak, (2000), was entitled ‘An Empirical Analysis of the Impact of Hedge Contracts on Bidding Behaviour in a Competitive Electricity Market’ (Summer 2000) Vol 14(2) International Economic Journal 1. In that paper he derived a model of bidding behaviour in a competitive electricity market which incorporated various sources of uncertainty and the impact of the electricity generator’s positions in the financial hedge contract market on their expected profit maximising bidding behaviour. The model was used to characterise the profit maximising market price that a generator would like set by its bidding strategy for several hedge contracts and spot sales combinations. It was applied to bid and contract data obtained from the first three months of operation of the National Electricity Market (which he called NEM1) in Australia. It was said to illustrate the sensitivity of expected profit maximising bidding strategies to the amount of financial hedge contracts held by the generating unit owner. It was also said to provide strong evidence for the effectiveness of financial hedge contracts as a means of mitigating market power during the initial stages of operation of a wholesale electricity market. Professor Wolak defined market power, relevant to generators, as ‘… the ability of a generating company to raise the market price by its bidding behaviour and to profit from this price increase’. To determine whether a generator possesses market power it is necessary, he said, to develop ‘… an accurate model of the optimal bidding behaviour for a generator competing in this market’. His paper demonstrated that a firm’s hedge contract position could exert a dramatic effect on its optimal bidding strategy and its short-term desire to raise the market price. For sufficiently high hedge contract levels a generator should therefore attempt to reduce market prices below its own marginal costs of production by its optimal short term bidding strategy. Importantly for the present case, he observed that:

‘Even given knowledge of a firm’s bidding behaviour in a competitive electricity market, without knowledge of a generator’s hedge contract position, it is difficult, if not impossible, to determine if the firm is able to exercise market power. For a specific bid function, there is often a hedge contract position that can rationalise that bid function as expected profit maximising. This result implies that the strategic value of actual bid functions to other competitors is significantly reduced because a key ingredient necessary to determine a firm’s profits from a given bidding strategy is unknown. Unfortunately, the monitoring value of actual bid functions to a regulator is also significantly reduced for the same reason.’

503               At the time of the study underpinning this paper, Queensland had not been connected to the National Grid. Victoria and New South Wales had been interconnected. The Australian Capital Territory was part of the New South Wales pool and South Australia was trading through the Victorian pool. The complexity of the decisions affecting market participants in the NEM even at that early stage of its history was well described at p 4 of the paper. There, in a passage which I mentioned earlier in connection with the LYPM Summer Strategy for 2000/2001, Professor Wolak said:

‘A competitive electricity market is an extremely complicated non-cooperative game with a very high-dimensional strategy space.’

He observed that a firm owning a single generating set competing in a market with half hourly prices must, at a minimum, decide how to set the daily price for the unit and the quantity bid for 48 half hours during the day. Firms are allowed to bid daily prices in half hourly quantities for ten bid increments per generating set. For a single generating set this amounted to what Professor Wolak called a ‘490-dimensional strategy space (10 prices and 480 half-hourly quantities)’. He referred to the range of bid prices from the minimum negative price to the then existing VoLL of $5,000 AU. Each of the quantity increments is to be greater than or equal to zero and their sum is less than or equal to the capacity of the generating set. As he said:

‘Most of the participants in this market own multiple gen sets, so the dimension of the strategy space for these firms is even larger.’

It is perhaps not surprising that he found sub-optimal, risk averse, bidding behaviour by generators in the market. Prices had fallen precipitously from the time of its formation. Before restructuring, the average price of a MWh of electricity in the area of NEM1 was $35. When separate markets were formed in the States of New South Wales and Victoria, prices in each settled at an average value of roughly $25/MWh. Following their interconnection and the formation of NEM1 in May 1997, average prices of the integrated market fell further to around $15/MWh. On Professor Wolak’s analysis, despite the fact that the marginal cost of generation for many of the large fossil-fuel generators was roughly $15/MWh, it was because of the large quantity of hedge contracts held by the major firms competing in the market that the short run profit maximising market price for those generators was very close to the actual market price set. He observed, against that background, that reductions in a generator’s contract position could significantly increase both the mean and standard deviation of the variable profits it earned from a profit maximising bidding strategy based on a reduced quantity of hedge contracts. He concluded that the presence of excess generation capacity and risk averse generating companies had contributed to low prices in NEM1. A reduction of the amount of excess capacity in the market could lead to higher prices. But such a strategy would only work if the generators found it optimal to contract less and in turn bid less aggressively. Less aggressive bidding would lead to higher prices. The bottom line was still that for all generators best reply prices to be above their marginal costs they must sell less contract cover than they produce in electricity.

504               A design implication of his findings was that effective price regulation could be imposed by forcing a large enough quantity of hedge contracts on newly privatised generators. This was reflected in the Vesting Contracts held by generators in the early stages of the NEM. While Wolak (2000) offers very interesting insights into the early operation of the NEM, it engenders a degree of unease about extrapolating behaviour from the embryonic stages of a new market to subsequent and future conduct in that market as it continues to develop. It engenders a degree of caution about any extrapolation to the present day based upon the observations in that study.

505               A second paper, important to the methodology applied in this case – Wolak (2003a) was entitled ‘Identification and Estimation of Cost Functions Using Observed Bid Data: An Application to Electricity Markets’ published in Dewatripont, Hansen and Turnovsky (eds) Advances in Economics and Econometrics: Theory and Applications, 8^th World Congress Vol II New York Cambridge University Press pp 133-169. This was a development of the study contained in Wolak (2000). It described itself in its Introduction as presenting ‘… several techniques for recovering cost function estimates for electricity generation from a model of optimal bidding behaviour in a competitive electricity market’. The procedures were applied to data from NEM1 to recover cost function estimates for a specific market participant. Professor Wolak found close agreement between the cost functions recovered from these procedures and those obtained from engineering estimates. The primary use for the procedures developed in his paper was said to be the measurement of market power possessed by a market participant using only bid information and market-clearing price and quantities. Professor Wolak referred to the peculiar qualities of electricity markets and observed:

‘… because of the manner in which electricity was sold to final customers during the former vertically integrated regime, the retail demand for electricity is very price inelastic on hour-ahead, day-ahead and even month-ahead time horizons. These features of the electricity production process and the insensitivity of retail demand to wholesale price fluctuations allow small defects in market design to enhance significantly the ability of generation unit owners to exercise market power.’

He added that seemingly innocuous changes in market rules can produce a large impact on market outcomes, so that market design is an extremely important aspect of the ongoing industry restructuring process. I accept the force of that observation which is reflected in the change to the rebidding rules in the Code following the summer of 2000/01.

506               The third paper, Wolak (2003b), referred to in Professor Wolak’s testimony was entitled ‘Measuring Unilateral Market Power in Wholesale Electricity Markets: The California Market, 1998 - 2000’ and was published in the American Economic Review in May 2003 at pp 425-430. In this paper Professor Wolak sought to measure the unilateral incentive of each of the five largest electricity suppliers in California to exercise market power in the State’s wholesale market during four-month periods from 1 June to 30 September 1998, 1999 and 2000. Using actual bids submitted to the California Independent System Operator he computed the hourly price elasticity of the ex-post residual demand curve, faced by each supplier, evaluated at the market clearing price for that hour. He relied upon the inverse of that elasticity to quantify the extent to which the supplier was able to raise the hourly real time energy price above its marginal cost of supplying the last MWh it sold in the real time energy market. The average hourly value of the inverse elasticity over the specified period of each year became ‘… a summary measure of the extent of unilateral market power possessed by each supplier’. The results presented were consistent with the proposition that ‘… the enormous increase in the amount of market power exercised in the California market beginning in June of 2000’ arose because of an increase in the amount of unilateral market power possessed by each of the five large suppliers in California. The California electricity crisis did not necessarily arise from coordinated actions by suppliers to raise prices in California. Professor Wolak’s results were said to be consistent with the price increases in the Californian market being the result of expected profit maximising responses of each of the five suppliers to the bidding behaviour of all other market participants in that market. The term ‘consistent’ figured more than once in Professor Wolak’s evidence. It is properly and carefully chosen. But its limits must be recognised for it does not itself convey interpretive power, albeit it is sometimes used in a context which implies such power.

507               The methodology applied in Wolak (2003b) appears to have been applied directly to Professor Wolak’s assessment of the market power of generators in the NEM for the purpose of these proceedings.

508               Professor Wolak’s first and principal report prepared for these proceedings set out what he called ‘three lines of analysis’ which he performed to assess the competitive impacts of the proposed acquisition. In doing so he adopted the assumption that the acquisition would be passive in the sense that AGL would have no control, direct or indirect, on the day to day dispatch and trading operations and longer term strategic behaviour of Loy Yang Power. While adopting the assumption he stated that he did not believe it likely to be true over the longer term as it would appear to deprive AGL of one of the major benefits of ownership, namely the ability to control how the Loy Yang power generation facilities are bid and operated.

509               The first part of his report sought to quantify the ability of Loy Yang Power, Hazelwood and Yallourn to raise market prices through their unilateral bidding behaviour in the NEM historically and specifically over the period January 1, 2000 to June 30, 2003. He concluded that all three suppliers were able to increase wholesale price substantially through their unilateral actions during a significant fraction of half-hours over that time.

510               The second part of his report sought to describe why three large base load suppliers would have a greater incentive to exploit their unilateral ability to raise wholesale electricity prices if AGL were to proceed with its acquisition. He computed likely counter-factual profit-maximising price increases over the period 1 January 2000 to 30 June 2003 that would have resulted from the level of forward contract obligations that the three base load generators would have held if AGL had owned a 35% stake in Loy Yang Power over that time. According to that analysis he found sizeable, counter-factual price increases for all three suppliers over the sample period. The term ‘counter-factual’ in this context I take to mean involving an assumption contrary to the fact, that in the relevant period AGL held the 35% interest which it now seeks to acquire.

511               The third part of his report sought to validate the magnitude of the counter-factual, unilateral price increase for a reduction in Loy Yang Power’s forward contract position based on the historical relationship between half-hourly price in Victoria and Loy Yang Power’s half-hourly forward contract holdings. According to this analysis Professor Wolak found predicted increases in the spot price as a result of reduction in the forward contract holdings of Loy Yang Power consistent with those found by his counter-factual unilateral price increase.

512               These three lines of analysis led to the overall conclusion that the proposed acquisition would be very likely to lead to a substantial lessening of competition in the wholesale electricity market.

513               In considering the ability of Loy Yang Power, Hazelwood and Yallourn to raise spot electricity prices, Professor Wolak used bid, spot price and dispatch data to calculate what he called ‘… the standard indicator in economics of the ability of a supplier to raise the market price’ namely the elasticity of the residual demand curve facing the supplier. The residual demand curve plots how, after all other suppliers have input supply, a change in price received by a firm correlates with a change in its sales. It embodies the constraints on a firm’s ability to raise prices due to the responses of its competitors and consumers. It is a well-established concept in economic theory and analysis.

514               Professor Wolak’s analysis centred upon an indicator representing the percentage increase in the spot price for a half-hour period of the day that would result from a one percent fall in a given generator’s dispatch during that half hour period. This is the inverse of the elasticity of a generator’s residual demand curve for that half hour. If it is low then a generator supplying less energy to the spot market enjoys very small increases in the price of electricity. Alternatively, if it is large in absolute value then supplying less energy to the spot market causes a large increase in price. It was not contentious that this figure can be a good indicator of a supplier’s power to increase prices by reducing supply. But as Professor Hogan, one of AGL’s expert witnesses, pointed out, the ability to increase spot prices which it measures does not does not necessarily carry with it the ability to maximise profits. As he said, by way of illustration, if one of the Victorian base load generators were to be turned off entirely prices would go up but there would be no benefit to the supplier. The extreme example illustrates a point which was not really in dispute.

515               The electricity market has the special feature that producers can only sell their energy through a bulk transmission network. This requires a centralised clearing mechanism. Suppliers submit to that mechanism, in this case NEMMCO, their willingness to sell electricity over a wide range of possible output levels, anticipating an uncertain demand for electricity. This special feature of the electricity market yields bid and output data from which empirical researchers can construct a supplier’s residual demand curve.

516               The residual demand curve faced by a generator is downward sloping. The generator must have regard to the increment to revenue from a higher price with the savings in cost from a reduced dispatch in formulating its expected profit maximising bidding strategy. The increment to revenue is the marginal revenue which can be calculated from its residual demand curve. The increment to cost is its short run marginal cost. That is the increase in a firm’s production costs as the result of a one unit change in its output. The qualification that the marginal cost is ‘short run’ refers to the fact that the fixed costs of starting a plant or restarting it, if it had been shut down, as well as sunk investment costs will not change over the period of relevance for setting bids, that is to say over a particular half-hour.

517               An important equation for the purpose of Professor Wolak’s first part analysis involves the following variables:

p - the spot price

c_k – generator k’s short run marginal costs

e_k (p) – the elasticity of the residual demand curve facing generator k evaluated at the spot price of p. This is defined by the relationship:

dDR (p)/dp. P/DR(p)

Where DR(p) is the demand for supplier k’s output at price p.

 

According to Professor Wolak the desire of a generator to maximise profits would lead it to bid to achieve spot prices to satisfy the following equation for each half-hour period of each day:

(p – c_k)/p = -1/e_k(p)

The right hand side of that equation is the negative inverse of the elasticity of residual demand. That quantity, -1/e_k(p), is said to measure the extent to which a generator is able to receive a price above its marginal cost during a given half hour. It gives the percentage increase in the spot price that would result from a 1% reduction in a generator’s actual dispatch in that half hour period as a result of the combined responses of other suppliers to a 1% reduction in supplier k’s output. On this theory a value of –1/e_k(p) equal to two implies that a 1% decrease in supplier k’s dispatch would lead to an increase in the spot market price of 2%. In his oral evidence however, Professor Wolak pointed out that the application of the equation to profit maximising behaviour rests upon four assumptions:

1. The supplier can observe its residual demand curve.

2. The supplier can chose its market price.

3. There are no system constraints.

4. There are no forward contracts.

None of these assumptions is true for generators in the NEM. The residual demand curve is seen ex post facto. The market price is fixed by NEMMCO. There are system constraints and there are forward contracts. That these conditions are not met does not deprive the equation of utility. It offers a way of measuring the ability of a supplier to price above its short run marginal cost in a given half-hour. In this context the distinction between ‘ability’ and ‘incentive’ was a feature of the economic debate in the proceedings. The above equation however cannot be directly applied to calculate elasticities in the electricity market. This is because the residual demand curve cannot be differentiated. As appears below, it is a step function not a smooth curve.

518               The quantitative data available from NEMMCO on a half-hourly basis regarding generator bids, dispatch and spot prices provided data enabling a quantitative assessment of the ability of a generator, defined by reference to the Wolak model, to raise spot prices. This can be done by computing the half-hourly value of the negative inverse elasticity of residual demand curve facing a particular supplier. Professor Wolak regarded this measure as a well regarded and accepted method for assessing the ability of a supplier to raise prices in wholesale electricity markets. It has been applied to markets in California, Texas, the NEM1 in Australia and other international markets. The derivation of his equation is set out in Appendix A to Professor Wolak’s first report. Once the appropriate definitions of the variables are put in place the derivation is trivial.

519               Professor Wolak calculated –1/e_k(p) for the three largest suppliers in Victoria namely Loy Yang Power, Hazelwood and Yallourn, for each half-hourly period from 1 January 2000 to 30 June 2003. The figures yielded by his computations were as follows:

	2000	2001	2002	2003
LYP	11.20	8.05	5.61	6.18
Hazelwood	7.04	6.22	4.05	4.56
Yallourn	6.95	5.42	3.51	4.21

 

These estimates were said to indicate that each of the three Victorian generators had a substantial ability to raise the spot price of electricity during all four years of the sample. Although I made a confidentiality order in respect of this table it is an important aspect of Professor Wolak’s evidence and represents the outcome of calculations which he has applied referenced to a number of variables. The historic character of the data upon which the calculations were based and the difficulty of reconstructing precise bidding behaviour by reference to them satisfies me that the confidentiality order should not be maintained.

520               Professor Wolak pointed out that he had made the conservative assumptions that there were no transmission constraints or line losses in the NEM and that effectively every single generator in the NEM was located at the same point in space. On this assumption the calculated values of -1/e_k(p) were likely to be lower than the real values because transmission constraints and transmission losses limit the ability of out of State suppliers to dispatch supplies into Victoria. His analysis also indicated that for certain time periods the apparent ability to raise the spot price of electricity is considerable. The monthly mean inverse elasticity of LYPM’s residual demand in December 2000 was said to be minus 31.82. If this were correct then it meant that for a 1% reduction in its half-hourly dispatch, LYPM would have been able to raise the spot price by almost 32%. On the other hand, the lowest estimate of –1/e_k(p) for LYPM in March 2002 showed that a 1% reduction in its half hourly dispatch would have led to an increase of 2.38% in the spot price. There is no doubt that at certain intervals of supply/demand imbalance there are opportunities for the kind of game playing in which LYP engaged in the summer of 2000/2001 which can lead to substantial price increases for short periods. As already discussed in the context of that summer strategy it does not follow that this reflects an ability to raise prices profitably on a sustained basis.

521               A point of contention between AGL and the ACCC with respect to Professor Wolak’s methodology was the way in which he derived the slopes for the residual demand curve which enabled him to calculate its elasticity at various points. Because demand for electricity in the NEM varies with each half-hour each day, the residual demand curve does not plot a fixed relationship. It is not a smooth curve but a jagged step function whose slopes, at various price points, are not able to be measured by the slope of tangents at those points as with a smooth curve. On a smooth curve, in two dimensions, the slope represents the rate of change of one variable with respect to the other. This is normally derived by the process of differentiation which on a step function may yield values of either zero or infinity.

522               To deal with the step-wise nature of the residual demand curve in the electricity market Professor Wolak adopted what is called the ‘arc elasticity method’ for measuring the price elasticity of residual demand. At each price for which he wished to calculate a slope he took an interval around that price defined by a slightly higher price and a slightly lower price. The change between the low price and the high price defined a change in demand. The change in the price divided by the corresponding change in demand yielded the slope.

523               Professor Wolak applied constant criteria to the selection of the price band width around a given market clearing price in order to avoid bias in the selection of slopes. This reflects the fact that in a rapidly changing residual demand curve the slopes between different pairs of points on the curve may vary radically according to where those points are located. Professor Wolak’s criteria were:

1. The difference between the two prices selected must be economically meaningful – that is to say it must reflect a change in demand.

2. The price increase must imply a physically feasible quantity change meaning a quantity change that the supplier would contemplate implementing within the relevant half-hour.

3. The price difference must be non-manipulable. That is to say the criteria for selection must not be biased by prior consideration of the data with a view to obtaining a particular result. Economically meaningful intervals could be quite small and no more than $1/MWh. This feeds into the requirement that the quantity change reflected by the price difference be physically feasible. A price increase of 10cents/MWh would imply a quantity reduction of 20MW which for Loy Yang with a generating capacity of about 2,000MW would be physically feasible. A price increase of a dollar/MWh would reflect an implied withheld quantity of 200MW. However this is at the extreme of the physically feasible output reduction that Loy Yang could implement in the half-hour. On the other hand, a price difference of $5 would imply withholding of 1,000MW, representing half of Loy Yang’s generating capacity. This would not be physically feasible. Nor would the quantity reduction implied by a price increase of $50/MWh used by Dr Webber to compute some of his elasticities. For that would be equivalent to a quantity reduction of 10,000MW.

The criteria which Professor Wolak applied to the selection of price bands have the virtue of consistency. He also claimed that they avoid asymmetry around the market clearing price. I observe however that symmetry may mean little in a rapidly varying demand curve.

524               Dr Webber, for AGL, on the other hand took slopes across different price bands at different portions of step functions measured across much larger intervals of price and quantity than the step functions relied upon by Professor Wolak. This was illustrated by a diagram which graphically contrasted their approaches. Dr Webber’s approach yielded lower slopes than Professor Wolak’s.

525               There is a question whether Professor Wolak’s ‘slopes’ which can change significantly over small intervals have any relevance to the residual demand curve viewed over more than individual half-hour periods. It is possible to view the step function, as it were, through a half-hour microscope, and see in it much local complexity. However looked at from a greater temporal ‘distance’ and by reference to the best fit smooth curve, it presents a very different picture over a longer time period. In the end I was left uncertain about the interpretive power of Professor Wolak’s slopes.

526               After calculating the negative inverse elasticities, as discussed above, Professor Wolak went on to consider the effect of forward contracts on a supplier’s ability to raise market price by supplying less output. This he described as a very important mitigating factor which causes a supplier to find it unilaterally profit maximising not to exploit the trade off embodied in its residual demand curve.

527               The profit maximising equation referred to earlier applies only to bidding in a spot price market. Where there is a quantity of supplier k’s output covered by swap contracts the demand curve is shifted to the left so that:

DR_k ^c(p) = (DR_k(p)-qc_k)

Here, qc_k is the quantity of output covered by contracts. DR^c_k(p) is the demand for k’s output net of forward contracts at price p. The elasticity of the residual demand curve, net of forward contracts is e_k^c(p). It is defined as:

e_k^c (p) = (dDR_k(p)/dp)(p/[DR_k(p) - qc_k])

Profit maximising behaviour by supplier k in a market in which the quantity qc_k is subject to forward contracts will, according to the theory, result in k bidding to set prices that satisfy:

(p - c_k)/p = -1/e_k^c(p)

This means that for the same residual demand curve a supplier with forward contracts will find it profit maximising to submit bids that result in market prices closer to their marginal cost than would be the case if it did not have forward contracts because –1/e_k^c(p) will be smaller the closer is the supplier’s forward contract position to its actual dispatch quantity.

528               Professor Wolak proceeded on the basis that forward contract cover was generally by way of swap contract which he described as ‘[t]he most popular forward contract traded by generators and retailers on a volume basis…’. The payment structure for swap contracts he described as having important implications for the expected profit maximising bidding strategy of a supplier because it changed the payoffs to the supplier associated with raising the market price. The supplier would effectively earn the spot price only on sales in the spot market above its forward contract quantity. So supplier k would only want to raise the spot market price p is it were to sell more than its swap contact quantity in the spot market.

529               Professor Wolak did not have information about the forward contract positions of the three major base load generator owners in Victoria, although he had been told that all of the suppliers in the NEM has significant forward contract positions negotiated in advance of delivery of the energy in real time. Because their forward swap contract positions could significantly limit the incentive suppliers would have to reduce their output in order to increase market price, knowledge of their forward contract positions was necessary to assess the likely impact on market prices of a reduction in the forward contract obligation of any of the large base load generators in Victoria. He sought to recover estimates of the half-hourly forward contracts positions of the Victorian base load generators by assuming a value for the marginal cost of supplier k and solving the profit maximising equation for the implied value of qc for that half-hour based on the value of the market clearing price, the elasticity of the residual demand curve at the market clearing price and the amount of energy produced by the supplier during that hour. This involved the application of the equation:

qc(implied) = [1+[(p -c_k)/p]e_k(p)]DR_k(p)

The equation for qc(implied) is derived by taking the elasticity of the residual demand curve net of forward contracts which is defined as:

e_k^c(p) = DR _k (p)dr/dp (p/(DR _k (p) – qc))

 

That equation is solved for qc and the term e_k^c(p) substituted by – p /(p - c_k) according to the profit maximising relationship above. The validity of the equation for qc(implied) rested on two assumptions. They are:

 

1. There is an accurate marginal cost estimate c_k for the supplier k.

2. Suppliers try to maximise expected profits given their forward contract positions so that the profit maximising equation net of forward contracts holds at market clearing price.

 

530              Professor Wolak operated on an assumed marginal cost for Loy Yang Power, Hazelwood and Yallourn which he had been given and also experimented with significantly larger values for that estimated marginal cost. Although differences in the estimated marginal cost did affect the half-hourly values of qc(implied) it did not significantly affect the level of the counter-factual profit maximising prices which formed the basis for his conclusions about the competitive consequences of the acquisition. This was the methodology he had previously applied to bids by a large base load supplier in the NEM1 over the period 15 May 1997 to 24 August 1997. He assessed its validity using a range of estimates of the marginal cost of producing electricity from the supplier’s units to compute half-hourly values for qc(implied) using the above equation. These were compared to the actual half-hourly forward contract values for the particular large base load supplier under consideration in the earlier study. That analysis showed a general agreement between half-hourly values of qc(implied) and qc(actual) across a variety of dimensions.

531              Professor Wolak considered the incentives for suppliers to sign forward contracts. He argued that retailers with fixed price retail load obligations have a strong incentive to enter into fixed price swap contracts to hedge their exposure to the spot price in the NEM. Secondly, a supplier takes on significant risk by limiting the magnitude of swap contracts that it sells. If its competitors sell additional forward contracts to make up for the ones it does not sell these competitors will have an incentive to drive the price below their marginal costs of production over a larger range of output from their generation units, because they will be net short relative to their new forward contract position over a wider range of output from those units. Such actions by the firms that sell additional forward swap contracts will lead to lower average wholesale prices so that the supplier that has decided to reduce its forward contract obligations earns a lower spot price for a larger fraction of its output. This provides a strong argument against a supplier being fully unhedged. It is not an argument in favour of a supplier being fully hedged because to do so eliminates its incentive to bid to raise spot prices.

532              Professor Wolak emphasised that ‘suppliers often leave a significant amount of their expected output uncontracted’. By way of example he referred to the position of LYP in the period January through March of 2001. By leaving much output uncontracted LYP was able to exploit the upward sloping residual demand curve that it faced to increase the price it received for its uncontracted quantity of energy. He said it appeared that LYP’s strategy of exploiting a slope of its residual demand curve over this time period was extremely successful. It resulted in an average price for that three month period of slightly less than $60/MWh as compared to the average price for all twelve months of 2001 of approximately $36/MWh. As to that, I interpolate that, having regard to the circumstances in which LYP reduced its contract cover and lifted spot prices in the summer of 2000/2001, I do not accept that its behaviour supports a generalisation about suppliers generally or LYP in the future. The reasons for that conclusion have already been set out in the discussion of the LYP bidding strategy for that period.

533              Professor Wolak then considered the likely change in wholesale market outcomes that would occur in his opinion as a result of AGL’s purchase of a 35% share in the Loy Yang Partners. He identified a number of assumptions underlying his analysis. They were as follows:

1. All retailers are sufficiently risk averse to variations in the spot price to ensure they are fully hedged against that risk.

2. AGL’s purchase of 35% of LYP would give it a ‘natural hedge’ against spot price risk which would result in it reducing its contract holdings.

3. The most straightforward calculation would proceed on the basis that because AGL is purchasing a 35% share in LYP it receives a hedge for 35% of LYP’s currently uncontracted output.

4. The acquiring consortium has modelled its valuation of LYP on the basis of a 75% hedge position. 75% of 2,000MW is 1,500MW. On this basis, and assuming a value of LYP’s average output of 1,900MW, the model yields a hedge quantity of 140MW being 35% of the difference between 1,900MW and 1,500MW. Variations of these values can be obtained by using other values for the LYP average output. According to Professor Wolak it is possible to obtain defensible values of the natural hedge from as high as 245MW to as low as 87.5MW. He accepts the most plausible value as 140MW based on LYP’s desired contracting level and its historical average annual output level.

534               The question which he then addressed was – if AGL acquired a 140MW natural hedge as a result of the acquisition what would it do? His best assessment was that it would reduce its forward holdings by slightly less than the quantity of the natural hedge, in part because the requirement that the acquisition be passive could provide it with the equivalent of a cap contract with an exercise price approximately equal to LYP’s average total cost following the acquisition. AGL’s passive acquisition of a 35% stake in LYP would resemble a cap contract because it would make an upfront payment to become a passive owner and in exchange receive 35% of the profits that LYP earned from selling electricity. Because of its passive ownership, it would be shielded from losses that LYP might earn from selling electricity except to the extent of its equity holdings. To a first approximation, if the average annual price that LYP receives from selling energy exceeds its average total cost, AGL could expect to receive 35% of the revenues in excess of those total costs. This implies an exercise price for this ‘cap’ approximately equal to LYP’s average total cost. Because the acquisition could be thought of as a cap contract with a strike price approximately equal to LYP’s average costs on a per MWh basis the natural hedge would have significantly more value to AGL than a conventional swap contract with the same quantity of MW at a price equal to LYP’s average total cost. Because of the magnitude and value of the natural hedge provided by AGL’s acquisition it should find it unilaterally profitable to reduce its forward contract quantity.

535               According to Professor Wolak, the decrease in the demand for forward contracts by AGL should result in reduced sales opportunities for the supplier that loses those contracts with AGL. On the assumption that all Victorian retailers are currently fully contracted they should only be willing to take on additional hedge contracts at reduced prices because of their inelastic demand curves for swap contracts. This would mean that the generators that lose forward contracts with AGL would be faced with the choice of selling electricity at a low price in the forward market or increasing exposure to the spot market and exploiting the trade off between reducing output and increasing price as quantified in the inverse of elasticity of the residual demand calculations described earlier.

536               Professor Wolak argued that AGL would have an incentive to reduce its forward contract holdings with the supplier most able to increase the profits earned by LYP. That is to say, it would want to reduce its forward contract holding with the supplier able to maximise the ex-post value of the notional cap contract it possesses with LYP as a result of its acquisition. On this logic, AGL would reduce its contract holdings with any of the large suppliers in Victoria able to influence market price through their bidding behaviour. By reducing its forward contracts with LYP, Hazelwood or Yallourn these suppliers would have a greater incentive to exploit their ability to raise wholesale electricity prices as quantified in the inverse of the elasticity of the residual demand calculations.

537               Professor Wolak then assessed the extent to which each of the three large Victorian suppliers could unilaterally raise the wholesale price as the result of a reduction in its current forward contract position. He undertook a counter-factual experiment. For each supplier he computed the difference between the amount of energy dispatched from its generation units during that a specific half-hour and the amount of forward swap contracts obligations that supplier had during that half-hour. In the case of LYP he used its actual forward swap contract quantity, but for the remaining suppliers he used the implied forward contract quantity. He then multiplied the half-hourly difference by the same fixed fraction for all hours during the sample period. That fraction was chosen to yield an average reduction of that supplier’s annual contract quantity at or near the lower end of the expected reduction in AGL’s annual contract quantity if the acquisition were to proceed. He wrote the equation:

DC_kh=a(Q_kh(actual) – qc_kh)

The letter a is a positive number less than one, Q_kh(actual) is the dispatch of supplier k in period h and qc_kc is the actual contract quantity for supplier k in period h for LYP and the implied contract quantity for supplier k in period h for Hazelwood and Yallourn. Given the half hourly values for DC_kh for each supplier, he computed the profit maximising price for supplier k for each hour h of the sample assuming that supplier k’s forward contract quantity during that hour was reduced by DC_kh. The counter-factual unilateral profit maximising price was computed using a simple grid search of prices from 0 to $300/MWh starting at the actual price for that hour. For the half-hours with actual prices above $300/MWh the profit maximising price was set to equal the actual price and no further search was performed. He imposed this constraint to make his computations more conservative for reasons which are set out in his report and which need not be further detailed here.

538               Professor Wolak prepared a table showing the annual average counter-factual profit maximising price p_kh(DC) for a value of a = 0.175 and the annual average value of p_h(actual), the actual price. The value of a was chosen so that the resulting averages of DC_kh for the three suppliers would be both above and below 100MW which he regarded as a conservative lower bound on the likely reduction in AGL’s forward contract holdings as a result of the acquisition of a 35% share of LYP.

539               The resulting table was said to illustrate that for all suppliers the average values of p_kh(DC) were significantly larger than their respective average annual price. The results were said to demonstrate that a unilateral reduction in the average swap contract quantity held by any of the three Victorian suppliers during any of the years from 2000 to 2003 would have caused them to find it unilaterally profit maximising to raise wholesale prices significantly. The lowest percentage rise across the three suppliers for 2001 was 20% for Hazelwood with the largest from LYP at 35%. This range of percentage increases in wholesale prices, according to Professor Wolak, provided strong evidence in favour of the view that there exists a high likelihood that the proposed acquisition will lead to a substantially lessening of wholesale competition in the NEM.

540               Professor Wolak described the price he computed for each supplier to be the ‘unilateral profit-maximising response of that supplier to the contract reduction given by DC, assuming no contract reduction from the remaining suppliers and no change in their corresponding bidding behaviour’. A change in the bidding behaviour of the supplier’s competitors would change the residual demand curve faced by it which would then change its profit maximising price with the lower level of forward contracts. He would expect a response from the supplier’s competitors but said there were several reasons why their response might not be to submit bids closer to their marginal cost curves. If the other suppliers believed that it was unilaterally profitable for them to exploit the steeper slope of their residual demand curve brought about by the less aggressive bidding by the supplier which had experienced a reduction in its forward contract position those price increases could be sustainable for a period of time long enough for substantial harm to competition to occur.

541               Professor Wolak referred to the NEMMCO Statement of Opportunities and its projection of demand growth as providing a basis upon which his apprehended scenario might occur. In so doing however, he accepted the statement by NEMMCO that no new scheduled generating plant capacity had been committed for the ten year outlook period covered by its statement. Under these circumstances he stated that all suppliers could be confident ‘… that there is insufficient available capacity relative to demand so that the strategy of further exploiting the trade off between selling less output and raising prices is preferable to a strategy of selling as much as possible at the prevailing market price’. For reasons which I have already given in relation to consideration of the NEMMCO forward projections, I do not accept that this hypothesis reflects any likely commercial reality. Nor do I accept that it is an hypothesis upon which the Court should act in assessing whether there is any real chance of a substantial lessening of competition in the relevant wholesale market.

542               To assess the validity of his counter-factual price increases arising from a reduction in swap contract quantity, Professor Wolak wanted to construct an economic model of the NEM and compute equilibrium responses to such a reduction in the level of swap contracts held by the large Victorian suppliers. He referred to similar work previously done by Bushnell, Mansur and Sarvia in their 2003 study of the impact of forward contracting levels by suppliers on market prices in electricity markets in the United States. They had modelled those markets under the assumption of quantity-setting Cournot behaviour by the large firms and price-taking behaviour by the smaller firms, and had computed equilibrium outcomes under a variety of levels of forward contracting. However such an analysis for Australia would take far more time than was made available to him to prepare the report. He therefore adopted what he described as a ‘less complex approach to assessing the equilibrium response’ which was possible given the special circumstances in the Victorian market. He observed that the price of basic import fuels used to produce the vast majority of energy sold in Victoria had not materially changed over the past four years. Victoria had experienced moderate demand growth over that period. Nevertheless, average prices during the months from 1 January 2000 to 30 June 2003 had ranged from as high as $94/MWh in February 2001 to as low as $15/MWh in April 2003. This, he said, raised the obvious question of what explained the large movements in prices over the sample period. Input cost changes could not completely explain them. Demand levels were important factors but in light of the discussion of the impact of forward swap contracts on the incentive of suppliers to attempt to exploit the slope of the residual demand curve they face, one explanatory variable for half-hourly differences in prices was the aggregate level of forward contracting by LYP, Hazelwood and Yallourn. Using this logic, Professor Wolak estimated a linear regression predicting the half hourly price in Victoria as a function of level of swap contracts held by LYP during that half hour and controlling for deterministic seasonal factors within the day, week and months of the sample period of 1 January 2000 to 30 June 2003.

543               According to the logic which he had discussed earlier he would expect that in the half hours when qc, LYP swap contract quantity, was low, its incentive to raise the spot price would be the greatest. This implied that after controlling for the level of half-hourly demand and other within-day, within-week and across-year factors, higher prices should be associated with lower levels of qc. Because he only had data on the actual contract position for LYP he could only run the regression with the contract quantity of LYP. He could not perform the analysis for Hazelwood and Yallourn.

544               He set out, in an appendix to his report, a full table of regression coefficients and a complete listing of the variables included in his model. The result of this analysis was as follows:

Regression Predicting Half Hourly Victoria Price as Function of Half-Hourly Swap Contract Position of LYP
Variable	Parameter Estimate	Standard Error	t-Value
LYP half-hourly swap contract quantity	-0.0492	0.0052	-9.43

545               The regression coefficient given in the above table was the predicted increase in the half hourly Victorian price that would result from a 1MW reduction in LYP’s half-hourly quantity of swap contracts sold. It implied that such a reduction would predict a $0.0491 increase in the Victorian spot price. This result implied that for a 100MW reduction in the contract quantity of LYP, well within the range of the expected magnitude of AGL’s forward contract reduction, the average wholesale price in Victoria was predicted to increase by $4.91/MWh which implied a 14% increase in the average price over the entire sample period of 1 January 2000 to 30 June 2003. These regression results, according to Professor Wolak confirmed the reasonableness of the counterfactual price increases he had earlier reported. His conclusions were as follows:

1. Each of LYP, Hazelwood and Yallourn had substantial ability to raise the spot price of electricity during each of 2000, 2001, 2002 and 2003. In 2001, LYP would have been able to raise the spot price by over 8% on average for a single percentage reduction in its half-hourly dispatch, Hazelwood by over 6% and Yallourn by over 5%. Further, in some months the ability of these generators to raise the spot price of electricity is very large. In December 2000 the monthly mean inverse elasticity of LYP’s residual demand was - 31.82. For a single percentage fall in its half-hourly dispatch, it would have been able to raise the spot price by almost 32%.

2. Forward contracting reduces the incentive for a base load generator with the unilateral ability to raise spot prices to exercise that power.

3. The acquisition will confer a natural hedge on AGL.

4. It is possible to obtain defensible values for the natural hedge from as high as 245MW to as low as 87.5MW. The most plausible value appears to be in the range of 140MW as this value of the natural hedge is determined based on LYP’s desired contract level of 75% and its historical average annual output level.

5. Following the acquisition AGL should respond by reducing its level of forward contracts because it will otherwise be over contracted.

6. It will be profit maximising for AGL to affect its reduction in contract holdings by reducing its contract holdings with any of LYP, Hazelwood or Yallourn.

7. As a result of the reduction in its forward contract holdings the Victorian base load generator will have a greater incentive to exercise its ability to raise the spot price.

8. The average annual price in 2001, had the contract holdings of any base load generator been reduced by the amount implied by the natural hedge, would have been significantly greater than they were for the relevant base load generator.

9. The forecast demand growth in Victoria and the absence of committed new scheduled generating plant for the ten-year outlook period suggests the potential for the exercise of power by one or more of the Victorian base load generators to raise the spot price and be profit maximising for a sustainable period as a result of the acquisition.

10. The analytical results do not provide an estimate of the ultimate equilibrium impact of the reduction in forward contract holdings but regression analysis, to estimate the relationship between Victorian half-hourly spot price and LYP’s half-hourly swap contract holdings, suggests the equilibrium impact of the acquisition would be a substantially lessening of competition in the wholesale electricity market in the NEM if LYP’s contract holdings were reduced because of AGL’s response to the natural hedge. If LYP’s swap contract holdings were reduced by 140MW, the regression predicts an equilibrium price increase of almost $7/MWh, ie approximately 20% of the average Victorian spot price over the sample period.

546               On these bases Professor Wolak expressed the opinion that the acquisition by AGL of the 35% passive interest in LYP was likely to have the effect of substantially lessening competition in the wholesale market for electricity in the NEM. Professor Wolak’s evidence was supported in its qualitative aspects by Professor Stephen King, the Professor of Management (Economics) at the Melbourne Business School at the University of Melbourne. Professor King’s evidence was largely qualitative in character and relied upon Professor Wolak’s quantitative analysis.

547               Professor King considered the effect of the proposed acquisition on AGL’s behaviour and said that AGL would alter its behaviour because the risk facing it would have altered. It would be likely to reduce its contract cover and this would in turn affect the behaviour of Loy Yang and other generators in their sales of electricity derivative contracts and their bidding behaviours for electricity in the NEM. This aspect of Professor King’s evidence essentially revisits the arguments about the effects of the natural hedge covered by Professor Wolak.

548               It is sufficient for present purposes to set out the essential conclusions offered by Professor King on his qualitative analysis which he illustrated graphically by reference to the behaviour of generators facing residual demand curves under various conditions of contract cover. His propositions were as follows:

1. The partial acquisition of LYP by AGL creates a ‘natural hedge’ for AGL.

2. This natural hedge is at least 150MW (this differs from the assessment by Professor Wolak but each of the witnesses was selecting a conservative figure from a range which they had assessed by reference to different assumptions about LYP’s unhedged capacity).

3. As a result of this natural hedge AGL will reduce its demand for arms length electricity derivative contracts.

4. This reduction in demand will be equal to the amount of the natural hedge.

5. The demand for these contracts by retailers is relatively insensitive so that the reduction in demand by AGL will most likely lead to an equivalent reduction in the volume of hedge contracts traded.

6. This will mean that generators in the NEM have increased unhedged generation capacity.

7. Because the natural hedge is with LYP and hedge contracts with different counter-party generators are imperfect substitutes, the most likely generators to have a reduction in hedged capacity are LYP and possibly other Victorian brown-coal, base load generators.

8. An increase in unhedged generation capacity creates an incentive for the relevant generators to try to raise the spot price of electricity.

9. The empirical results of Professor Wolak show that the brown-coal, base load generators in Victoria have the ability to influence and raise the spot price of electricity.

10. It is therefore likely that the reduction in the volume of hedge contracts as a result of the acquisition will lead to an increase in the average wholesale spot price of electricity in the NEM.

Through these steps Professor King concluded that the predicted increase in the average spot price of electricity would be evidence that the acquisition of 35% of LYP by AGL was likely to lead to a substantial lessening of competition in the NEM. He accepted that for an acquisition to lead to a substantial lessening of competition it must normally be the case that the lessening of competition is not transitory short-lived. To the extent that lessening of competition would be reflected in a significant increase in the spot price of electricity the lessening of competition would be substantial only if the increase in the spot price of electricity was not transitory and short-lived. That increase would be short-lived if entry by new generation facilities quickly eroded the ability of LYP or other generators to raise the spot price. In such a situation it would be unlikely that there would be a substantial lessening of competition.

549               In Professor King’s opinion the ability of entry by new generation facilities to offset any lessening of competition was limited. Supply side changes in base load electricity generation and supply were likely to be relatively slow and he relied upon Mr Ergas’ evidence in this respect. He therefore considered that the ability of LYP and other relevant generators to influence and raise the spot price was unlikely to be constrained by new competitive generation in the short to medium term. On this basis he considered that the acquisition was likely to lead to a substantially lessening of competition in relation to the provision of wholesale electricity.

550               He also argued that the acquisition would create incentives for other retailers and generators to merge and therefore could lead to increased vertical integration in the future. To the degree that increased vertical integration would reduce the trade in hedge contracts, a new retailer might find it necessary to enter both retailing and generation simultaneously in order to gain protection from spot price volatility. This was the burden of Dr Small’s evidence which is dealt with separately in these reasons.

551               In considering Professor Wolak’s conclusions, based on residual demand analysis, that Victorian base load generators have a substantial degree of market power, Mr Ergas for AGL said that although that methodology has its merits it raised a number of complex issues. He argued that Professor Wolak had not given enough attention to those issues. He referred to the origins of residual demand analysis in a paper published in 1988 – Baker, JB and Bresnahan, TF, ‘Estimating the Residual Demand Curve Facing a Single Firm’ (1988) 6 International Journal of Industrial Organisation at p 286 where the authors remarked:

‘This notion is familiar in the case of a dominant firm with a price-taking fringe.’

A similar observation about the assumptions underpinning the use of residual demand elasticities was made by Kahai, SK and Kaserman, DL in ‘Is the “Dominant Firm” Dominant? An Empirical Analysis of AT&T Market Power’ (1996) 39 The Journal of Law and Economics pp 499-517. In a review article published in 1991, Froeb and Werden drew attention to the assumptions that a modeller had to make to correctly infer market power from a residual demand analysis. After referring to the risks of the method they said:

‘The moral of all of the foregoing is that residual demand elasticities are relevant, but they do not tell anywhere near the whole story. We cannot calculate an implied monopoly mark-up from a residual demand elasticity without making several assumptions, and the predicted mark up is highly sensitive to these assumptions. The easiest case to deal with is that in which it is reasonable to suppose that a hypothetical monopolist would act as a dominant firm and to treat marginal cost as constant, and even in that case, we need to assume that the residual demand elasticity is constant (or varies in a particular way) in order to calculate a monopoly mark-up. Thus, in the best of circumstances, an accurate estimate of the residual demand elasticity can yield only a fairly rough estimate of the monopoly mark-up.’

Froeb, LM and Werden, GJ ‘Residual Demand Estimation for Market Delineation: Complications and Limitations’ (1991) 6 Review of the Industrial Organisation at p 38

 

552               Mr Ergas concluded that the method applied by Professor Wolak is designed to estimate elasticities for a situation where the firm that has or may have a substantial degree of market power faces one or more other firms that themselves lack market power. The situation considered by Professor Wolak in the electricity markets in Victoria involve at least three firms with a substantial degree of market power. To estimate the elasticities of demand that they face, according to Mr Ergas, he would have to explicitly model their interdependence. Instead he models the firms as if they were unaware of the interdependence between them. Mr Ergas did not believe that to be a plausible assumption. He believed the use of a model only appropriate for one set of circumstances to model another is incorrect as a matter of economic methodology and must lead to doubt about its results. He argued that there is a range of indicators relevant in addressing the question whether LYP has a substantial degree of market power and that they are not consistent with that characterisation.

553               He did not believe market power could be inferred from occasional price spikes and did not accept that it was sufficient to demonstrate that a firm could durably set price above marginal cost in order to find that the firm has a substantial degree of market power. Where there are substantial fixed costs, pricing above marginal costs may be necessary for firms to recoup the entirety of their costs. Although Professor Wolak’s results amounted to a finding that the base load generators in Victoria could set price at more than marginal cost he did not show that they could durably price above average cost and hence earn economic profits. Although he had agreed that once a generator has entered into a swap contract its incentives would change there were two points to be made about this. First, even a contracted generator has incentives to increase pool prices if by so doing it can increase contract prices in future periods. So if generators had a substantial degree of market power one would expect this to be evidence in their actual pricing behaviour. The second point was that if generators, by entering into contracts, are foregoing material profits as Professors Wolak and King suggested, it would be necessary to explain why they were doing so. The explanation given by Professors Wolak and King was that it is competition in the contract market that forces the generators to enter into the contracts. They said that generators collectively and individually would prefer not to contract but each would be worse off if it does not enter into contracts while its rivals do than it would be if it entered into a contract. This was the so-called ‘prisoners dilemma’ to which much reference was made in the course of the hearing. Its essence is that even though all the players in the game have selected the outcome which is collectively inferior for them no player has an incentive to unilaterally change its behaviour independently of a concerted change by the other players. For there to be a dilemma explaining why generators enter into contracts that reduce their prices and profits it would have to be true, according to Mr Ergas, that once each generator had signed contracts no generator could increase its own profits by unilaterally reducing the extent of its contract cover. This explanation however was not open to Professors Wolak and King because if Professor Wolak’s model were correct it should estimate the best expectation that each generator holds about what would happen if it slightly reduced its contract cover taking full account of how other generators will respond. If it did not do this, then the model does not in fact estimate each generators residual demand elasticity. According to Mr Ergas, if Professor Wolak’s model were correct, its clear result would be that each of the base load generators could very materially increase its profits by unilaterally reducing its contract cover. Most economists believe that profit maximising firms will not persistently ignore opportunities to substantial increase their profits. Professor Wolak’s results implied generators were passing up such opportunities and must, in Mr Ergas’ opinion, cast further doubt on their plausibility. He did not believe that generators in fact have the very great degree of unilateral market power that Professor Wolak estimated, nor that they had entered into contracts which induced them to behave ‘as if’ they did not have such a degree of power and so did not translate it into monopoly profits.

554               Mr Ergas addressed the argument that the proposed transaction would create a natural hedge leading to reduction of AGL’s contract cover resulting in generators in the NEM having increased unhedged generation capacity reflecting the amount of AGL’s natural hedge. He was of the view that vertical integration by a retailer can provide a natural hedge against pool price volatility. That was a view accepted by ACL in its submissions. AGL expressly accepted that the proposed transaction may give rise to a natural earnings hedge.

555               Mr Ergas did not accept that the extent of the hedge acquired by a retailer has a simple relationship to the extent of its ownership stake in a generator, especially one that it does not dispatch. He did not accept Professor King’s argument that a stake in a generator is equivalent to having purchased a price cap contract. Unlike a price cap contract, an ownership stake exposes the retailer to both upside and downside risk as a fall in price below the generator’s costs causes losses in the value of the generator’s equity stake. And although a price cap contract is firm the natural hedge provided by integration is not. A plant outage could mean the retailer would be uncovered by the vertical ownership hedge. Under a price cap contract the risk of outage would usually be borne by the generator. On this basis the purchase of a 100MW cap contract at a given price would confer far more protection on a risk averse retailer than the equivalent purchase of a passive ownership stake in 100MW of generation. Three conclusions flowed from this:

1. Professors Wolak and King had not established that protection provided by the natural hedge was equivalent to that provided by an equal amount of capacity under contract.

2. They had not therefore established that the extent of that protection resulting from the purchase of a passive equity stake in Loy Yang was such as to materially alter AGL’s optimal level of contract cover; and

3. Any contention that implies that AGL will reduce the extent of its contract cover to a material or even discernable extent was not established from an economic perspective.

556               Mr Fraser in his second affidavit offered a commercial perspective. He said that he had never previously encountered or contemplated the idea that a passive equity interest of 35% held by a retailer in a generator would lead to de-contracting by the retailer. Subject to the AGL risk management policy and the direction of the Board, he and his Energy Trading Team at AGL are responsible for managing pool price exposure. He had no plan and to his knowledge there was no plan within AGL to reduce contract cover if they acquire a 35% equity interest in LYP. Based upon his knowledge of AGL’s business and the electricity industry he could not envisage it having a lower level of contract cover in Victoria by reason of that acquisition because to do so would involve increased risks. He referred to the effect of the acquisition on AGL’s contract cover in the first quarter of 2004. It would not bear upon the quantum of that cover. On the basis of demand forecast by the AGL Structure Desk for the first quarter a contract portfolio to manage pool price exposure has already been established. In the ordinary course demand forecasts and contract portfolios will be continuously reviewed until the point immediately preceding the relevant dispatch interval. AGL has currently forecast the time of its maximum demand during the first quarter, although its accuracy will only be known after the first quarter has ended. It has a current forward contract position for the relevant period which could change at any point up to the time of maximum demand depending upon changes in expected demand. At this point based on current demand forecast that is the level of contract cover AGL intends to hold. AGL has decided to take some level of exposure to spot price because the cost of purchasing contracts to mitigate the spot price risk would exceed the cost of remaining exposed to the spot price even if set at VoLL. This is what Mr Fraser called the economic level of exposure. By this he means an amount that will not cause AGL to incur significantly greater costs in the event that the worst outcome in the spot market is realised, that is VoLL occurs for all periods during which AGL has a volume of exposure to the spot market. If AGL were to de-contract to the extent of the alleged natural hedge throughout the first quarter, the level of its exposure would exceed what he would consider to be a prudent level. If during one or more of the settlement periods in which AGL was exposed to the pool price Victoria experienced an extended period of prices at VoLL, AGL could be exposed to significantly increased costs to the extent that it was unhedged. If such an event occurred even for a few hours on the day of AGL’s highest customer demand, assuming its forecast is correct, the costs could run into millions of dollars. AGL would not be indifferent to that result on the basis of its natural hedge. For a start, AGL would not be sharing in the revenues of LYP. As a shareholder in GEAC, it would expect to receive a share in the profits as opposed to the revenues. Mr Fraser said:

‘I could not know that any risk incurred by AGL being under-contracted would be matched by payments made by Loy Yang Power to AGL in its capacity as a shareholder, because AGL cannot know, or reliably predict, that Loy Yang Power would be earning pool revenues equal to or exceeding AGL’s losses during any particular settlement period(s).’

The outcome for AGL would depend on the actual level of cover for LYP in any half-hour period and the volume for which it was dispatched, which would be a function of its bidding behaviour. Any knowledge of average or desired contract cover would be inadequate to provide assurance to AGL that its pool price losses would be matched by suggested shareholder revenues.

557               On the assumption that LYP contracts to only 75% of capacity in all periods during the period of highest system demand in Victoria, it would have 500MW exposed to the spot price when generating at near full capacity. With spot price at VoLL for five hours LYP would earn $25 million from the 500MW exposed to the pool. A 35% share of that revenue would be no more than $8.7 million. Mr Fraser gave confidential evidence to indicate that AGL uncontracted would still have a shortfall. He also made the point that to match AGL’s contract position to that of LYP there is a number of variables AGL would need to know or be able accurately to predict in advance of changing its contracting behaviour to be confident that it could de-contract but retain income protection from LYP during periods of exposure to the spot price. It would need to know that LYP would not be contracted for more than 75% of its registered capacity particularly during peak demand periods. It would need to know that LYP would not suffer outages or forced reductions in output at times of peak demand and that it would be dispatched for a quantity of its net contractual position plus 25% of its registered capacity for each settlement period at such times. The level of LYP’s contract will not be known by AGL under current GEAC shareholders’ arrangements. The other two variables, according to Mr Fraser, are demonstrably unpredictable and unable to be estimated with any accuracy on a half-hourly basis by either AGL or LYP. Having regard to the risks to AGL of being under-contracted by an additional 140MW, Mr Fraser would not direct his trading staff to engage in the strategy which it is suggested that they would undertake.

558               Although Mr Fraser was a witness who was in a sense an advocate in his own cause, his evidence on this issue had a ring of commercial reality about it. It certainly seemed to reflect the untidy realities of decision-making in this marketplace and the pressures in favour of risk management which would not be significantly mitigated by the natural hedge.

559               Mr Ergas argued that his conclusions and those of Mr Fraser cast doubt on the contentions that the acquisition would result in a reduction in the volume of hedges written by generators. He pointed to other assumptions in relation to that inference which he said were incorrect. There were as follows:

1. That AGL, having reduced its contract cover, the generator less hedged would be unable to increase the volume of hedges it sold to third parties. This amounted to an assumption that the demand for hedging cover was highly inelastic relative to supply. This assumption was unsupported by evidence or theory.

2. Secondly, Professor King argued that there was some differentiation in the market for cover and that some forms of cover provided less protection than others because they involved greater basis risk. If that were accepted, according to Mr Ergas, one would expect contracts written at the Victorian Node to be especially attractive to Victorian retailers as involving less price risk exposure.

Another point made by Mr Ergas was that neither Professor Wolak nor Professor King discussed in any detail the determination of prices in the market for contract cover. They advanced no theory nor empirical model to explain how those prices are determined, although the workings of the market for contract cover were fundamental to their argument. It would seem strange if generators with the very great market power they say the Victorian base loaders have did not significantly mark up prices for contract cover. He also made the point, borne out by a large empirical literature, that those who finance investment in generation seek and demand from generators a high level of hedge cover to protect providers of debt.

560               Mr Ergas concluded that the evidence did not provide reasonable grounds for concluding from an economic perspective that the acquisition would result in an increase in the volume of unhedged base load generation in Victoria. Even if it did, he rejected the argument that there would be an incentive and ability for AGL and generators to increase spot and contract prices. Even under current conditions it could not be assumed that generators are extremely myopic and take no account of the impact of current pool prices on the prices at which they will sell contracts in future. No such assumption could be sustained in a more realistic portrayal of generator behaviour. Even with some impact of contract cover on generator behaviour it was difficult to believe that a very small change in the extent of cover could induce a very substantial and durable increase in pool prices. The basis for such a claimed large effect was the econometric analysis set out by Professor Wolak. This was a model devised for one set of circumstances applied to a completely different set of circumstances. In Mr Ergas’ view it was quite inappropriate to rely on the results of that analysis for predictive purposes. If Professor Wolak’s estimates were correct, each of the base load generators would have an incentive to unilaterally reduce its contract cover to low levels. The absence of any such fall cast doubt on the reasonableness of using those estimates to predict outcomes were the transaction to proceed. These concerns were heightened by the fact that the market frequently experiences changes in contract cover of the magnitude that it is claimed will arise from the transaction. If changes of this magnitude in contract cover could result in large and sustained price increases, it would seem reasonable to expect that base load generators would have discovered this and acted upon it.

561               AGL submitted, having regard to the evidence of Dr Price, Mr Ergas and Mr Fraser that, if the price of hedge contracts were reduced, retailers may seek to win more customers and to obtain further hedge products at a cheaper price to cover the increased demand for electricity. Dr Hieronymus said that even if retailers were fully hedged their hedge position would consist of a combination of swaps and other products such as caps. If the price of swaps decreased, retailers might choose to purchase more swap contracts than the inferior hedge product represented by caps. On this basis it was contended that the demand for swap contracts is more elastic than claimed by Professor Wolak.

562               In any event there was evidence from Dr Price that the demand for electricity in Victoria is growing. Mr Denton described this growth as outstripping the growth in supply so that the supply/demand balance is tightening. On that basis the demand for contracts is rising relative to supply. Therefore the claimed reduction in demand for contracts consequent on the transaction would be capable of being absorbed simply by natural growth.

563               Reference should be made also at this point briefly to the evidence of Mr Smart, a witness called by AGL who is a principal and Executive Director of the Network Economics Consulting Group Pty Ltd. Mr Smart tested Professor Wolak’s assertion that when LYP swap contract quantity is low its incentive to raise the spot price is the greatest. He referred to an instance in 2003 in which a non-transient reduction in the level of LYP’s contract cover took place. He examined pool prices before and after the change. In his witness statement he demonstrated that his comparison showed that pool prices did not increase when LYP’s level of contract cover was reduced to 88MW. Instead pool prices actually decreased. Seasonal effects could not explain the observed price decrease for two reasons. First, the before and after time periods were relatively short and adjacent. Secondly, to the extent that seasonal effects are reflected in changes to overall demand levels on the NEM, Mr Smart had adjusted for them by finding an empirical relationship between price and demand separately for the before and after periods and comparing the price demand relationships themselves.

564               In my opinion, it cannot be said that it is likely that the proposed acquisition will lead to a thinning of the hedge contract market and the pricing consequences for which the ACCC contends. To the extent that that contention rests upon Professor Wolak’s analysis which I have outlined, that analysis does not provide what I would regard as a reliable guide to a determination one way or the other in these proceedings.

565               I am not persuaded that the slope selection and the calculation of negative inverse elasticities of demand of the residual supply curve undertaken by Professor Wolak can be translated to the realities of bidding and pricing behaviour in the relevant wholesale market. The complex slope curve selected by Professor Wolak gives one picture at a resolution of half-hour intervals and another over greater time periods. The criteria for selecting price band widths to measure the ‘slopes’ at various price points on the curve no doubt ensure internal consistency. I am not persuaded that they give weight to the interpretive power of the slopes so measured and the elasticities calculated by reference to them in terms of commercial behaviour over an extended period. To the extent that Professor Wolak relied, albeit by way of example, upon the LYP bidding behaviour for summer 2000/2001, he was pointing to a risky strategy induced by financial pressures and supported in the event by a constellation of fortuitous circumstances.

566               Professor Wolak’s model did not take any account of the supply side response to a contraction in the demand for forward contracts. He said in cross-examination that there is a potential supply response and to him that would be the real issue. He said it was the fundamental issue:

‘Q. And your model does not model the potential supply response?

A. There is no model that I’m aware of that does that.’

 

AGL contended that, because Professor Wolak’s model leaves out of account the potential supply response to any unilateral change by LYP or any other generator, it is not capable of being used to test the likely outcome in the real world of the behaviour he prospectively attributes to LYP. I agree with that submission. It is not in the end a criticism of the model which seeks to present one aspect of behaviour by a market participant.

567               Professor Wolak also relied upon the NEMMCO Statement of Opportunities to assume that all suppliers would be confident that there would be insufficient available capacity related to demand. For reasons already given, I do not accept the hypothesis about the future generation capacity that underpins that conclusion. Importantly, I consider that the evidence given by Professors Wolak and King does not do justice to what Professor Wolak has referred to as the very high dimensional strategy space in which generator decision-makers on bidding and pricing and contract cover operate. The complexities of that ‘space’ were well demonstrated by Mr Fraser in his evidence on the consequences of the natural hedge to be gained by reason of the acquisition. It is of course always possible that the scenario envisaged by Professors Wolak and King and the ACCC could occur. In my opinion, however, that is a possibility which does not amount to likelihood at a level which could attract the prohibition for which s 50 provides.

568               I am fortified in the preceding conclusions by the evidence of industry witnesses. Mr Bonwick of Power Direct said that his company currently sources a variety of hedge products including peak and non-peak swaps from a variety of generators and through broker-facilitated transactions. Typically, Power Direct will purchase simple hedging products through a broker and will seek individually negotiated terms for more complex products. On the basis of his experience, which included time as the Director of Sales and Marketing at Yallourn Energy Pty Ltd and previously Mercury Energy Ltd in New Zealand, he had formed the view that retailers adopt a wide variety of strategies to hedge their load. This includes the routinely buying flat swaps and also peak and off-peak swaps, weekend peaked swaps, shaped swaps, caps, collars, floors and also European and American options on most or all of these products and Asian swaps together with more highly structured products. Base load generators are capable of selling most or all of the above products as are intermediate and peak generators and importantly retailers. It was his opinion that given the current contracting practices of Power Direct and the availability of hedge contract cover in Victoria, he did not believe that the acquisition of a partial interest by AGL in LYP would adversely affect the ability of Power Direct to obtain hedge contract cover for its Victorian load, either currently or as it expands in the future.

569               Mr Trevor Lee, Group Manager, Regulatory Affairs of Energex Ltd, a Queensland Government owned corporation, which is a retailer and distributor of electricity, said that under its risk management policy it enters into hedge contracts with NEM participants which substantially cover its obligations to settle its electricity sales at NEM pool prices. Its purchase costs of electricity are largely determined by its hedge contract position. Most of its portfolio consists of swap contracts. The company commenced electricity retailing operations in Victoria in 1998. These are principally directed to commercial and industrial customers as opposed to residential and small business customers. The company competes with a number of other retailers in Victoria. In order to win retail customers it has to offer prices on a very low margin over the costs at which it acquires electricity. The reason it has not expanded its electricity retail operations in Victoria to supply residential and small business customers is principally because the regulation of retail tariffs in Victoria is such that there is little scope to offer competitive services to that sector. Energex has not had any difficulty in obtaining hedge cover for its New South Wales and Victorian customer load. There is a number of electricity generators in Victoria which have an insufficient credit rating for its purposes including LYP and for that reason it does not deal directly with LYP as a counter-party. Nevertheless it has not had difficulty in obtaining hedging coverage from other counter-parties. Mr Lee said he has considered the proposed acquisition and the resultant ownership structure of LYP. Based on that consideration Energex may be more prepared to deal with the generator following the proposed acquisition as it can be expected to substantially improve LYP’s capital position and credit worthiness.

The Effect of the Acquisition Upon Vertical Integration in the Relevant Markets

570               The ACCC contended that the acquisition by AGL of an interest in Loy Yang Power would encourage what was described variously as ‘cascading vertical integration’ and a ‘band-wagon effect’ which would have anti-competitive effects. It relied, particularly on this issue, upon the evidence of Dr John Small who is the Head of the Department of Economics at the University of Auckland and Director of the Centre for Research in Network, Economics and Communications at that university. Dr Small was asked by the ACCC to give evidence about the likely effects on competition of vertical integration in the electricity industry with specific reference to the New Zealand experience.

571               Dr Small referred in a footnote, to his statement, to a paper by Ordover, Saloner and Salop, ‘Equilibrium Vertical Foreclosure’ published in the March 1990 edition of the American Economic Review at pp 127-142. He described the paper as a ‘classic contribution’. The paper set out a number of well known criticisms of the theory that vertical integration is likely to give rise to vertical foreclosure, ‘the exclusion that results when unintegrated downstream rivals are foreclosed from the input supplies controlled by the firm that integrates’. Vertical foreclosure may also be said to occur when unintegrated upstream competitors are foreclosed from selling to the downstream division of the integrated firm.

572               The authors of the Ordover paper formulated what they described in the abstract to the paper as ‘… a complete, but analytically simple, equilibrium model of vertical mergers to evaluate the logic of standard vertical foreclosure claims and the criticisms made of those claims’. Their principal conclusion was that anti-competitive foreclosure arose as an equilibrium phenomenon in a coherent model where sophisticated firms use a wide range of strategies and counter-strategies. On their analysis the central condition for successful foreclosure was that the unintegrated upstream firm’s gain should exceed the downstream firm’s loss. This was a condition that could be satisfied in a model with differentiated products and price setting because the sum of the profits of the foreclosed firm and its supply increase if the foreclosed firm’s price rises. They said, however:

‘In a quantity-setting, homogeneous goods model, by contrast, the sum of those profits decrease if the foreclosed firm’s output decreases and the rival’s adjusts optimally.’

573               Directly relevantly to the present debate, they also observed in the last paragraph of the paper:

‘The controversy surrounding the vertical foreclosure argument has traditionally been cast in the context of vertical integration. In order to address the questions raised in that debate we have, for the most part, kept to that formulation. Yet, as modern contract theory emphasises, vertical integration is but one particular form of vertical contract. Indeed, as we have shown, there may exist incentives for firms to structure vertical relationships in ways that fall short of full vertical integration if they are so able. This richer formulation raises a number of interesting issues that we have only begun to explore.’

574               Dr Small explained his approach by reference to a simple ‘informal’ model of the electricity industry assuming a market with two generators and two retailers. If one generator and one retailer were to merge this would leave the second generator and retailer as the only participants in the spot market because the first generator could supply internally all demand states of the first retailer. It would have no incentive to supply any excess capacity to the spot market. It would also have no incentive to offer the second retailer capacity through a forward contract. This is because supplying the second retailer via forward contracts would reduce the input price for that retailer and so increase the capacity that the second retailer could use in the retail market against the first. The existence of capacity constraints allows for input price and input quantity foreclosure.

575               Instead of practising customer foreclosure the merged entity would have an incentive to secure as great a share of the second generator’s generation as possible. This would reduce the potential for aggressive retailing by the second retailer and would enable the first to gain market share on its own. In other words the vertically integrated firm would have an excellent chance of increasing its own retail market share. By increasing its rival’s capacity cost, it would reduce the second retailer’s incentives to invest in retail acquisition which makes retail expansion easier for the merged entities own retail unit. On this basis a vertical merger could be understood as a means of acquiring retail market share at a lower cost than competing head-to-head with a rival retailer. Dr Small also contended that a change in market structure through a vertical merger would have an impact on the profitability of entry in the industry both at the generation and the retail level. The vertically integrated firm’s foreclosure of rivals could lead to the entrenchment of market power by raising barriers to entry and so, in effect, force new entrants to enter both at the retail and generation level simultaneously.

576               The electricity industry is, in Dr Small’s opinion, most exposed to the risk of tacit collusion between firms to support prices above marginal cost given that in many countries prices are set at half hour intervals. The logic of such collusion requires that firms agree to high prices and enforce them by threat of punishing deviation by a rival through future price wars. Vertical integration increases the risk to separate retailers and generators of being foreclosed and so their exposure to an additional punishment device for deviation from tacit collusion in price support.

577               Dr Small observed that the impact of a vertical merger should not only take into account the short-term effect, but also look at the medium and long term. It was in this context that he addressed the issue of consequential vertical integration. Changes in vertical ownership structure affect the incentives and profitability of other potential acquisitions in the industry. If a merger between a generator and retailer proceeds it will increase the wholesale price and decrease the capacity availability of the independent retailer. This will lead to loss of market share and fall in profits in the medium term. The only way for the second retailer to save its competitiveness will be to gain direct access to generation and thereby eliminate the negative impact of the merger on its wholesale price and capacity. On this basis, Dr Small proposed that it is most likely that a vertical merger between a generator and a retailer will induce a counter-balancing merger between the remaining generator and retailer in a simple market defined by two generators and two retailers. He said:

‘More generally, non-integrated firms that are threatened with foreclosure by the merging parties will have an incentive to counter-mergers by vertically integrating themselves. This could give rise to a merger wave or cascade during which the industry witnesses a swift transition from a completely disintegrated industry structure to a market structure with competing vertical chains. The transition should indeed be swift because the non-integrated retailer risks losing important retail market shares.

 

As a consequence the relevant comparison is no longer between a vertically disintegrated structure and a structure with one vertically integrated firm but the comparison between complete vertical integration and no vertical integration.’

 

578               In a market with competing vertical chains the retailing unit would have complete discretion over its capacity to the available capacity of the generator. Leaving one generator’s capacity unused would be less costly because the competing chain would be unable to buy it and gain market share. Each competing chain would therefore withhold capacity from the market and provide the risk-adjusted capacity it requires to serve its customers. This would result in higher retail prices than under vertical disintegration. A more serious consequence of going from a vertically disintegrated structure to one with vertical integration is that the vertical ownership would cement market shares in the long run. Barriers to entry are increased with one integrated firm but this still leaves some scope for a retail entrant to bid for capacity with the remaining generators. With competing vertical chains, retail entry would be impossible and two-tier entry the only and costly option.

579               Against this theoretical background, Dr Small described the historical development of the New Zealand electricity industry. Generation and retail functions are highly integrated in New Zealand so that the observation of competitive processes in that country could reveal effects of vertical integration that might otherwise be less obvious. The experience of AGL there further highlighted several features of it. He set out the history of the privatisation of the industry and in particular the enactment of the Electricity Industry Reform Act in 1998 which split the Electricity Corporation of New Zealand Ltd (ECNZ), a government owned body, into three competing State owned enterprises. It also disaggregated ownership of lines and transmission from generation and retail. Retail and line businesses had to be separated. Because of this all of the distribution/retail companies sold their retail businesses. In most cases they were sold to existing generators. The detail of that sequence of events is set out in Dr Small’s report and it is not necessary to repeat it here.

580               The last major stand-alone retailer of electricity was NGC. In 2001, NGC launched OnEnergy as its new energy retail brand. But within a few months this brand had been withdrawn from the market and all of the retail electricity customers transferred to the large generators, Meridian and Genesis. Dr Small set out a table summarising the events, which he said, caused vertical integration in New Zealand. He described as striking features of the table the large number of transactions that occurred and the speed with which most of the acquisition activity took place. Taken together, those facts strongly suggested that the generators viewed the purchase of retail business as a commercial imperative and that this common view led to a cascade of transactions.

581               The structure of the New Zealand electricity industry as at December 2002 was as follows:

. 7 vertically integrated generator-retailers controlling 93% of generation capacity and almost 100% of customers;

. independent generators (with no retail operations) controlling 7% of generation capacity;

. a single State-owned high voltage transmission network operator (Transpower)

. 29 local (low voltage) regulated monopoly lines companies.

582               Since December 2002 there has been, according to Dr Small, a clear trend towards even greater balance between generation and retail in the New Zealand industry. This was evidenced by the conduct of Trustpower, the firm with the largest retail imbalance as at December 2002. Since that time that firm has been implementing a deliberate strategy of contracting its retail business. It announced on 30 May 2003 its decision to withdraw from the Wellington and Christchurch residential markets and attributed its decision to ‘ongoing difficulties competing against dominant retailers/generators’. In June 2003, it sold 8,500 customers to Mercury Energy completing a process which it had commenced in March 2002 when ‘… Trustpower and Mercury both signalled their intention to exit markets where the two companies believed it would be increasingly difficult to acquire competitive scale’. Dr Small saw these statements, while not referring explicitly to the contract market, as lending empirical weight to his theoretical arguments. To remain active in the Wellington and Christchurch retail markets Trustpower would need to secure contract cover from its vertically integrated competitors.

583               Hedge market information published by the New Zealand equivalent to the NEM, known as the New Zealand Electricity Market (NZEM), was apparently of insufficient reliability and relevance to assist Dr Small’s conclusions one way or the other. The spot price of electricity in New Zealand, as in other jurisdictions, is volatile. He observed that since 1997 there have been four full ‘normal’ years and one drought year, which was 2001. Spot prices increased dramatically during the winter of 2001 as unusually low rainfall created a fear that hydro generators would be unable to supply energy to the market to back their contract position. He said the average spot price for 2003 would also be high for the same reason.

584               Dr Small then presented evidence on the competitive effects of the vertically integrated structure of the New Zealand industry. He referred to published spot prices at each of three Reference Nodes and compiled data sets comprising the fixed and variable retail prices facing a typical residential consumer in each of 44 locations at dates in August 1999 and the annual average prices for 1998 and 1999. He also compiled a matching series of average wholesale spot prices at each of three Reference Nodes. He allocated the lines companies across the three Reference Nodes and derived the retail margin for each firm supplying power through the distribution network of every lines company. He had information about the number of retailers serving every region, the identity of the incumbent retailer in each region, and the fixed monthly connection charge applying in each region.

585               Dr Small analysed these data using econometric methods. He found that retail margins had been increasing over time at an average rate of .15 cents per KWh every 100 days. The result was statistically significant. He found that the retail margin in 1998 prior to any vertical integration of generation and retail was 0.73 cents per KWh lower than it was over the post-integration period. He also found retail margins on average were 0.13 cents per KWh higher for every additional retailer serving a region. Both of these results he described as statistically significant. The results were obtained from models that controlled for variation in the spot price. The results were said to show consistently that even when controlling for variation in the spot price of electricity, retail prices had increased in New Zealand over the last five years and materially since the vertical integration of the industry.

586               Dr Small referred to evidence of the difficulty that major users of electricity had experienced in obtaining hedge cover for their load. Thin hedge markets would affect large purchasers of electricity and potentially net retailers of electricity. When reporting a $300 million loss in August 2001, NGC said it had exited the business because ‘the risks for a net retailer are unacceptable due to the way the hedge and spot markets are functioning’. Since the exit of NGC there had been no significant new entry into retail activities by a stand-alone firm. The reasons were apparent having regard to the statements of major users to which Dr Small referred. Stand-alone retailers would also require hedge cover and lack of supply into the hedge market could be expected to deter entry into retailing by unintegrated firms. Similar arguments would apply to entry into generation by a stand-alone player. Such a potential entrant would be slightly better off than a new retailer because vertical integration had made it somewhat easier to sell hedges in New Zealand than to buy them. Nevertheless the lack of a transparent and liquid hedge market did make entry into generation by a stand-alone firm less attractive.

587               In his conclusion Dr Small said his analysis had been focussed entirely on the New Zealand electricity industry, which had become tightly integrated between the generation and retail sectors during 1998 and 1999. He acknowledged that the policy environment in which that integration occurred may have contributed to the speed with which vertical integration took hold. The New Zealand experience did involve what he thought could be described as a ‘cascade’. That is to say once initial integration deals were struck the practice became widespread very quickly. As a consequence of the integration several major users and a stand-alone retailer had made public statements indicating their difficulty in obtaining hedge cover from integrated generator/retailers. In his opinion it was clear that vertical integration had reduced liquidity in the hedge market relative to what it would have been if integration had not occurred. Relatively low levels of supply to that market had made competition correspondingly less intense. A second longer run effect of thin hedge markets was that it deters entry by new suppliers in generation and in retail. It restricts the ability for a stand-alone generator to sell output. The position of a potential stand-alone entrant into retail is arguably worse, since such firms would need to secure a supply contract from the generation affiliate of a retail rival. So entry would be deterred overall because firms would effectively need to enter both the generation and the retail sectors simultaneously. At the retail level Dr Small’s empirical analysis showed that average margins available to retailers in each region were now higher than they were before the industry was integrated. On average there had been an increase in the difference between reported retail prices and the corresponding average spot price over the last five years. This was consistent with the theory that vertical integration extends capacity constraints from the generation sector to the retail market, which reduces the intensity of competition.

588               In cross-examination by counsel for AGL, Dr Small said he had developed his ‘informal’ model within the last couple of months specifically in relation to the electricity industry and for the purpose of these proceedings. It was not simply an application of the Ordover model for it was concerned with a market characterised by upstream capacity constraints.

589               It was put to Dr Small that the Ordover paper identified a central condition for the success of market foreclosure by vertical integration which applied in the case of differentiated products. Electricity is a homogeneous product. He disagreed with the proposition that the central condition for successful foreclosure could not be fulfilled in such a market. He explained his view by saying that volatility in spot prices would not affect the integrated chain. It would give it a competitive advantage over the unintegrated actors. It was not apparently immediately how this related to the satisfaction of the central condition for successful foreclosure and it was not explicit in Dr Small’s written statement of evidence.

590               The ‘informal’ model proffered by Dr Small is no doubt a useful starting point for analysis of vertical integration in the electricity industry. But it describes a very simple case, that of duopoly in the wholesale and retail sectors, albeit it imposes a capacity constraint condition specific to the industry. A qualitative leap from analysis of vertical integration in a duopoly to its analysis in an oligopoly or a multi-participant competitive market is somewhat analogous to the difference between writing an equation to describe the motion of two bodies in a plane and writing an equation for the motion of many bodies in three dimensions. In the electricity market many actors operate in a ‘high dimensional strategy space’. Dr Small accepted in cross-examination that the New Zealand market is inherently oligopolistic and that his theory was likely to say little or nothing about an oligopoly.

591               While Dr Small’s model has some intuitive appeal, I cannot accept it as a basis for drawing inferences or assessing the likelihood of particular affects of vertical integration in electricity markets. A fortiori it does not assist in assessing the effect of the particular acquisition proposed in this case even on the assumption that AGL were not constrained by the provisions limiting its involvement in LYP’s operations. The authors of the Ordover paper noted that the analysis of vertical relationships falling short of full vertical integration was one ‘… we have only begun to explore’. There was no reference to any further such exploration by Dr Small and, in fairness to him, he was not asked to embark upon that exercise. His brief was confined to consideration of full vertical integration. It was that kind of integration which he considered in the context of the New Zealand experience.

592               In relation to the New Zealand market, Dr Small accepted that the primary driver of large price spikes is the weather, that is to say how hot and dry it is. He accepted that proposition in the context of New Zealand’s reliance upon hydro-powered generators. He said:

‘… most of the hydro generating capacity is located in the bottom of the South Island and if the in-flows to those lakes are low as a result of a dry season, then historically prices, particularly since the market has been in operation, wholesale prices have responded to that.’

593               In my opinion the experience of the New Zealand electricity market does not support any inference that AGL’s proposed acquisition of a 35% interest in Loy Yang, whether held as a passive investor or otherwise, would trigger the process of vertical integration within the market place that would not have occurred in the absence of that acquisition.

594               I have already found, after consideration of AGL’s vertical integration strategy document and the existing market characteristics, that there is a natural tendency on the part of major retailers in the NEM to undertake some degree of vertical integration at the level of peaking or intermediate plant. As I observed earlier, the practical considerations which generate a commercial pressure in that direction will operate independently of any acquisition by AGL. There are also more ways of vertically integrating than by the acquisition of shares in a generator. Considering the wholesale market for electricity and derivative contracts with and without the proposed acquisition, I do not consider there to be any real chance that the acquisition will make any difference to the trend to vertical integration in the variety of ways which is already apparent.

Retail Markets in Victoria

595               The ACCC submitted that the Victorian Residential and Small Business Customer Market is highly concentrated and vulnerable to an increase in barriers to entry which would substantially lessen competition. Furthermore, it submitted that retailers who compete in the Victorian Commercial and Industrial Customer Market are not immediate potential entrants. The evidence was said to show that the Victorian Residential and Small Business Customer Market is characterised by a large number of customers each consuming relatively small quantities of electricity. The other market is characterised by a small number of customers consuming very large quantities of electricity. Because of these differences the business systems, operations and know how required to supply electricity in the former market are distinct from that required for the latter.

596               Secondly, it was submitted that the quantity of electricity consumed in the smaller customer market is highly variable having regard to temperature and the use of air conditioning. In the other market it is more consistent. Retailers serving the smaller customer market therefore need different types of electricity derivative contracts in order to hedge their variable load.

597               The contestability of the smaller customer market was said to be affected by the inertia of customers in that market. A relatively small proportion had changed retail suppliers since the commencement of market contestability.

598               By reason of the preceding factors it was said that the Residential and Small Business Customer Market is highly concentrated, primarily being supplied by three incumbent retailers in Victoria, namely AGL, Origin and TXU. The availability of hedge contracts was said to be a substantial barrier to new entry. Retailers face substantial risks of spot market volatility and would seek to fully hedge that risk. It was said to be clear that if hedges are not readily available at competitive prices retailers would not enter the market or would only do so as vertically integrated entities. It was evident that these submissions depended upon the proposition that the proposed acquisition would lead to a thinning in the hedge market because of AGL’s withdrawal of forward contract cover on account of its acquisition of a natural hedge through its equity in LYP. For reasons which I have already outlined, I do not accept that such a withdrawal is a reasonable hypothesis. For that reason I do not accept the flow-on effects into the retail market propounded by the ACCC.

Conclusions on Whether the Proposed Acquisition Contravenes Section 50

599               On the basis of the findings which I have made concerning the position in the relevant markets with and without the acquisition, I am satisfied on the balance of probabilities that the proposed acquisition would not be likely to have the effect of substantially lessening competition in any market. The question that remains is whether AGL should have the declarations which it seeks.

 

The Grant of Relief – Discretionary Issues

600               The ACCC submitted that for a number of reasons the Court should decline, in the exercise of its discretion, to make the declarations sought by AGL. The first reason advanced was that the scheme of the Trade Practices Act in connection with merger matters makes it inappropriate for a proposed acquirer of shares or assets, concerned about a possible contravention of s 50, who has not sought authorisation, to apply for declaratory relief against such a future potential risk. The ACCC argued that the scheme of the Act is that such a proposed acquirer should either seek authorisation under Pt VII of the Act, including if necessary, exercising the rights of review given under Pt IX, or alternatively should make the acquisition unless restrained by an injunction under s 80.

601               The ACCC submitted that AGL seeks substantial and, arguably illegitimate, forensic and juridical advantages. Apart from greatly foreshortening the ordinary length of the ACCC’s preparation for trial because of the very short timetable for preparation in the present case, AGL was able to prevent the ACCC from having any opportunity to wait and see whether there was likely to be in the future a substantial lessening of competition in any market. It was argued that in ordinary circumstances, in the absence of a declaration or authorisation, the ACCC would have the full term of the limitation period in s 81 in which to decide whether to apply to the Court for relief. Section 81(2) of the Act provides that an application for divestiture may be made at any time within three years after the date on which the contravention occurred. The juridical advantage conferred upon the ACCC expressly by s 81(2) was said to have been removed by the procedure, which AGL has invoked in the present case. This, it was submitted, is contrary to the statutory scheme of the Act. Linked to this proposition was the argument that courts have declined declarations where an equally or more appropriate remedy is available. The authorisation procedure was said to provide a more appropriate means for AGL to achieve the certainty which it seeks having regard to the evident policy of the Act in relation to merger matters. This is particularly so if AGL had a real doubt as to its lawfulness.

602               It was also submitted for the ACCC that the uncertain and hypothetical nature of the issues raised particularly concerning the permanence of the contractual provisions on which AGL places heavy reliance are factors weighing against any exercise of the Court’s discretion to grant the declarations. This is particularly so in circumstances where the declaration sought concerns a matter involving the public interest and not merely the private interests of the parties. The public interest, it was said, is not facilitated by having an accelerated and incomplete consideration of the competition issues predicated on the uncertain and hypothetical foundation, which exists in the present case.

603               It was further put that there is doubt concerning the utility of the declarations. They will not bind any third parties who, after the acquisition, might allege that there has been a contravention of s 50. A third party can apply for divestiture under s 81(1) or seek damages under s 82. A third party is only precluded from seeking an injunction – s 80(1A). Doubts as to whether any res judicata would arise may turn upon whether the proposed acquisition is precisely the same as the actual acquisition.

604               It is surprising to hear from the ACCC the proposition that it is ‘arguably illegitimate’ for a corporation to approach the Court to obtain a declaration as to the lawfulness of a proposed acquisition. The availability and utility of such a procedure in other areas of the law has been endorsed in the High Court. In Commonwealth v Stirling Nicholas Duty Free Pty Ltd (1972) 126 CLR 297 at 305, Barwick CJ observed that the capacity of courts to declare that conduct which has not yet taken place will not be in breach of a contract or a law ‘contributes enormously to the utility of the jurisdiction’. The High Court in Bass v Permanent Trustee Company Ltd (1999) 198 CLR 334 said, of the jurisdiction to grant declaratory relief, at 356, that:

‘The jurisdiction includes the power to declare that conduct which has not yet taken place will not be in breach of a contract or a law and such a declaration will not be hypothetical in the relevant sense.’

 

I have already referred to these authorities in connection with the ACCC’s unsuccessful challenge to the jurisdiction of the Court.

605               The submission that AGL gains some ‘arguably illegitimate advantage by proceeding in this way’ may be contrasted with the observation of the Swanson Committee in that part of its 1976 Report which gave consideration to clearance procedures in respect of mergers. In rejecting submissions that there ought to be provision for clearance appeals to the Tribunal, the Committee said at 8.39:

‘Many submissions urged on us the absolute need for expedition in processing merger clearance matters, and an appeal procedure would inevitably frustrate this. We think a party disappointed with the Commission’s decision may take its own decision whether to proceed at risk or seek a declaration from the Court. In this regard we note the Trade Practices Act Amendment Bill 1976 would incorporate a new section into the Act, section 163A which would allow any person to approach the Court to seek a declaration on these matters.’

 

606               In considering the availability of authorisation as an alternative remedy, affecting the discretion to grant relief, it is necessary to have regard to the nature of the authorisation process and the basis upon which authorisation decisions are made. It respect of acquisitions of shares in a body corporate, s 88(9) of the Trade Practices Act provides:

‘Subject to this Part, the Commission may, upon application by or on behalf of a person:

(a) grant an authorisation to the person to acquire shares in the capital of a body corporate or to acquire assets of a person.

…

and, while such an authorisation remains in force:

(c) in the case of an authorisation under paragraph (a) – section 50 does not prevent the person from acquiring shares or assets in accordance with the authorisation;’

Section 90(9) provides:

‘90(9) The Commission shall not make a determination granting an authorisation under subsection 88(9) in respect of a proposed acquisition of shares in the capital of a body corporate or of assets of a person or in respect of the acquisition of a controlling interest in a body corporate within the meaning of section 50A unless it is satisfied in all the circumstances that the proposed acquisition would result, or be likely to result, in such a benefit to the public that the acquisition should be allowed to take place.’

Section 90(9A) specifies, non-exhaustively, certain matters which must be taken into account as benefits to the public, these being matters which relate to Australia’s international trade and competitiveness. It is clear that the authorisation process vests in the Commission a power to make an evaluative assessment of the benefit flowing from the proposed acquisition without necessarily determining that the acquisition would contravene s 50. So much has been recognised in Tribunal decisions on the nature of authorisation. In QCMA, as noted earlier, the Tribunal said that it did not believe that an application for authorisation could carry with it any presumption as to liability under s 50 or the presence of significant or substantial anti-competitive effects. The task of the Tribunal is not first to inquire whether the acquisition is ‘likely to have the effect of substantially lessening competition … because the issues for determination by the Court in the event of prosecution are different from those for determination by the Tribunal when authorisation is sought’ – Re QIW Ltd (1995) 132 ALR 225 at 235 citing QCMA.

607               An application for authorisation need not answer the question whether a proposed acquisition if it proceeded would contravene s 50. It brings to bear as an important consideration, quite extraneous to the construction and application of s 50, the question whether there is public benefit resulting or likely to result from the acquisition. That is a matter for which the Act provides a means of assessment which is administrative in both a functional and constitutional sense. It involves polycentric decision-making of a kind which the Court is not institutionally competent, nor authorised by statute or the Constitution, to undertake. As I have written elsewhere:

‘Courts have neither the resources nor, as a general rule, personnel with the skills and experience necessary to undertake wide ranging inquiries of the kind that may be necessary for the resolution of public benefit or efficiency issues in authorisation applications. The investigative process, the receipt of submissions from interested groups and parties, the evaluation of their roles and interests in the relevant market and the striking of compromises which may be reflected in the conditions attached to authorisations are not within the functions to which courts are confined.’

French, ‘Role of Courts in the Development of Australian Trade Practices Law’ in Hanks and Williams (eds)Trade Practices Act – A Twenty Five Year Stocktake Federation Press 2001 pp 98- 116 at 108

 

608               I conclude that the availability of the authorisation procedure affords no discretionary bar to the right of a corporation to claim a declaration with respect to a proposed acquisition. That is not to say that there may not be other reasons for refusing relief according to the circumstances of the particular case.

609               When writing to AGL on 12 June 2003 the ACCC said it had ‘conducted a thorough investigation so as to determine the likely effect of the transaction on both the generation and retail sectors of the electricity supply chain’. In its letter of 5 September, the ACCC expressed its ‘significant reservations’ and that it reserved its right to take action for contravention of s 50 including penalty and divestiture action.

610               In its press release of 8 September 2003, the ACCC was ‘… firmly of the view that the proposed acquisition creates substantial competition concerns which are potentially in breach of s 50 of the Trade Practices Act 1974.’ It stated unequivocally that the acquisition ‘… would lead to a less competitive and less efficient market structure in Victoria and potentially, in the National Electricity Market’. It stated that the ACCC would oppose AGL acquiring an interest in LYP, that it remained ‘strongly opposed to this transaction’, and would ‘continue to build its enforcement case should AGL, the Commonwealth Bank and TEPCo decide to close the transaction without providing undertakings satisfactory to the ACCC.’ In the light of these statements it is difficult to understand why the ACCC should now complain that it is deprived of the opportunity to wait up to three years before bringing a divestiture action.

611               In my opinion there is no ‘arguably illegitimate forensic or juridical advantage’ accruing to AGL by reason of bringing these proceedings nor is there any unfair loss of a juridical or forensic advantage to the ACCC. Indeed in these proceedings the ACCC enjoys the forensic advantage, which it would not have enjoyed, had it brought injunction proceedings, that AGL must disprove the likelihood of a substantial lessening of competition flowing from its acquisition.

612               Notwithstanding my conclusions about the question of discretion in the present case, there may be cases in which discretion would bar the grant of declaratory relief. A corporation would be unlikely to succeed in persuading the Court to grant it a declaration on the basis of a transaction merely in contemplation absent any controversy. A justiciable controversy can arise, as in this case, where the proposed acquirer has approached the ACCC which after consideration of the proposed transaction, which is real and not merely hypothetical, has stated its opposition to it and its intention to act against it. A like controversy could arise by reason of the opposition of a third party. Each case will fall to be decided on its own facts. In this case the opposition of the ACCC is unequivocal. It has not proceeded to claim injunctive relief but has threatened post-acquisition divestiture action. It is not in the least surprising that AGL would not wish to enter into this major transaction with that sword of Damocles hanging over it and the other members of the consortium. Indeed it is difficult to see how, if the transaction were to proceed in the face of such a threat, the public interest would be served with such uncertainty hanging over the operation of a major public utility.

The Form of Relief

613               The undertaking offered to the Court in my opinion, provides, in substance, a significant constraint upon the exercise of the kind of influence or control by AGL upon the operations of LYP that might conceivably given rise to anti-competitive effects. It can no doubt be argued that an undertaking of greater precision could be formulated. I am satisfied however, that the terms of the undertaking offered in its interaction with the agreements which refer to it, constitute a significant safeguard against AGL having access to confidential information relating to its retail competitors or the day to day operations of LYP. It also safeguards to a substantial degree against such a level of influence and control of LYP’s dispatch and bidding operations as might have any real adverse affect upon the competitive process expressed in the day-to-day bidding and hedging practices of LYP.

614               In so saying, I do not regard the undertaking as the only form of safeguard against that eventuality. The presence of other major shareholders, particularly TEPCo, the constraints which I find are likely to be imposed by external lenders and the fact that any further acquisition would be subject to potential s 50 action combine to lead to the conclusion that it is appropriate in the circumstances to grant the declarations sought. Nevertheless the undertaking is, in my opinion, an appropriate condition for the grant of the declarations. That is subject to two qualifications.

615               The first qualification relates to the last two parts of the undertaking which concern the circumstances under which it may be discharged or varied. They appear to imply that the Court may have an information-gathering role in relation to compliance with the undertaking. While these matters may be relevant to the form of the undertaking as offered to the ACCC under s 87B of the Trade Practices Act, they are not appropriate as a part of the undertaking offered to the Court. I would therefore accept an undertaking deleting those provisions and definitions relevant only to them but subject to liberty to AGL to apply, on notice to the ACCC to discharge or vary the undertaking at any time without any present specification of the grounds upon which such an application may be brought.

616               The second qualification is that the undertaking refers, in prospect, to AGL’s entry into the Loy Yang Consortium Agreements. I will accept the undertaking offered in these terms for the moment upon the basis that it is taken to refer to agreements already executed as well as agreements yet to be executed. In a sense the agreements are all in prospect until the conditions precedent are satisfied. The Court will entertain a variation of the undertaking for the purpose of tidying up that aspect of its drafting.

617               I also propose to make a provision in the order which would allow the ACCC, at its option, to accept a renewal of the undertaking previously offered to it under s 87B in which event the undertaking to the Court will be discharged. In the form previously offered to the ACCC, which is the form annexed to these reasons for judgment, there was a facility for the ACCC to continue to have a monitoring role in relation to compliance with the undertaking. Should it wish to now assume that role, it is open to it to do so. Alternatively, it is open to AGL and the ACCC to negotiate a varied or substituted form of undertaking pursuant to s 87B of the Act which would allow the ACCC to have an ongoing monitoring role should it wish to do so.

618               AGL having been successful in the action, the costs should follow the event. The ACCC will be ordered to pay AGL’s costs of the proceedings.

I certify that the preceding six hundred and eighteen (618) numbered paragraphs are a true copy of the Reasons for Judgment herein of the Honourable Justice French.

Associate:

Dated: 19 December 2003

Counsel for the Applicant:	Mr TF Bathurst QC and Mr SJ Gageler SC with Mr JD Elliott and Mr AJ Payne
Solicitor for the Applicant:	Gilbert and Tobin
Counsel for the Respondent:	Mr J Beach QC and Mr N O'Bryan SC with Mr PJ Cosgrave and Mr MJ O'Bryan
Solicitor for the Respondent:	Phillips Fox
Date of Hearing:	18-21, 24-28 November, 1-2, 4-5 December 2003
Date of Judgment:	19 December 2003

Annexure 1

Scheduled generators in the NEM

 

Registered Participant

Station

Class

Type

Participant’s total state capacity

Registered capacity (MW)

Capacity Factor

 

NSW

 

Delta Electricity	Mt Piper Power Station	MS	Coal	4240MW	2*660MW = 1320MW	84.1%
	Munmoarah Power Station	MS	Coal		2*300MW = 600MW	9.3%
	Vales Point “B” Power Station	MS	Coal		2*660MW = 1320MW	56.0%
	Wallerawang “C” Power Station	MS	Coal		2*500MW = 1000MW	60.1%
Eraring Energy	Eraring Power Station	MS	Coal	2905MW	4*660MW = 2640MW	64.0%
	Hume Power Station	MS	Hydro		25MW	25.6%
	Shoalhaven Power Station (Bendeela and Kangaroo Valley)	MS	Hydro		2*40MW+2*80MW = 240MW	2.9%
Macquarie Generation	Bayswater	MS	Coal	4690Mw	4*660MW = 2640 MW	74.7%
	Hunter Valley Gas Turbine	MS	Liquid Fuel		2*25MW = 50MW	0.0%
	Liddell	MS	Coal		4*500MW = 2000MW	43.4%
Redbank Project Pty Ltd	Redbank Power Station	MS	Coal	150MW	150MW	79.7%
Snowy Hydro Limited	Blowering 1x80MW	MS	Hydro	80MW	80MW	19.4%
Sithe Australia Power	Smithfield Energy Facility	NMS	Gas	176MW	3*38MW+1*62MW = 176MW	63.9%
Total				12,241MW

 

Victoria

 

AGL Electricity Limited	Somerton Power Station	MS	Gas	160MW	4*40MW = 160MW	0.5%
Duke Energy Bairnsdale Operations Pty Ltd	Bairnsdale Power Station	MS	Gas	92MW	2*46MW = 92MW1	14.7%
Ecogen Energy Pty Ltd	Jeeralang “A” Power Station	MS	Gas	932MW	4*51MW = 204MW	0.6%
	Jeeralang “B” Power Station	MS	Gas		3*76MW = 228MW	1.8%
	Newport Power Station	MS	Gas		500MW	13.3%
Edison Mission Energy	Loy Yang B Power Station	MS	Coal	1000MW	2*500MW = 1000MW	99.7%
Energy Brix Australia	Energy Brix Complex	MS	Coal	195MW	1*24MW+2*33MW+1*30MW+1*75MW = 195MW	62.4%
Eraring Energy	Hume Power Station	MS	Hydro	25MW	25MW	55.3%
Hazelwood Power	Hazelwood Power Station	MS	Coal	1600MW	8*200MW = 1600MW	84.5%
Loy Yang Power Management Pty Ltd	Loy Yang A Power Station	MS	Coal	200MW	4*500MW = 2000MW	88.0%
SECV	Angelsea Power Station	MS	Coal	150MW	150MW	103.3%2
Southern Hydro Partnership	Darmouth Power Station	MS	Hydro	422MW	150MW	47.3%
	Eildon Power Station	MS	Hydro		2*60MW = 120MW3	10.6%
	McKay Power Station	MS	Hydro		2*45MW = 90MW	4.3%
	West Kiewa Power Station	MS	Hydro		2*31MW = 62MW	18.1%
Valley Power Pty Ltd	Valley Power Peaking Facility	MS	Gas	300MW	6*50MW = 300MW	1.8%
Yallourn Energy	Yallourn W Power Station	MS	Coal	1450MW	2*350MW=2*375MW = 1450MW	89.4%
Total				8,326MW

Queensland

Callide Power Trading Pty Ltd	Callide Power Plant	MS	Coal	840MW	2*420MW = 840MW	79.8%
CS Energy	Callide A Power Station	MS	Coal	2150MW	4*30MW = 120MW	0.0%
	Callide B Power Station	MS	Coal		2*350MW = 700MW	95.4%
	Swanbank A Power Station (mothballed during 2002)	MS	Coal		6*68MW = 408MW	6.3%
	Swanbank B Power Station	MS	Coal		4*125MW = 500MW	51.8%
	Swanbank D Gas Turbine	MS	Liquid Fuel		37MW	0.1%
	Swanbank E Gas Turbine	MS	Gas		385MW	8.0%
Millmerran Energy Trader Pty Ltd	Millmerran Power Plant	MS	Coal	852MW	2*426MW = 852MW	9.7%
Origin Energy Electricity Limited	Roma Gas Turbine Station	MS	Gas	80MW	2*40MW = 80MW	6.1%
Eneretrade (QPTC)4	Barcaldine Power Station	MS	Gas	2668MW	57MW	36.4%
	Collinsville Power Station	MS	Coal		4*30MW=1*60MW = 180MW	44.3%
	Gladstone	MS	Coal		6*280MW = 1680MW	65.5%
	Oakey Power Station	MS	Liquid Fuel		2*141MW = 282MW	0.2%
	Mt Stuart Gas Turbine	MS	Liquid Fuel		2*152MW = 304MW	2.8%
	Townsville Gas Turbine	MS	Liquid Fuel		165MW	1.2%
Stanwell Corporation	Barron Gorge	MS	Hydro	1566MW	2*30MW = 60MW	10.4%
	Kareeya	MS	Hydro		4*18MW = 72MW	22.5%
	Mackay Gas Turbine	MS	Liquid Fuel		34MW	0.1%
	Stanwell Power Station	MS	Coal		4*350MW = 1400MW	82.6%
Tarong Energy	Tarong Power Station	MS	Coal	2315MW	3*350MW+1*315MW = 1365MW	95.2%
	Tarong North Power Station	MS	Coal		450MW	5.6%
	Wivenhoe Power Station	MS	Hydro		2*250MW = 500MW	4.3%
Total				10,471MW

 

South Australia

AGL Electricity Limited	Hallett Power Station	MS	Liquid Fuel	220MW	8*16.8MW+2*25.2MW+2*17.9MW = 220MW	0.5%
NRG Flinders Operating Services Pty Ltd	Northern Power Station	MS	Coal	950MW	2*265MW = 530MW	89.1%
	Osborne Power Station	MS	Gas		1*118MW=1*62MW = 180MW	82.5%
	Playford B Power Station	MS	Coal		4*60MW = 240MW	2.8%
Origin Energy Electricity Limited	Ladbroke Grove Power Station	MS	Gas	176MW	2*40MW = 80MW	81.6%
	Quarantine Power Station	MS	Gas		4*24MW = 96MW	8.8*
Pelican Point Power Limited	Pelican Point Power Station	MS	Gas	478MW	2*160MW+1*158MW = 478MW	65.1%
Synergen Power Pty Ltd	Dry Creek Gas Turbine Station	MS	Gas	359MW	3*52MW=156MW	0.5%
	Mintaro Gas Turbine Station	MS	Gas		90MW	1.3%
	Port Lincoln Gas Turbine	MS	Liquid Fuel		2*25MW = 50MW	0.1%
	Snuggery Power Station	MS	Liquid Fuel		3*21MW = 63MW	0.5%
TXU (South Australia) Pty Ltd	Torrens Island Power Station “A”	MS	Gas	1280MW	4*120MW = 480MW	6.1%
	Torrens Island Power Station “B”	MS	Gas		4*200MW = 800MW	33.0%
Total				3463MW

 

Snowy

Snowy Hydro Limited	Guthega	MS	Hydro	3676MW	2*30MW = 60MW	22.9%
	Murray	MS	Hydro		10*95MW+4*137.5MW = 1500MW	11.3%
	Tumut	MS	Hydro		4*82.4MW+4*71.6MW +6*250MW = 2116MW	0.5%
Total				3676MW

The source of this table, put in evidence through Mr Price, is the NEMMCO List of Generators and Scheduled Loads, Version 73 dated 7 August 2003 and Version 27, 12 September 2000 for Swanbank A (removed from Version 73 due to mothballing of the plant in 2000). The capacity factors are derived from analysis of NEMMCO trading interval generation and registered capacities. The table includes all scheduled generators regardless of whether they are market or non-market generators. The capacity factors have been calculated using the total energy, on an as generated basis, produced by each power station during the 2002 calendar year.

Annexure 2

 

Table 1: Share of Scheduled Registered Capacity and Energy in the NEM (at the generator terminals)

 

Participant	% of total capacity	% of total generation (2002 calendar year)
Macquarie Generation	12.42%	13.97%
Delta Electricity	11.23%	12.32%
Snowy Hydro Limited	9.84%	2.07%
Eraring Energy	7.76%	8.43%
Enertrade	7.06%	5.95%
International Power	6.45%	8.18%
Tarong Energy	6.13%	6.61%
TXU Pty Ltd	5.86%	1.79%
CS Energy	5.72%	6.35%
Loy Yang Power Management Pty Ltd	5.30%	8.64%
Stanwell Corporation	4.15%	5.79%
Yallourn Energy	3.84%	6.37%
Edison Mission Energy	3.44%	4.93%
Intergen	3.37%	2.05%
NRG Flinders Operating Services Pty Ltd	2.52%	3.08%
Southern Hydro Partnership	1.12%	0.48%
AGL Electricity Limited	1.01%	0.01%
Origin Energy Electricity Limited	0.68%	0.39%
Energy Brix Australia	0.52%	0.60%
Sithe Australia Power	0.47%	0.55%
Redbank Project Pty Ltd	0.40%	0.59%
SECV	0.40%	0.76%
Duke Energy Bairnsdale Operations Pty Ltd	0.24%	0.07%
Total	100%	100%

 

Table 2: Share of Scheduled Registered Capacity and Energy in Victoria (at the generator terminals)

 

Participant	% of total capacity	% of total generation (2002 calendar year)
Interconnectors	21.79%	29.89%
Loy Yang Power Management Pty Ltd	18.79%	22.97%
International Power	15.03%	22.97%
Yallourn Energy	13.62%	22.03%
Edison Mission Energy	12.21%	17.03%
TXU Pty Ltd	8.75%	1.22%
Southern Hydro Partnership	3.96%	1.68%
Energy Brix Australia	1.83%	2.07%
AGL Electricity Limited	1.50%	0.01%
SECV	1.41%	2.63%
Duke Energy Bairnsdale Operations Pty Ltd	0.86%	0.23%
Eraring Energy	0.23%	0.23%
Total	100.0%	100.0%

Annexure 3

Corporate Groups with Retail Licenses in the NEM
Retailer	Company Background	Licences
		NSW	Vic	Qld	SA	ACT
ActewAGL Retail	Franchise retailer and stapled to DNSP in ACT	ü	ü	ü	ü	ü
AGL	Franchise retailer and stapled to DNSP in metropolitan and regional Victoria, franchise retailer in SA	ü	ü	ü	ü	ü
Aurora Electricity	Franchise retailer and stapled to DNSP in Tasmania	ü	ü
Australian Energy Services (Power Direct)	Independent retailer	ü	ü	ü
Australian Inland Energy and Water	Franchise retailer and stapled to DNSP in regional NSW	ü
Country Energy	Franchise retailer and stapled to DNSP in regional NSW	ü	ü	ü	ü	ü
CS Energy	Queensland generator			ü
Delta Electricity	NSW generator	ü
Energex Retail	Franchise retailer and stapled to DNSP in metropolitan Queensland	ü	ü	ü	ü	ü
Energy Australia	Franchise retailer and stapled to DNSP in metropolitan NSW	ü	ü	ü	ü	ü
Eraring Energy	NSW generator	ü
Ergon Energy	Franchise retailer and stapled to DNSP in regional Queensland	ü	ü	ü		ü
Ferrier Hodgson Electricity	Independent retailer	ü		ü		ü
Integral Energy	Franchise retailer and stapled to DNSP in metropolitan NSW	ü	ü	ü		ü
Jackgreen (International)	Independent retailer	ü
NRG	SA generator				ü
Retailer	Company Background	Licences
		NSW	Vic	Qld	SA	ACT
Origin Energy (including Citipower)	Franchise retailer in metropolitan Victoria	ü	ü	ü	ü	ü
Tarong Energy	Queensland generator			ü	ü
TXU Electricity (including TXU)	Franchise retailer and stapled to DNSP in metropolitan and regional Victoria	ü	ü	ü	ü	ü
Victoria Electricity	Independent retailer		ü
Yallourn Energy (AusPower)	Victorian generator	ü	ü	ü	ü	ü

Annexure 4

 

Annexure 5

Annexure 6

Annexure 7

Annexure 8

UNDERTAKINGS

BY THE AUSTRALIAN GAS LIGHT COMPANY

 

1. BACKGROUND

 

1.1 A consortium comprising The Australian Gas Light Company Limited (AGL), The Tokyo Electric Power Company, Incorporated (TEPCO) and certain financial investors including the Commonwealth Bank of Australia (Financial Investors) propose to acquire the electricity generation business in Victoria currently conducted by the partners in the Loy Yang Power Partnership and known as Loy Yang A, with AGL holding a 35% interest, the Financial Investors holding 30% interest and TEPCO holding a 35% interest in that business.

1.2 This Undertaking has been given to the Commission and/or the Court in order to facilitate the acquisition by the consortium referred to in clause 1.1 of Loy Yang A on the basis outlined in clause 1.1.

2. DEFINITIONS

 

Act means the Trade Practices Act 1974 (Cth).

AGL means The Australian Gas Light Company ABN 95052 167 405 and/or its Related Bodies Corporate.

Aggregate Information means any information:

(a) derived from Confidential Generator Information or Confidential Customer Information (such as a sum, average, statistical analysis, comparison or general qualitative description); but

(b) from which the underlying Confidential Generator Information or Confidential Customer Information cannot reasonably be derived or ascertained.

Business Day means a day other than a Saturday, Sunday or public holiday in Victoria.

Commission means the Australian Competition and Consumer Commission.

Contracted Capacity means Registered Capacity in respect of which a party has a contractual right to control the Dispatch and Market Activities of that Registered Capacity.

Confidential Customer Information means the details (including identity of each counter-party, price, term and volume) of any specific Customer Contracts with the Marketing Management Company, but excluding:

(a) information which is generally known; and

(b) AGL’s own Customer Contracts with the Marketing Management Company.

Court means the Federal Court of Australia.

Customer Contract means an electricity derivative contract or power purchase agreement:

(a) entered into by the Market Management Company; or

(b) considered, proposed or likely to be entered into by the Market Management Company.

Confidential Generator Information means:

(a) details (including identity of each counter-party, price, term and volume) of Dispatch and Marketing Activities; and

(b) details (including quantities, dates and times) of any reductions or expected reductions in the availability of the Loy Yang plant to less than the plant’s Registered Capacity,

but excluding information which is generally known.

Confidentiality Regime means the arrangements described in paragraph 3.6.

Consortium means the structure by which the Consortium Members hold interests in the Loy Yang Business, from time to time.

Consortium Members means The Australian Gas Light Company, Tokyo Electric Power Company and the Financial Investors, or any of their Related Bodies Corporate or successors.

Dispatch and Marketing Activities means:

(a) the determination and management of the scheduling of available capacity at the Loy Yang Plant;

(b) the determination and management of trading, dispatch and re-bidding and contracting strategies;

(c) the placement of dispatch offers and re-bids;

(d) entering into Customer Contracts; and

(e) regulatory policy and dealings with relevant economic and competition regulators (including any issues arising under the Code or the Act) for the Loy Yang Business.

Economic Interest

(a) means interests in a company or partnership, including, shares, voting rights, rights to receive dividends, rights to receive other distributions of income or capital, rights to receive a share of proceeds on winding up; but

(b) excludes:

(i) rights to purchase the interest in the Loy Yang Business or the Loy Yang Assets of a Consortium Member seeking to divest its interest the exercise of which are subject to AGL obtaining approval (on a formal or informal basis) from the Commission or the Australian Competition Tribunal; and

(ii) any rights AGL has to prevent approval of decisions in respect of Permitted Matters.

Financial Investors means Commonwealth Bank of Australia or other equity investors in the Loy Yang Business.

Force Majeure means any event or circumstance not within the control of AGL and which, by the exercise of reasonable diligence, AGL is not reasonably able to prevent or overcome including (but not limited to) the following events or circumstances:

(a) acts of God, including without limitation, earthquakes, floods, washouts, landslides, lightning, storms and the elements;

(b) acts of enemy, wars, blockades or insurrections, riots and civil disturbances or acts of terrorism;

(c) fire or explosion;

(d) strikes, lockouts, bans or other industrial disturbances; and

(e) acts or omissions of any government or governmental agency or authority.

FM Affected Obligation is defined in clause 5.2(a)(i).

Intermediary has the meaning given in the National Electricity Code.

Loy Yang Assets means the electricity generating plant and the coal mine used in the operation of the Loy Yang Business.

Loy Yang Business means the electricity generation business operating in Victoria, known as Loy Yang A.

Loy Yang Consortium Agreements means a Shareholders Agreement, a Partnership Agreement, and any other arrangements between the Consortium Members (or their Related Bodies Corporate) or between entities which manage the Loy Yang Business.

Loy Yang Plant means the electricity generating plant used in the operation of the Loy Yang Business.

Market Management Company means a company appointed in accordance with clause 3.2(a).

Material Change of Circumstances means any matter that causes or results in a material change to the competitive position of the Loy Yang Business or to AGL’s competitive position in the generation or retailing of electricity.

NEM means National Electricity Market.

Permitted Matters means the arrangements described in paragraph 3.4.

Related Bodies Corporate has the meaning given in sub-section 4A(5) of the Act.

Risk Management Policy has the meaning given to it in Appendix A.

Registered Capacity means the capacity, measured in megawatts, of a generator that is registered by the National Electricity Market Management Company Limited.

Victorian Generation Registered Capacity means the total of:

(a) the Registered Capacity of Generators situated in Victoria; and

(b) the Registered Capacity of Market Network Service Providers with Interconnects into Victoria in respect only of capacity that may be able to be supplied by such Providers into Victoria.

The terms Megawatts, Generators, Market Network Service Providers and Interconnects have the meaning given to them in the National Electricity Code.

3. UNDERTAKINGS

 

3.1 AGL undertakes that the combined Economic Interest of AGL in the Loy Yang Assets and/or the Loy Yang Business will not exceed 35%.

3.2 AGL undertakes that the Loy Yang Consortium Agreements, which it will enter into with the Consortium Members, will require the following arrangements regarding the governance of the Loy Yang Business to be put in place and maintained until these Undertakings cease in accordance with paragraph 4.2:

(a) the Consortium Members will appoint or will procure the appointment of a company (Marketing Management Company) which will be solely responsible for undertaking Dispatch and Marketing Activities as the agent of the Consortium Members.

(b) AGL will be prohibited from having any Economic Interest in the Marketing Management Company. It also will not enter into any contracts, arrangements or understandings with the shareholders of the Marketing Management Company which would in effect confer on it such an Economic Interest.

(c) The terms upon which the Marketing Management Company is appointed in respect of the matters in paragraph (a) will include a requirement that any dealings between it and AGL are to be conducted at arms length.

(d) AGL will not participate in the appointment or supervision of the executive management of the Marketing Management Company.

(e) For the avoidance of doubt Dispatch and Marketing Activities which the Marketing Management Company undertakes as the agent of the Consortium Members will not give rise to a contravention of this undertaking provided that paragraphs 3.2(a) to (d) are otherwise complied with.

3.3 AGL undertakes not to be otherwise involved in:

(a) the Dispatch and Marketing Activities of the Loy Yang Business; or

(b) any board or management decision-making (including at the level at which AGL holds its Economic Interest in the Loy Yang Business or at the Marketing Management Company level) in respect of Dispatch and Marketing Activities,

in each case except in so far as the subject matter of the decision-making involves a Permitted Matter.

3.4 For the avoidance of doubt, the Consortium Agreements may provide that certain specific matters relating to the conduct of the Loy Yang Business require the agreement and/or participation of AGL. Those matters are (Permitted Matters):

(a) the financing and capital structure of the Loy Yang Business;

(b) variations to the structure by which the Consortium Members hold interests in the Loy Yang Business;

(c) expansion of the Loy Yang Business and acquisitions, construction and disposals of substantial assets;

(d) appointment or change of auditors for the Loy Yang Business;

(e) distribution policy of the Loy Yang Business;

(f) key corporate changes, including a merger, trade sale, initial public offering, dissolution, suspension or winding up of the Loy Yang Business or constitutional amendments in respect of the Loy Yang Business, except in circumstances where that event would case AGL to breach paragraphs 3.1, 3.2 and 3.3 of these Undertakings;

(g) annual budgets (including capital and operating expenditure) and approvals of expenditures and liabilities outside of budget of the Loy Yang Business;

(h) the Risk Management Policy of the Loy Yang Business;

(i) environmental policies, occupational health and safety policies and industrial relations policies of the Loy Yang Business; and

(j) commencing, defending or settling claims of the Loy Yang Business.

3.5 For the avoidance of doubt, AGL may hold an Economic Interest not exceeding 35% in a company which provides the following services to the Loy Yang Business:

(a) management, labour, engineering and other services for the physical day-to-day operation and maintenance of the Loy Yang Assets; and

(b) services in relation to the matters referred to in paragraph 3.4(i).

3.6 AGL undertakes that the Loy Yang Consortium Agreements will make provision for the adoption of, and compliance with, a Confidentiality Regime which prohibits AGL having access to:

(a) Confidential Customer Information; and

(b) Confidential Generator Information.

For the avoidance of doubt the Confidentiality Regime does not preclude AGL from having access to Aggregate Information reasonably necessary for AGL to assess compliance with policies regarding Permitted Matters, so long as AGL uses the Aggregate Information solely for the purpose of assessing compliance with policies regarding Permitted Matters.

3.7 Notwithstanding paragraph 3.6, if a company is appointed pursuant to paragraph 3.5 then:

(a) paragraph 3.6 does not apply in respect of information which falls within paragraph (b) of the definition of Confidential Generator Information provided to any director appointed by AGL to that company; but only if

(b) any such director has provided a written undertaking to not disclose to AGL, its officers, employees or agents any confidential information within paragraph (b) of the definition of Confidential Generator Information.

4. COMMENCEMENT AND DURATION OF UNDERTAKINGS

4.1 Commencement of Undertakings

These Undertakings commence on the date on which the latest of the following events occurs:

(a) their execution by AGL;

(b) their acceptance by the Commission or the Court, as the case may be; and

(c) the completion of the acquisition of the Loy Yang Business by a consortium which includes AGL.

4.2 Cessation of Undertakings

These Undertakings cease in the event that any of the following events occurs:

(a) AGL and its Related Bodies Corporate cease to have an Economic Interest in the Loy Yang Business and the Loy Yang Assets;

(b) AGL and its Related Bodies Corporate cease to retail electricity in Victoria;

(c) the aggregate of:

(i) the Registered Capacity of the Loy Yang Business and other Registered Capacity of any other business in which AGL hold more than a 35% Economic Interest; and

(ii) any Registered Capacity in which AGL has Contracted Capacity,

is less than 20% of Victorian Generation Registered Capacity; or

(d) the Commission or the Court, as the case may be, determines that it is no longer necessary for AGL to keep the Undertaking in place.

5. ANCILLARY MATTERS

5.1 Review of Undertakings

(a) If AGL:

(i) is unable to comply with its obligations under these Undertakings; or

(ii) believes it is necessary to seek some modification due to changed circumstances (including without limitation a Material Change in Circumstances,

then:

A. where this Undertaking has been given to the Commission, AGL and the Commission agree that they will review these Undertakings and negotiate in good faith the variation or revocation of all or any of the Undertakings in light of such circumstances having regard to the need to maintain competition in the retail electricity industry in the NEM; or

B. where this Undertaking has been given to the Court, AGL will be entitled to ask the Court to review the Undertakings in light of such circumstances having regard to the need to maintain competition in the retail electricity industry in the NEM.

(b) Paragraph 5.1(a) will not operate in respect of changed circumstances that are:

(i) known to or reasonably foreseeable by AGL at the date of this undertaking; or

(ii) arising, whether directly or indirectly, from an act, matter or thing done by or on behalf of AGL or a failure by AGL to do an act, matter or thing which is within AGL’s control.

5.2 Force Majeure

(a) If, and only for as long as, AGL:

(i) is prevented from performing one or more obligations under these Undertakings or parts of an obligation due to any Force Majeure (the FM Affected Obligation); and

(ii) is complying with paragraph (b) in respect of any Force Majeure,

then the FM Affected Obligation does not apply. For the avoidance of doubt, all the other obligations or parts of obligations as the case may be of these Undertakings continue to apply with full force and effect.

(b) AGL must, as expeditiously as possible, use all reasonable diligence and employ all reasonable means to remedy, abate and minimise the scope of any Force Majeure.

(c) If pursuant to paragraph (a) AGL is relieved from an obligation, AGL must notify the Commission or the Court in writing within 5 Business Days specifying:

(i) the FM Affected Obligation and the cause and extent of the non-performance;

(ii) the date of commencement of Force Majeure and its expected duration; and

(iii) the means proposed and adopted to remedy abate and minimise the scope of the Force Majeure.

6. INFORMATION AND REPORTING

6.1 The Commission or the Court, as the case may be, may at any time request in writing that AGL provide such written information that the Commission or the Court reasonably requires for the purposes of monitoring compliance with paragraphs 3.1 to 3.4.

6.2 AGL must promptly comply with any reasonable requests made by the Commission or the Court, as the case may be, under paragraph 6.1.

7. ACKNOWLEDGMENTS

AGL acknowledges that these Undertakings may be made available for public inspection and may, from time to time, be publicly referred to.

THE COMMON SEAL of THE AUSTRALIAN

GAS LIGHT COMPANY is fixed in the

Presence of:

Signature of Director Signature of Director/Secretary

Name of Director (print) Name of Director/Secretary (print)

DATED:

APPENDIX A TO ACCC UNDERTAKING

 

RISK MANAGEMENT POLICY

 

 

The term Risk Management Policy, in the context of an electricity generation business, means a policy document which has the following characteristics:

1. It is a document which records a formal policy adopted by the highest governance body within a business entity (for example a Board of Directors or Partnership Committee).

2. Its purpose is to preserve the value of the business’s assets and ability to deliver budgeted outcomes by:

(a) setting global limits and controls on the business’s exposure to; and

(b) establishing the internal governance structure and management philosophy for managing,

specific categories of business risk.

3. The specific risk areas dealt with would include:

(a) major asset (physical) risk (for example risks of losses arising from poor maintenance on key productive asserts (turbines));

(b) trading risks (price, volume and credit risk) (for example risk of losses from or total contract portfolio exceeding available plant capacity or a disproportionate exposure to spot prices);

(c) operational risk (for example breakdown in human resources, processes or technology); and

(d) legal and compliance risk (for example poor contract management systems, absence of compliance with Trade Practices Act, Corporations Act or National Electricity Code).

4. The global limits and controls may be quantitative or qualitative in nature and would operate to limit or control the total level of risk (often measured in financial terms) that a specific business activity or division may incur. For example:

(a) trading risk (price and volume) may be subject to limits on the proportion of financial budget forecasts which may be put at risk as a consequences of contract portfolio exceeding available plant capacity or a disproportionate exposure to spot prices;

(b) credit risk (counterparty default risk) would be subject to limits on the financial exposure to counterparties of varying categories of credit worthiness; and

(c) legal contract risk would be subject to requirements on use of ISDA pro forma contracts and ISDA optional force majeure clauses in those contracts.

5. The internal governance structure and management philosophy for managing risks would provide for:

(a) organisational structures which segregated relevant functions (such as staff responsible for trading, transaction confirmation, settlements, risk management, accounting and financial reporting, and internal audit); and

(b) delegations of authorities regarding activities with exposure to specified risks and allocations of responsibility for risk management functions.

6. In relation to trading, credit and legal risk the Risk Management Policy would provide for the delegation to a Risk Management Committee responsibility:

(a) for overseeing risk management operations and procedures (within the parameters of the Risk Management Policy) of activities involving energy market exposure; and

(b) for setting individual trader limits.

7. A Risk Management Policy is not a Risk Management Framework. A Risk Management Framework, in the context of an electricity generation business, would:

(a) be approved by the Risk Management Committee; and

(b) specify the management procedures, rules and specific controls for implementing the Risk Management Policy.