Federal Court of Australia

HBSY Pty Ltd v Lewis (No 2) [2025] FCAFC 44

ORDERS

THE COURT ORDERS THAT:

1.    The appeal be dismissed.

Note:    Entry of orders is dealt with in Rule 39.32 of the Federal Court Rules 2011.

REASONS FOR JUDGMENT

THE COURT:

the facts and the issue

1    This case concerns the estate of Marjorie Lewis, who died on 15 August 2008 (the estate). The biggest asset of the estate was an amount of around $550,000 due to it from the Sir Moses Montefiore Jewish Home (the Montefiore sum).

2    Apart from a legacy of $5,000, the estate fell into residue. The residuary beneficiaries were Marjorie’s brother Allan Lewis and his four sons. Allan and one of his sons, Anthony, were named as executors but renounced their executorship before obtaining probate. Letters of administration were granted to another son, Geoffrey, in January 2009 (at which time the value of the estate was identified as $786,453.31). In these reasons we will continue the practice of referring to members of the Lewis family by their first names.

3    Ten days after Marjorie’s death Anthony, purporting to be the executor of her will, contacted the debtor and procured payment of the Montefiore sum. On 2 September 2008 the Montefiore sum was deposited into an account that Anthony had opened in the name of the estate with an investment company (Lewis Securities), which he controlled and operated. On 9 September 2008 that sum, together with $20,000 from a pre-existing account in Marjorie’s name (the additional sum), was transferred into Lewis Securities’ general trading account. The total amount of misappropriated funds was around $570,000.

4    On 29 October 2008, Lewis Securities went into voluntary administration. On 6 February 2009 it went into liquidation. The liquidators denied a request on behalf of the estate for payment of the total mentioned above and took the position that the estate was an unsecured creditor. The liquidators later paid the estate three separate dividends amounting to around $85,662.74, plus a final dividend of $17,760.75 paid between the hearing and the judgment in the court below.

5    On 2 April 2009, Anthony was declared bankrupt. Geoffrey, as administrator, lodged a proof of debt with Anthony’s trustee in bankruptcy. The trustee ultimately declared a nil dividend and the estate has not recovered any funds from Anthony. Anthony was discharged from bankruptcy in April 2012.

6    On 21 July 2011 Anthony’s trustee in bankruptcy entered into an agreement with the present appellant (HBSY) to sell certain assets including Anthony’s interest in the estate. The purchase price was $275,000. On 9 November 2015 Anthony became the registered owner of all of the share capital of HBSY.

7    The issue in the appeal, in short, is whether HBSY, as Anthony’s assignee, is entitled to receive his share of the estate without the monies misappropriated by Anthony having been repaid to the estate. Central to that issue is the question whether Anthony was released from his obligation to the estate upon his discharge from bankruptcy.

the proceedings so far

8    HBSY commenced proceedings in the Supreme Court of New South Wales by a statement of claim filed on 23 August 2019. It sought the removal and replacement of Geoffrey as trustee. Geoffrey cross-claimed, seeking the following declarations.

1.    If the Court determines that the Cross-Defendant received a valid and enforceable assignment of an interest of Anthony Richard Lewis in the estate of the late Marjorie Lewis who died on 15 August 2008 (“the Estate”) (which is denied), the Cross-Claimant seeks:

(a)    a declaration that the Cross-Defendant is deemed to have received a distribution from the Estate in the sum of $502,554.74 (“the Net Sum”) and interest thereon (or such other amount as the Court may determine);

or further or in the alternative:

(b)    a declaration that the Cross-Defendant cannot participate in or receive a distribution from the Estate without having first paid the Net Sum and interest thereon to the Estate (or such other amount as the Court may determine).

2.    A declaration that the Estate is not liable to pay any legacy to the Cross-Defendant. …

(Emphasis in original.)

9    Geoffrey also obtained an order for security for costs against HBSY. HBSY failed to provide that security, and on 24 April 2020 its proceeding against Geoffrey was stayed until further order. Only the cross-claim therefore came before the primary judge for determination.

10    The primary judge upheld the cross-claim. His Honour made a declaration substantially in accordance with proposed declaration 1(b) above, save that a lesser sum was specified. (While this aspect of the final orders was not the subject of separate reasons, we infer from the reasons at [151]-[152] and [180] that his Honour deducted from the “Net Sum” the amount of Anthony’s entitlement and an amount of $300,000 for which Geoffrey had sought to prove in Anthony’s bankruptcy. We have some reservations concerning these deductions but no complaint was made about them in this Court.) His Honour also formally dismissed HBSY’s proceeding.

11    On 2 September 2022, HBSY filed an application for an extension of time to appeal to this Court. HBSY was out of time in part because its lawyers had prepared the papers for an appeal to the New South Wales Court of Appeal, but then came to the view that the appeal should come to this Court.

12    On 2 and 3 May 2023 we heard argument on all of the issues bearing upon the extension of time application and the substantive appeal. We delivered a judgment on 14 July 2023 (HBSY Pty Ltd v Lewis [2023] FCAFC 109; 298 FCR 303 (HBSY No 1)).

13    We would have granted the extension of time, but for the view that we reached as to the competency of the proceedings (at [15]). We held that s 7(5) of the Jurisdiction of Courts (Cross-vesting) Act 1987 (Cth) did not require the appeal to be commenced in this Court and there was no other potential source of appellate jurisdiction. We therefore held that either the extension application was incompetent (at [54]) or the application should be refused because the substantive appeal would be incompetent (at [55]).

14    HBSY then commenced proceedings in the High Court, seeking constitutional writ relief in relation to the orders of this Court. On 9 October 2024 the High Court granted the relief sought by HBSY (HBSY v Lewis [2024] HCA 35; 98 ALJR 1211 (HBSY HCA)). That relief included a writ of mandamus directing this Court and the judges thereof “to hear and determine [HBSY]’s appeal from the whole of the judgment of the Supreme Court of New South Wales”.

15    The form of this writ required this Court to grant the extension of time to appeal sought by HBSY (since an extension is necessary in order for the appeal to be heard (Federal Court Rules 2011 (Cth) (the Rules), r 36.03) and the High Court did not purport to exercise the power in r 36.05 of the Rules itself). We infer that the writ was framed in this way because it was clear from HBSY No 1 that, but for the jurisdictional issue, we would have granted the extension of time.

16    At a case management hearing on 5 November 2024, the parties were informed that we proposed to determine the appeal on the basis of the documents already filed and the oral submissions made in May 2023. HBSY sought leave to file short supplementary written submissions. Orders were then made on 6 November 2024 granting this leave to both parties and formally extending the time for filing the notice of appeal.

the appeal

The position before Anthony’s bankruptcy

17    HBSY accepts that, from the time when Lewis Securities went into liquidation and the Montefiore sum and the additional sum was lost to the estate, Anthony was liable to the estate in that sum. This was on the basis that, by his intermeddling in the recovery of the Montefiore sum, he had rendered himself an executor de son tort and had dealt with that sum in a way that was manifestly inconsistent with the duties of an executor. His renunciation of the office of executor could not free him from the obligation to make good his default. The primary judge made findings to this effect at [126]-[134], which we do not understand to be challenged and which in any event are clearly correct.

18    Immediately before Anthony became bankrupt, therefore, it is clear that he came within the following statement of principle by the learned authors of Jacobs’ Law of Trusts in Australia (8th ed, LexisNexis Butterworths, 2016) (Jacobs) at [21-27].

A trustee who has committed a breach of trust cannot claim, as against the other cestuis que trust, any beneficial interest in the trust estate until the default has been made good.

19    Several equitable principles, which were canvassed in the submissions before the primary judge and on appeal, support this general statement. These were argued by reference to Re Dacre; Whitaker v Dacre [1916] 1 Ch 344 (Re Dacre), Morris v Livie (1842) 1 Y & C Ch Cas 380; 62 ER 934 (Morris v Livie) and Cherry v Boultbee (1839) 4 My & Cr 442; 41 ER 171 (Cherry v Boultbee). It is useful to note these principles briefly at this point.

20    In Re Dacre, Henry Dacre was appointed executor under the will of JG Womack. The will bequeathed a legacy of £2,000 to Henry’s wife, Alice. She died without having received that legacy. Henry was both the sole executor and beneficiary of her estate. Henry died insolvent, never having proved in Alice’s estate. It was discovered after his death that Henry had misappropriated £1,500 in his role as executor of JG Womack’s estate. The surviving executor of JG Womack’s estate sought a declaration that he was entitled to set off or retain any part of the legacy of £2,000 that had devolved beneficially to Henry, to the extent necessary to satisfy Henry’s liability to make good the misappropriated sum. The executor succeeded at first instance and in the Court of Appeal, where Lord Cozens-Hardy MR said (at 347):

… the true principle, I think, is that which is laid down by Sir George Jessel in Jacubs v Rylance (L.R. 17 Eq. 341), and which is again affirmed by Stirling J. in the case of [Doering v Doering (1889) 42 Ch D 203 (Doering v Doering)], and lastly emphatically affirmed by Parker J. in re Towndrow ([1911] 1 Ch. 662, 666, 668), where during the course of the argument he said this: “The real principle is that where there is an aggregate fund in which the trustee is beneficially interested and to which he owes something, he must be taken to have paid himself that amount on account of his share.” That principle he lays down again in the same judgment where, quoting the words of Stirling J., he says: “The theory on which that rule is based is that the Court treats the trustee as having received his share by anticipation, and the answer to any claim made by the trustee is this: ‘You have already received your share; you have it in your own hands.’” That doctrine has been applied not merely as against the defaulting trustee, but against his assignees for value.

21    Thirteen years earlier, the Full Bench of the Supreme Court of Victoria (also referring to Doering v Doering) had stated a “rule” to much the same effect: Cumming v Austin (1903) 28 VLR 622 at 628-629 (Holroyd J, Willams and Hodges JJ agreeing).

22    In Morris v Livie, the elder Robert Livie bequeathed £4,000 to his brother-in-law, Alexander Champion and his partner, Robert Livie the younger, to be held on trust by them for the sole benefit of his sister, Catherine Primrose (or, in the event of her death, her son). The funds were to be invested in stocks and the dividends paid to Catherine. Alexander Champion died and Robert Livie the younger sold the stocks for his own benefit. Approximately £2,200 of dividends owed to Catherine remained unpaid. The will also bequeathed a £5,000 legacy to the younger Robert Livie which, prior to selling the stock, he had assigned to another person for valuable consideration. Catherine’s executors sought to reclaim what was owed out of the younger Robert’s legacy.

23    It was held that a legacy given to the younger Robert was subject to a condition that he would properly perform his duties as trustee. That condition “if existing, accompanied his legacy until its discharge, and applied to it as much after as before its assignment” (at 389; ER 938). Thus, his right to participate could not be exercised (including by the assignee) until the breach was remedied. Isaacs J later expressed the point in Cock v Aitken (No 2) (1912) 15 CLR 373 (Cock v Aitken) at 384 as follows.

The case of Morris v. Livie… shows that where a legacy is given to a trustee there is a condition implied by law and attaching to the legacy until due, that he shall fulfil his fiduciary obligations. That condition, as Knight-Bruce V.C. says, accompanies the legacy until its discharge, and applies as much to it after as before assignment. That is the risk the assignee runs.

(Emphasis in original.)

24    In Cherry v Boultbee, Catherine Boultbee was a creditor of her brother Thomas, who became bankrupt. She did not prove in his bankruptcy. Catherine died and bequeathed a legacy to Thomas which was greater than the sum he owed to her. Mr Cherry, the assignee of Thomas’s bankrupt estate, claimed the legacy. Catherine’s executors claimed a right to set off the amount owed by Thomas to Catherine against the legacy. The Lord Chancellor (Lord Cottenham) observed that the terminology of “set-off” was inaccurate and continued (at 447; ER 173):

The right of an executor of a creditor to retain a sufficient part of a legacy given by the creditor to the debtor, to pay a debt due from him to the creditor’s estate, is rather a right to pay out of the fund in hand, than a right of set-off. Such right of payment, therefore, can only arise where there is a right to receive the debt so to be paid; and the legacy or fund, so to be applied in payment of the debt, must be payable by the person entitled to receive the debt.

25    The claim by Catherine’s executors failed, however, because Thomas had become bankrupt before she died and the executors were never entitled to claim more from him than the dividend upon the debt. There was never a time when the same person was both entitled to the legacy and liable to pay the entire debt (at 447-448; ER 173).

26    In Re Akerman; Akerman v Akerman [1891] 3 Ch 212 (Re Akerman) at 219-220, Kekewich J described the principle “laid down in Cherry v. Boultbee” in the following way.

A person who owes an estate money, that is to say, who is bound to increase the general mass of the estate by a contribution of his own, cannot claim an aliquot share given to him out of that mass without first making the contribution which completes it. Nothing is in truth retained by the representative of the estate; nothing is in strict language set off; but the contributor is paid by holding in his own hand a part of the mass, which, if the mass were completed, he would receive back.

27    The point was put in very similar terms in Re Peruvian Railway Construction Co Ltd [1915] 2 Ch 144 at 150 (Sargant J) (appeal dismissed: [1915] 2 Ch 442) and in Gray v Guardian Trust Australia [2002] NSWSC 1218 at [108] (Austin J).

28    The primary judge held that, aside from the effect of Anthony’s bankruptcy, each of these principles was engaged. Grounds 1, 2 and 9 of the notice of appeal appear to take issue with these conclusions. However, the gravamen of HBSY’s submissions appears to be that the principles do not operate in the face of s 153(1) of the Bankruptcy Act 1966 (Cth) (the Bankruptcy Act). This was the main area of disagreement in the appeal.

Election

29    Before engaging with the effect of Anthony’s bankruptcy and subsequent discharge on the rights of the estate, we deal with a logically anterior issue: whether Geoffrey in his capacity as trustee is precluded from relying on Cherry v Boultbee, Re Dacre and Morris v Livie by conduct inconsistent with such reliance.

30    In the proceedings below HBSY submitted that, by proving in Anthony’s bankruptcy, Geoffrey communicated an election not to rely on the equitable principles. The primary judge rejected this argument (at [151]). Grounds 3 and 4 in the notice of appeal took issue with this conclusion. However, the argument was withdrawn in HBSY’s written submissions in reply and ground 4 was consequently not pressed.

31    Instead, leave was sought to support ground 3 by way of a new argument: that Geoffrey acted inconsistently with the assertion of any right of retainer by claiming payment of the Montefiore sum and the additional sum from the liquidators of Lewis Securities and receiving dividends in response to that claim. We refused that leave. Our reasons for doing so were as follows.

32    First, the argument was raised very late and without any explanation of why it had not been raised earlier. Geoffrey’s cross-claim, which raised the question of his entitlement to withhold Anthony’s share of the estate, was filed on 23 October 2019, heard by the primary judge in August 2021, judgment delivered on 24 June 2022 and final orders made on 1 July 2022. There is no apparent reason (and none was suggested other than inadvertence) why the point could not have been raised in the trial. Work had begun on an appeal from the judgment by the end of July 2022 and the application for an extension of time to appeal to this Court (with a draft notice of appeal) was filed in September 2022. It was only in the written submissions in reply, filed on 26 April 2023 (around a week before the hearing), that the point first appeared. The nature of the interests involved in the proceeding do not call for any relaxation of the general principle that parties are bound by the way they run their cases at first instance. Hence, even putting aside the next point, it is not “expedient in the interests of justice” (to adopt language used in VUAX v Minister for Immigration and Multicultural and Indigenous Affairs [2004] FCAFC 158; 238 FCR 588 (Kiefel, Weinberg and Stone JJ) at [46] and cited in many subsequent cases) for the new argument to be determined in the appeal.

33    Secondly, although the point was presented by HBSY as an issue of law arising on uncontroversial facts, this was contested by counsel for Geoffrey. Without embarking on a detailed consideration of the point, we are not able to be satisfied that there was not additional evidence that might have been adduced in response if it had been raised at the trial. For example, if the point is one of election, the extent of Geoffrey’s awareness of the existence of alternative remedies may be relevant. Generally at least, a point will not be allowed to be raised for the first time on appeal if evidence could have been led which “by any possibility” could have prevented it from succeeding at first instance: Coulton v Holcombe (1986) 162 CLR 1 at 7-8 (Gibbs CJ, Wilson, Brennan and Dawson JJ).

Effects of Anthony’s bankruptcy

34    The complicating factor is thus Anthony’s bankruptcy and the consequent application of the provisions of the Bankruptcy Act.

35    First, with certain presently irrelevant exceptions, Anthony’s property and rights exercisable by him became property divisible among his creditors (s 116) and vested in his trustee in bankruptcy (s 58). Anthony’s right to share in the distribution of the estate therefore passed to his trustee. That right, as noted earlier, was subsequently assigned for value to HBSY. The validity of the assignment was not challenged.

36    This does not in itself present particular difficulties, because each of the principles referred to above applies equally to a person claiming through the defaulting trustee (including an assignee for value): Re Dacre at 347; Cock v Aitken at 383-384; Jacobs at [21-11] (citing Burridge v Row (1842) 1 Y & C Ch Cas 183; 62 ER 846). HBSY cannot have a better claim against the estate than Anthony did. As the primary judge observed at [142], it took its interest subject to the equitable principles.

37    Secondly, upon discharge from bankruptcy, Anthony was, with some exceptions, freed from all debts provable in the bankruptcy by s 153. That section provides, relevantly, as follows.

153 Effect of discharge

(1)    Subject to this section, where a bankrupt is discharged from a bankruptcy, the discharge operates to release him or her from all debts (including secured debts) provable in the bankruptcy, whether or not, in the case of a secured debt, the secured creditor has surrendered his or her security for the benefit of creditors generally.

(2)    The discharge of a bankrupt from a bankruptcy does not:

…

(b)    release the bankrupt from a debt incurred by means of fraud or a fraudulent breach of trust to which he or she was a party or a debt of which he or she has obtained forbearance by fraud; or

…

38    Debts provable in the bankruptcy are identified by s 82. Relevantly for present purposes, s 82(1) provides as follows.

(1)    Subject to this Division, all debts and liabilities, present or future, certain or contingent, to which a bankrupt was subject at the date of the bankruptcy, or to which he or she may become subject before his or her discharge by reason of an obligation incurred before the date of the bankruptcy, are provable in his or her bankruptcy.

39    Two issues arise in respect of the operation of s 153.

40    The first issue is whether Anthony’s obligation to make good the consequences of his default, or more accurately the right of Geoffrey as administrator to withhold Anthony’s share until he did so, was a “debt … provable in the bankruptcy” to which s 153(1) applies. The second issue is whether, if s 153(1) does apply, the debt was incurred “by means of … a fraudulent breach of trust” within the meaning of s 153(2)(b).

41    As to the first of these issues, the primary judge considered that s 153 “clearly extinguishes liability with respect to the rule in Cherry v Boultbee” (at [153]). His Honour acknowledged that there was less certainty in relation to Re Dacre and Morris v Livie, but then observed that s 153 “operates against debts incurred by breach of trust” and it was therefore necessary to consider s 153(2)(b) (at [154]). These conclusions are challenged by a notice of contention filed by Geoffrey on 16 March 2023.

42    As to the second issue, his Honour held that s 153(2)(b) applied to “fraud in the equitable sense” (at [156]) and that Anthony’s conduct came within this concept (at [158]-[159]). These conclusions are challenged by grounds 5 to 8 in the notice of appeal.

43    Section 86 of the Bankruptcy Act should also be noted at this point, although the primary judge held that it was not applicable in the present case and that holding is not challenged. Section 86 is as follows.

86 Mutual credit and set-off

(1)    Subject to this section, where there have been mutual credits, mutual debts or other mutual dealings between a person who has become a bankrupt and a person claiming to prove a debt in the bankruptcy:

(a)    an account shall be taken of what is due from the one party to the other in respect of those mutual dealings;

(b)    the sum due from the one party shall be set off against any sum due from the other party; and

(c)    only the balance of the account may be claimed in the bankruptcy, or is payable to the trustee in the bankruptcy, as the case may be.

(2)    A person is not entitled under this section to claim the benefit of a set-off if, at the time of giving credit to the person who has become a bankrupt or at the time of receiving credit from that person, he or she had notice of an available act of bankruptcy committed by that person.

The application of s 153(1) in this case

44    By his notice of contention, Geoffrey puts in issue the primary judge’s acceptance that s 153(1) would be an answer to his case but for the exception in s 153(2)(b). For the reasons that follow, we respectfully differ from the conclusions of the primary judge in this respect.

45    It can be accepted that a claim by Geoffrey (as administrator) against Anthony, seeking orders for the payment by Anthony to the estate of any part of the Montefiore sum and the additional sum, could not now succeed as a result of s 153(1). Anthony’s obligation to pay those amounts to the estate was a “debt” that was “provable in the bankruptcy”, from which he was released upon discharge. However, that is not what the present case is about. In the proceeding below Geoffrey cross-claimed, not for payment of any amount owing by Anthony, but for declarations identifying what he claimed was a precondition to the assertion of Anthony’s rights as a beneficiary. The difference is emphasised by the fact that Anthony is not a party to the proceeding. HBSY commenced the proceeding in its capacity as the assignee of Anthony’s rights; no duty is sought to be enforced against him.

Cherry v Boultbee

46    The general principle stated in Cherry v Boultbee (but not applied on the facts of that case) depends on the beneficiary owing the estate money rather than on any default by the beneficiary in the performance of duties owed to the estate. The formulations in Cherry v Boultbee itself and in Re Akerman emphasise that this is not seen as a case of setting off mutual debts; rather, the beneficiary is treated as having already received their entitlement.

47    Nevertheless, it was held in Re Sewell; White v Sewell [1909] 1 Ch 806 (Re Sewell) that no sum could be withheld by the new trustees as against former, defaulting, trustees who had agreed a composition with their creditors (which extinguished their debts). Parker J said (at 808-809):

In my opinion the equity which has been asserted… depends upon the existence of a liability on the part of the person against whom it is asserted to pay money to an estate in the distribution of which he is interested. This liability may be legal or equitable only, and it may or may not be a liability which is enforceable by action. But in my opinion the liability must exist when the equity is asserted. It is not enough that it previously existed and has been released prior to the time when the equity is asserted. …

I do not think it makes any difference whether the liability be to executors for a debt or to trustees for a breach of trust.

48    This reasoning was applied by the Full Court of the Supreme Court of Tasmania in Re Hurburgh; National Executors and Trustees Co of Tasmania Ltd v Hurburgh [1959] Tas SR 25 (Re Hurburgh) at 48-49 (Crawford J, Green J agreeing; see also at 31-32 (Gibson J)). Re Hurburgh is a decision of an Australian intermediate appellate court which no party submitted was wrongly decided. We note, however, that the facts in Re Hurburgh did not involve any breach of fiduciary duty; rather, the question was simply whether the trustees of a deceased estate could retain part of the interests of the testator’s sons who had owed money to the testator (and whose debts had been extinguished under the predecessor of the Bankruptcy Act). We will return to the significance of this point later, but for present purposes we accept that the release of a bankrupt beneficiary from debts provable in the bankruptcy can prevent such a debt from becoming the basis for retention of funds by the trustee pursuant to the broad Cherry v Boultbee principle.

49    Questions of timing can be significant, however. In Cherry v Boultbee itself (where the trustees’ argument failed), the beneficiary had become bankrupt before the death of the testator; and the trustees of her estate therefore never acquired any right to claim more than the dividend on the debt. No right to retain the full amount ever arose. When the testators’ sons in Re Hurburgh became bankrupt, they had only a reversionary interest in the estate; their debts were thus extinguished before there was any legacy from which the trustee of the estate could claim a right to retain anything. Conversely, there is authority that, if the beneficiary becomes bankrupt after a right to retain part of the legacy has vested in the trustee, that right is not affected by the release of the beneficiary from their debts: see eg, Otis Elevator Co Pty Ltd v Guide Rails Pty Ltd (in liq) [2004] NSWSC 383; 49 ACSR 531 (Otis Elevator) at [35]-[39] (Palmer J).

50    This can be explained on the basis that, if the trustee has a pre-existing right to retain a portion of the beneficiary’s share of the estate, that portion never truly becomes part of the beneficiary’s bankrupt estate so as to become divisible among their creditors: Re Melton; Milk v Towers [1918] 1 Ch 37 at 56-57 (Warrington LJ), 60-61 (Scrutton LJ). Thus, if at the time of bankruptcy the beneficiary’s share of the estate is $1,000, but the beneficiary owes the estate $900, the true position is that the beneficiary’s share is worth $100; not that the $1,000 is payable in full while the trustee’s only right is to prove in the bankruptcy. Correspondingly, once the Cherry v Boultbee principle attaches to the debt owed to the trust estate (the $900 in this example), that amount is properly seen as a qualification on the beneficiary’s rights under the trust rather than a separate debt provable in the bankruptcy. This understanding is consistent with the policy reflected in s 86(1) of the Bankruptcy Act.

51    These considerations indicate that the statement in Re Sewell set out above, despite its apparent approval in Re Hurburgh, goes too far. As a matter of fairness and practicality it is hard to see why the liability should be required still to exist (if by that it is meant that it must be enforceable) when the relevant equity is asserted (which may not be possible for years or even decades), as opposed to when the equity arises. The decision in Re Sewell may be more satisfactorily explained on the basis that the new trustees had proved for and accepted the composition and received a dividend, thereby (in his Lordship’s view) assenting to it (at 809). They had, on one view at least, elected to pursue a remedy that was inconsistent with reliance on a right to retain part of the former trustees’ share of the estate. This is how Palmer J appears to have understood the case in Otis Elevator at [48] (see also Re Sirrah Pty Ltd (in liq) [2024] NSWSC 784 at [75], [90] (Black J)).

52    It should not be forgotten that the question ultimately concerns the construction and application of s 153(1). The cases referred to above have relevance because they establish the nature of the right relied on by Geoffrey and (as to the older decisions) because part of the statutory context of s 153(1) is the long history of bankruptcy and insolvency legislation and the case law applying it. The area is one in which common law, equity and statute have been so intertwined for so long that (for example) neither Palmer J in Otis Elevator nor the judges who decided Re Hurburgh found it necessary to set out the terms of the relevant statutory provisions. The better view of the case law is that the “equity” entitling a trustee to treat a debtor/beneficiary as having already been paid to the extent of their debt—and thus retain the amount of that debt when the estate is distributed—is related to but distinct from the debt itself. Section 153(1) should be construed on that basis: the “equity” is not a “debt” that is “provable in the bankruptcy”, within the meaning of s 153(1) (read with s 82(1)); and, if it comes into existence before the debt is extinguished, it survives that extinguishment.

53    This understanding is supported by s 86. It reflects an intention that, where there are mutual monetary obligations, the person who owes money to the bankrupt is not required to pay the whole of their debt to the trustee in bankruptcy and then hope to receive a portion of what they are owed in return. Instead, the obligations are set off and the net debt is either provable or distributable as the case may be. It is difficult to see any sensible reason why the same policy would not have been intended to apply where, instead of arising from “mutual credits, mutual debts or other mutual dealings”, the relevant obligations arise out of a relationship of trustee and beneficiary. Section 153(1) aligns with and gives effect to such a policy if the phrase “debts … provable in the bankruptcy” is construed so as not to include the right of a trustee to withhold part or all of a beneficiary’s share of the estate on distribution to account for a debt owed by the beneficiary to the trustee as such.

Re Dacre and Morris v Livie

54    The position is somewhat clearer in respect of the principles deriving from Re Dacre and Morris v Livie. While Cherry v Boultbee concerns a beneficiary who is also a debtor (for whatever reason), Re Dacre and Morris v Livie concern fiduciary obligations.

55    Equity (as the primary judge observed at [140] and as HBSY appears to have accepted) looks more severely upon a defaulting trustee seeking to claim a beneficial entitlement than a mere debtor. This observation does not in itself resolve any question of statutory construction, but it does point towards the correct understanding of the relevant rights for the purpose of applying the statute.

56    Stated broadly, Anthony’s obligation to “do equity” required, at the very least, that he should not assert any right against the estate unless he had made good his earlier default. What Geoffrey was seeking in the proceeding below was declaratory relief confirming the existence of that impediment or precondition to the enjoyment of Anthony’s rights as a beneficiary (rights which had been assigned—necessarily subject to the equitable principles that bound Anthony—to HBSY).

57    The statements of principle in the cases, set out at [20]-[27] above, make this point in more focused ways.

(a)    Re Dacre holds that a trustee-beneficiary who has misappropriated funds is taken to have been paid their entitlement to that extent. This is a concise way of explaining why in these circumstances equity does not allow the trustee-beneficiary, having already received money from the estate, to insist on being paid again.

(b)    The principle in Morris v Livie (as encapsulated by Isaacs J in Cock v Aitken at 384) establishes a “condition implied by law and attaching to the legacy until due” that the trustee “shall fulfil his fiduciary obligations”. That condition, as Knight-Bruce VC had said in Morris v Livie, “accompanied his legacy until its discharge”.

58    Both principles (if they are separate principles) treat the performance of the defaulting trustee’s fiduciary obligations as a precondition for the assertion of their rights qua beneficiary. To the extent that that precondition can be said to vest a right in the current trustee, it is not in our view something that is properly characterised as a “debt”, let alone one that could be proved in the bankruptcy. As one commentator put it (Derham R, The Law of Set-Off (4th ed, Oxford University Press, 2010), quoted by the primary judge at [54]):

[14.85] One consequence of both [the Re Dacre and Morris v Livie] formulations is that, unlike under Cherry v Boultbee, it should not make any difference that the liability to contribute otherwise ceased to exist before the right to participate became due and payable, for example because of a discharge from bankruptcy, or as a result of a composition or scheme of arrangement. The discharge should not be relevant, because the breach would mean that the right to participate has already been impugned (under Morris v Livie), or it has already been satisfied by a notional payment in advance to the defaulting trustee. Similarly, in the case of Morris v Livie, the fact of proving the debt and receiving a dividend in the defaulting trustee’s bankruptcy should not constitute a waiver of the right to invoke the principle either before or after discharge, since the trustee’s right should still be impugned until such time as the default is remedied in toto. While Re Sewell is authority to the contrary, the point may be made that Parker J did not analyze the case in terms of the defaulting trustee formulations.

(Emphasis in original.)

59    An identical passage appears in the 5th edition of the same work (Oxford University Press, 2025) at [14.106].

60    The facts of Re Sewell involved defaulting trustees (although as Derham observes, Parker J did not approach the case by reference to their obligations as trustees). For reasons outlined above, we do not think the statement in Re Sewell concerning the effect of a release under bankruptcy legislation can be fully accepted even in respect of Cherry v Boultbee. It is more clearly incorrect, with respect, in so far as it purports to qualify the Re Dacre and Morris v Livie formulations; although (as noted above) the actual decision in Re Sewell appears supportable on other grounds. As we have also noted above, Re Hurburgh did not concern a defaulting trustee. It therefore does not stand in the way of the conclusion we have reached concerning the application of s 153(1) in such a case.

Conclusion

61    The declaratory relief sought by Geoffrey was not precluded by Anthony having been released from “debts … provable in the bankruptcy” pursuant to s 153(1), regardless of whether s 153(2)(b) applies. Geoffrey is entitled to succeed on this basis.

The application of s 153(2)(b)

62    The conclusion we have reached on s 153(1) makes it strictly unnecessary to reach a conclusion on s 153(2)(b). However, it is appropriate that we set out our views on the issue.

63    In its defence to Geoffrey’s amended cross-claim, HBSY pleaded that any liability of Anthony to the estate was “extinguished by his discharge from bankruptcy pursuant to s 153 Bankruptcy Act 1966 (Cth)”. Geoffrey’s reply did not plead s 153(2)(b) in response to that allegation. Section 153(2)(b) was raised in Geoffrey’s written submissions in the court below and HBSY objected that it had not been pleaded. Ultimately the primary judge allowed an argument based on s 153(2)(b) to be put only if it was confined to equitable fraud (in the sense that any conflict of interest on the part of a fiduciary gives rise to a fraud in equity) and involved no consideration of Anthony’s state of mind.

64    Ground 5 in the notice of appeal alleges that the primary judge erred in allowing s 153(2)(b) to be raised on this basis. This was in substance an issue of case management, and as such was within the discretion of the primary judge. Although there were clearly arguments in favour of holding Geoffrey strictly to the pleaded case, the course that his Honour took effectively reduced the issue to one of law and minimised the prejudice to HBSY. We do not consider that this course involved any error of principle. We therefore turn to the substantive issue (which is agitated by grounds 6 to 8).

65    As a result of the limited basis on which s 153(2)(b) was permitted to be raised, no case was run to the effect that Anthony had (for example) made any statement he knew to be false or engaged in conscious wrongdoing. Geoffrey can succeed under s 153(2)(b) only if “fraud or fraudulent breach of trust” can be established without proving that the conduct giving rise to the liability involved any particular state of mind on the part of the former bankrupt. The construction of the phrase “fraud or fraudulent breach of trust” is therefore critical in this part of the case.

66    The primary judge approached this issue primarily by reference to Maxwell v Chittick (Maxwell), which he considered binding, although his Honour also set out alternative reasoning against the possibility that this understanding of Maxwell was wrong. (The judgment of the New South Wales Court of Appeal in Maxwell, dated 23 August 1994, was unreported and was later allocated the medium neutral citation [1994] NSWCA 196. In what follows we refer to the version of the Court’s reasons that appears on the AustLII website.)

67    Mr Maxwell, who was a solicitor, suggested to the Chitticks (who were his parents in law) that they build a house on land owned by him and his wife. It was agreed that they could stay on the land as long as they wished and that, if they moved, the Maxwells would pay them the value of their improvements. A deed was executed to reflect this agreement. Without informing the Chitticks (and having assured them at several points that there was nothing to worry about), Mr Maxwell repeatedly mortgaged the land to support his business dealings. Eventually a mortgagee took possession of the land and sold it, so that the Chitticks had to leave. Mr Maxwell became bankrupt and his trustee entered into a composition with his creditors. The Chitticks alleged (among other things) breach of fiduciary duty by Mr Maxwell. One of the issues that arose was whether Mr Maxwell’s liability to the Chitticks survived his composition with creditors on the basis that it was incurred by means of “fraud or a fraudulent breach of trust” within the meaning of s 153(2)(b).

68    Over the course of 12 years or so, Mr Maxwell did several things that could be described as fraud in various senses. Mahoney JA (with whom Priestley and Powell JJA agreed) described his conduct in the following terms at 5 (we have italicised some passages and split some paragraphs in order to emphasise the sequence).

In my opinion, what Mr Maxwell did in his relationships with Mr and Mrs Chittick involved fraud of various kinds. When, in 1979-1980, he induced them to spend money in the erection of a home upon the land owned by him and his wife, he was involved in fraud in the sense of conscious deceit. At the time of the discussion the land was subject to a mortgage.

During the course of the discussion of and the erection of the home, the then existing mortgage was discharged and another mortgage given over the land. These facts were not revealed to Mr and Mrs Chittick. It was obviously in their interest that they be told of the existence of such mortgages. Not to tell them of the mortgages involved, in my opinion, a deceit of them. There was, in the extended or equitable sense, a fraud committed upon them. It is not necessary for present purposes to determine whether, at that stage, Mr Maxwell had the intention of defrauding them in the sense of defeating the interest which he held out to them in respect of their continued residence upon the land and in the house. It may be that it was then his intention to carry out the terms of the mortgage so that they would remain undisturbed in their occupation of the land. But, in my opinion, however that be, the mortgage transaction entered into by him in the context of what occurred involved deceit of them and, in the sense to which I have referred, fraud.

…

The deed was plainly intended to create legal relationships. In that context, it was necessary that they be informed of the existence of the mortgage upon the land. It was deceit of them by Mr Maxwell not to tell them of it. Again, in the sense to which I have referred, there was at that time a further fraud committed.

In addition, there was, in my opinion, the commission of further frauds each time Mr Maxwell, without the consent of the plaintiffs, executed further mortgages upon the land. The fact alone of the creation of fresh mortgages in 1984, 1986, and 1989 involved on each occasion deceit of the plaintiffs.

I infer that, at least by 1987, Mr Maxwell had the subjective intention to defeat the interests of the plaintiffs if that should be necessary. At the time, in 1987 and again in Spring 1988, his representations to them in relation to the visit by the pest inspector and the valuer were to his knowledge false. He made those representations with the purpose of misleading the plaintiffs and diverting their attention from the fact that mortgages existed or were to be granted. There was, in what was then done, conscious fraud and the contemplation that, should his financial circumstances become such, Mr and Mrs Chittick’s rights in relation to the land would be defeated. There was, in my opinion, in respect of these various matters a fraud.

(Emphasis added.)

69    As to the effect of the Bankruptcy Act, Mahoney JA said (at 7-8):

The learned judge concluded, and I agree, that the “debt” owed to the plaintiffs was incurred by means of fraud or a fraudulent breach of trust. As the learned judge has pointed out, the term “fraud” in the bankruptcy provisions in this regard has been given a broad interpretation. It is not necessary to pursue the authorities which were referred to by the judge and in argument before this Court. So much is accepted. The obligations incurred by Mr Maxwell, however they arose and whatever point in time they arose, involved fraud of the relevant kind. The obligations which arose under and by virtue of the original proposal to build were of an equitable nature and, as I have indicated, there was deceit in the nature of fraud at that time. That deceit was repeated at the time the deed was executed. There were, when the various mortgages were entered into, breaches of equitable obligations arising from deceit in the nature of fraud. In my opinion, on each of the relevant occasions, there was fraud within the meaning of the statute. I agree with the judge’s conclusions in this regard.

(Emphasis added.)

70    It is apparent that all of the instances of deceit (in the sense of the Chitticks being deceived) occasioned by Mr Maxwell’s actions were treated as constituting fraud for the purposes of s 153(2)(b), even though only some of them involved fraud in the sense of “conscious deceit”. Those which constituted “fraud in the extended or equitable sense” were included in the “fraud” that was found to have occurred. Mahoney JA plainly did not see a need to isolate a particular act or acts as giving rise to Mr Maxwell’s liability to compensate the Chitticks, evidently because for the purpose of s 153(2)(b) there was no relevant difference between the various acts.

71    The primary judge, at [156], concluded that:

Maxwell is authority for the proposition that fraud in the equitable sense, i.e. not requiring the defendant to have an actual intention to defraud the plaintiff, will meet the description “fraud or a fraudulent breach of trust” for the purposes of s 153(2)(b).

72    His Honour then observed that as an executor de son tort Anthony owed duties to the estate (including not to use his position to obtain a personal benefit) and breached those duties, before concluding (at [159]):

In short, the Court has no difficulty in concluding that Mr Lewis, at the very least, speculated with money that was not his. As I have already observed, the fact that he may have subjectively seen nothing wrong with it or hoped that it would be to the benefit of the Estate is irrelevant. Mr Lewis thereby engaged in conduct which was fraudulent in the equitable sense, and applying Maxwell, the Court finds therefore gave rise to a debt to the Estate which was "incurred by means of fraud or a fraudulent breach of trust to which [Mr Lewis] was a party" for the purposes of s 153(2)(b).

73    Against the possibility that the matter might go further, the primary judge recorded that he would have reached the same conclusion even if Maxwell were not binding (at [160]) and then explained that conclusion by reference to Australian and English authorities.

74    HBSY’s submissions in relation to s 153(2)(b) in the appeal are directed to:

(a)    the proposition of law for which Maxwell stands;

(b)    the proper construction of s 153(2)(b); and

(c)    the primary judge’s characterisation of Anthony’s conduct.

The authority of Maxwell

75    The paragraph extracted from Mahoney JA’s reasons in Maxwell at [69] above contains an important qualification:

It is not necessary to pursue the authorities which were referred to by the judge and in argument before this Court. So much is accepted.

76    We agree with HBSY’s submission that these sentences should be taken to mean that the construction of s 153(2)(b)—at least as to whether “fraud” or “fraudulent breach of trust” require deliberate deception or conscious misconduct—was not contested in the Court of Appeal. The proposition that the primary judge extracted from Maxwell at [156] (quoted at [71] above) is, therefore, one that was uncontested in that case rather than one arrived at by consideration of competing arguments. A proposition of this nature does not bind lower courts: CSR Ltd v Eddy [2005] HCA 64; 226 CLR 1 at [13] (Gleeson CJ, Gummow and Heydon JJ); Bell Lawyers Pty Ltd v Pentelow [2019] HCA 29; 269 CLR 333 at [28] (Kiefel CJ, Bell, Keane and Gordon JJ).

77    Maxwell therefore did not bind the primary judge to hold that any type of “fraud in the equitable sense” was sufficient to engage s 153(2)(b). Nor, so far as this issue is concerned, does Maxwell engage the general rule that this Court should follow a decision of another intermediate appellate court on the construction of Commonwealth legislation unless convinced that it is plainly wrong (as to which see, eg, Hill v Zuda Pty Ltd [2022] HCA 21; 275 CLR 24 at [25] (Kiefel CJ, Gageler, Keane, Gordon, Edelman, Steward and Gleeson JJ)). It is therefore necessary to examine the history of s 153(2)(b) and the cases decided prior to Maxwell.

78    It should also be noted that in the judgment at first instance in Maxwell (Chittick v Maxwell (1993) 118 ALR 728) Young J, after referring to some nineteenth-century instances of debts found to have been incurred by fraud, found “fraud” on the part of Mr Maxwell only “in being involved in mortgaging the land and making misrepresentations to the lenders … when he knew the contrary was the case” (at 740).

79    Nor, so far as we have been able to ascertain, does any other authority in Australia bear directly upon the present issue. Whether “actual fraud” is needed or “constructive fraud” is sufficient for the purpose of s 153(2)(b) was expressly left open by all of the judges who constituted the Full Court of the Supreme Court of Western Australia in Cornelius v Barewa Oil & Mining (NL) (in liq) [1982] WAR 311, a case referred to by the primary judge but not the subject of argument before us. That case involved a preliminary issue concerning pleadings and their Honours were comfortably satisfied that the matters alleged against Mr Cornelius involved fraud on any view: at 313 (Burt CJ), 316-317 (Wickham J), 319-320 (Kennedy J). It is therefore necessary to address the construction of s 153(2)(b) and consider the correctness of the primary judge’s alternative reasoning.

The construction of s 153(2)(b)

80    The task of statutory construction, as has often been observed, must begin and end with the text.

81    Section 153(2)(b) refers, disjunctively, to two concepts: “fraud” and “fraudulent breach of trust”. However, if the adjective “fraudulent” limits the relevant breaches of trust to those involving fraud in a strict sense (actual dishonesty or consciously acting against the beneficiary’s interests), a “fraudulent breach of trust” becomes no more than a subspecies of “fraud”. That construction therefore renders part of the statutory language redundant, which is an indication that it is not the correct construction. All of the words in a statutory provision are to be given effect if possible: eg, Project Blue Sky Inc v Australian Broadcasting Authority [1998] HCA 28; 194 CLR 355 at [71] (McHugh, Gummow, Kirby and Hayne JJ). Textual analysis of the paragraph therefore indicates that “fraudulent”, as a qualifier applied to a breach of trust, has a broader connotation than “fraud”. This is consistent with the broader usage of the term “fraud” in equitable, as distinct from common law, contexts (discussed further below).

82    The text must, of course, be read in context. Part of that context is the history of the relevant provision.

83    Section 153(2)(b) has not been relevantly amended since the Bankruptcy Act was enacted in 1966. Materially identical language (“incurred by means of any fraud or fraudulent breach of trust to which he was a party”) was used in the equivalent provision of the predecessor legislation, s 121(1)(b) of the Bankruptcy Act 1924 (Cth) (the 1924 Act).

84    The 1924 Act was largely based on the Bankruptcy Act 1883 (UK) (the 1883 Act), although it incorporated some innovations from the laws of the various States and adaptations to local conditions (Allsop JLB and Dargan L, “The History of Bankruptcy and Insolvency Law in England and Australia” in Gleeson JT, Watson JA and Peden E, Historical Foundations of Australian Law Volume II: Commercial Common Law (Federation Press, 2013) 415 at 455-456). Section 30(1) of the 1883 Act performed the role now performed by s 153(2)(b) and also used the expression “incurred by means of any fraud or fraudulent breach of trust to which he was a party”. The equivalent provision in the previous United Kingdom (UK) legislation, s 49 of the Bankruptcy Act 1869 (UK) (the 1869 Act) had referred to a “debt incurred by means of fraud or breach of trust” (emphasis added).

85    Jessell MR considered the 1869 provision in Emma Silver Mining Company v Grant (1880) 17 Ch D 122. Mr Grant was the promoter of the plaintiff company and took a secret commission from the vendor of an asset that the company purchased. The company sued Mr Grant and obtained a judgment in respect of the profit he had made. Mr Grant then obtained the benefit of a scheme of arrangement with his creditors under the 1869 Act. When the company sued on its earlier judgment, the question that arose was whether it was caught by s 49 of that Act. Jessell MR concluded at 127-128 that there was “no other word to characterise the transaction except fraud”. His Lordship continued (at 128):

It is said, however, that Mr. Albert Grant did not believe it to be a fraud: that may be so. A man may commit a fraud without believing it to be a fraud, that is, without believing it to be a fraud for which he can be held responsible in law; though I am by no means convinced that even Mr. Albert Grant could have treated this originally as an equitable or moral transaction. But the question is, What has his belief got to do with it? … A man must take the consequences of his acts, and he cannot be allowed in a Court of Justice to say he had failed to estimate correctly either their moral or their legal bearings.

(Emphasis in original.)

86    Pausing here, we observe that this passage does not appear to take matters very far. His Lordship appears to have accepted that Mr Grant might not have realised his actions would be characterised by the law as fraudulent. However, that is a different question from whether those actions were undertaken with an intention that properly attracted the label “fraud”. Mr Grant had secretly arranged with the vendor of the asset to inflate its purchase price so that he personally could profit, and could not have been unaware that this was to the detriment of the company he was supposed to be representing. It could scarcely have been suggested that such conduct was not consciously deceitful, even if Mr Grant did not see it as wrong.

87    Jessell MR also considered whether Mr Grant had committed a “breach of trust”, while noting that it was not necessary to decide the point. His Lordship concluded that those words should be given “their proper technical legal meaning” (at 128-129).

88    The 1883 Act introduced the qualifiers “fraudulent” and “to which he was a party”, in relation to “breach of trust”, which have been picked up in the 1924 and 1966 Acts in Australia (and have also continued to appear in the UK legislation (currently s 281(3) of the Insolvency Act 1986 (UK) (the UK Act))). The primary judge at [79]-[87] advanced the theory that these words were inserted in response to the decision of the Court of Appeal in Cooper v Prichard (1883) 11 QBD 351, which was delivered while the Bill for the 1883 Act was in committee. In the absence of concrete references to relevant extrinsic materials (and detailed argument on the legitimacy of using legislative history as an aid to interpreting UK legislation of this era), this approach appears to us to be problematic. While it may well be correct that some parliamentarians were aware of, and motivated by, the decision in Cooper v Prichard, the search for legislative intention is not an inquiry into the subjective intentions of individual legislators: Mondelez Australia Pty Ltd v Automotive, Food, Metals, Engineering, Printing and Kindred Industries Union [2020] HCA 29; 271 CLR 495 at [98] (Edelman J).

89    Clearly, however, the intention to be inferred from the insertion of these words in the 1883 Act is that not every debt arising from “breach of trust” would be excluded from the general provision for release from debts. The exclusion was only to apply when the bankrupt had been a “party” to the breach (semble not when they had been held liable in respect of a breach by somebody else, such as a partner); and it was only to apply where some feature of the breach justified its description as “fraudulent”. The words that were added should be taken to have been intended to do some work; ie, to make the exclusion more limited than the previous iteration. The Court of Appeal seems to have taken this view of s 30(1) in Re Smith; Hands v Andrews [1893] 2 Ch 1 at 16 (Lindley LJ, Lord Esher MR and Lopes LJ agreeing) (Re Smith). The Court allowed the appeal of a trustee whose failing was to have given money to his co-trustee who misapplied it. (The co-trustee had not been made bankrupt, so that his conduct did not need to be considered for the purposes of s 30(1).)

90    In adopting the same language as s 30(1) of the 1883 Act, the Australian Parliament in enacting s 121(1)(b) of the 1924 Act is presumed to have adopted the interpretation assigned to the earlier enactment in any relevant judicial decisions: Brisbane City Council v Amos [2019] HCA 27; 266 CLR 593 at [24] (Kiefel CJ and Edelman J). The same is true of s 153(2)(b) of the Bankruptcy Act in so far as it adopts the same language as s 121(1)(b). Re Smith provides some guidance as to the meaning of “fraudulent breach of trust” in that it indicates that obtuseness or laxity on the part of a trustee is not within the scope of the paragraph. However, we have not been referred to any decision at appellate level on the 1883 Act or the Australian legislation that provides real guidance as to whether the concept of a “fraudulent breach of trust” involves the trustee having any particular state of mind.

91    The need (on the one hand) to give “fraudulent” some effect within the phrase “fraudulent breach of trust” and the need (on the other hand) to give the phrase itself some effect within s 153(2)(b) can be reconciled by observing that the trust is an equitable concept and the term “fraud” has never been regarded by equity as limited to actual dishonesty. As Lord Haldane LC observed in Nocton v Lord Ashburton [1914] AC 932 at 954:

[W]hen fraud is referred to in … Chancery in describing cases which were within its exclusive jurisdiction, it is a mistake to suppose that an actual intention to cheat must always be proved. A man may misconceive the extent of the obligation which a Court of Equity imposes on him. His fault is that he has violated, however innocently because of his ignorance, an obligation which he must be taken by the Court to have known, and his conduct has in that sense always been called fraudulent, even in such a case as a technical fraud on a power. It was thus that the expression “constructive fraud” came into existence. The trustee who purchases the trust estate, the solicitor who makes a bargain with his client that cannot stand, have all for several centuries run the risk of the word fraudulent being applied to them. What it really means in this connection is, not moral fraud in the ordinary sense, but breach of the sort of obligation which is enforced by a Court that from the beginning regarded itself as a Court of conscience.

(Emphasis added.)

92    With this in mind, a “fraudulent” breach of trust can be understood to mean a breach of trust involving failure to comply with the fiduciary obligations that equity imposes on a trustee, prominent among which are disinterestedness and devotion to the beneficiaries’ interests rather than one’s own. Breach of those standards, which are conceived as matters of conscience, involve an element of moral failing and are relevantly “fraudulent”; they are of a different character from negligence and incompetence. This understanding (which aligns with the approach of the primary judge at [171]-[177]) gives effect to all of the words of the provision, in a manner at least broadly consistent with how the term “fraudulent” has been understood in the courts administering equity in Australia and the UK.

93    We note that the construction proposed in the previous paragraph is inconsistent with the position at which the English courts appear to have arrived in relation to s 281(3) of the UK Act (which, as noted above, uses the same language as s 153(2)(b)).

94    In Mander v Evans [2001] 1 WLR 2378 (Mander) Ferris J considered a claim to recover against a discharged bankrupt certain loans which, it was alleged, had been obtained by undue influence. This was not said to amount to a “fraudulent breach of trust” (at [25]), so that his Lordship appears to have been concerned only with the concept of “fraud”. He held that “fraud” meant “actual fraud in the Derry v Peek 14 App Cas 377 sense” (at [25]). We are inclined to agree with that conclusion and note that it is not in conflict with our preferred construction of “fraudulent breach of trust”.

95    Ferris J considered the concept of “fraudulent breach of trust” in Woodland-Ferrari v UCL Group Retirement Benefits Scheme [2003] Ch 115 (Woodland-Ferrari). A former trustee of a pension scheme was found by the Pensions Ombudsman to have committed breaches of trust in the early 1990s, resulting in a diminution of the assets of the scheme, and directed to repay the shortfall. In the meantime he had been made bankrupt and discharged from bankruptcy. His failure to repay the amount ordered resulted in a statutory demand for the amount, which he applied to have set aside. The source of the liability was thus the Pensions Ombudsman’s determination rather than a curial determination of breach of any fiduciary duty. His Lordship held that “dishonesty” was an “essential ingredient” of a fraudulent breach of trust (at [50]) and that the findings in the Pension Ombudsman’s determination did not establish the requisite dishonesty (at [68]).

96    In reaching the first of these conclusions, Ferris J traced the history of the UK legislation through the nineteenth and twentieth centuries, noting the shift from retributive to rehabilitative objectives that had occurred during that period, particularly in the test for obtaining a discharge from bankruptcy (eg, at [43]). He referred to several cases which included observations concerning fraudulent breaches of trust in the context of applications for discharge, which appear to us to relate only indirectly to the present issue (which concerns the effect of discharge). More generally, where the legislature uses exactly the same words in successive iterations of a legislative regime (each time in a provision performing the same function), that is in our view a strong indication that the effect of the particular provision was intended to remain the same. Broad observations about the evolving social policy objectives of bankruptcy legislation are therefore of limited utility in resolving the present issue of construction.

97    Ferris J then referred to the following passage in Armitage v Nurse [1998] Ch 241 at 250-251 (Millett LJ) (Armitage):

The common law knows no generalised tort of fraud. Derry v. Peek 14 App.Cas. 377 was an action for damages for deceit, that is to say, for fraudulent misrepresentation. In such a case fraud must be proved by showing that the false representation was made knowingly, that is to say, without an honest belief in its truth, or recklessly, that is to say, not caring whether it was true or false. Care needs to be taken when these concepts are applied not to a representation but to a breach of trust. Breaches of trust are of many different kinds. A breach of trust may be deliberate or inadvertent; it may consist of an actual misappropriation or misapplication of the trust property or merely of an investment or other dealing which is outside the trustee’s powers; it may consist of a failure to carry out a positive obligation of the trustees or merely of a want of skill and care on their part in the management of the trust property; it may be injurious to the interests of the beneficiaries or be actually to their benefit. By consciously acting beyond their powers (as, for example, by making an investment which they know to be unauthorised) the trustees may deliberately commit a breach of trust; but if they do so in good faith and in the honest belief that they are acting in the interest of the beneficiaries their conduct is not fraudulent. So a deliberate breach of trust is not necessarily fraudulent. Hence the remark famously attributed to Selwyn LJ by Sir Nathaniel Lindley MR in the course of argument in Perrins v. Bellamy [1899] 1 Ch 797, 798; “My old master, the late Selwyn LJ, used to say, ‘The main duty of a trustee is to commit judicious breaches of trust.’”

(Emphasis in original.)

98    This passage is found in the part of Millett LJ’s reasons in Armitage dealing with the construction of an exclusion clause in a trust settlement, and thus has little if anything to say concerning the construction of the Bankruptcy Act. Elsewhere in his reasons in that case, Millett LJ addressed s 21(1)(a) of the Limitation Act 1980 (UK), which provided that no limitation period applied to an action by a beneficiary in respect of “any fraud or fraudulent breach of trust to which the trustee was a party or privy”, and said (at 260-261):

The first [question] is whether section 21(1)(a) is limited to cases of fraud or fraudulent breach of trust properly so called, that is to say to cases involving dishonesty. The judge held that it is. In my judgment, he was plainly right for the reasons which he gave. I have explained the meaning of the word “fraud” in a trustee exemption clause, and there is no reason to ascribe a different meaning to the word where it appears in section 21(1)(a) of the Limitation Act 1980. Moreover, the meaning of the subsection is not free from authority. Its predecessor, section 26 of the Limitation Act 1939, was held to “mean what it says” and to be limited to cases where fraud was an ingredient of the wrong: see Beaman v. A.R.T.S. Ltd. [1949] 1 K.B. 550, 558, per Lord Greene M.R. The meaning of the words “fraud” and “fraudulent” in section 21(1)(a) is not distorted by the meaning of the expression “concealed fraud” formerly used in section 26 of the Act of 1939 and which was given a very special meaning but has been replaced in the Act of 1980 by the more accurate expression “deliberate concealment”. The result is that in the absence of deliberate concealment liability for an honest breach of trust is statute-barred after six years, but liability for a dishonest breach of trust endures without limitation of time.

99    This aspect of Armitage was referred to by Brereton J in Re Auzhair Supplies Pty Ltd (in liq) [2013] NSWSC 1 at [13] on an issue concerning s 47(1)(a) of the Limitation Act 1969 (NSW).

100    Mander and Woodland-Ferrari were followed by Judge Simon Barker QC (sitting as a Judge of the High Court) in Templeton Insurance Ltd v Brunswick [2012] EWHC 1522 at [46]-[49], holding that “the touchstone in each case” (ie, as to both “fraud” and “fraudulent breach of trust”) is that “actual dishonesty on the part of the bankrupt is a necessary ingredient”. The claims in that case did not involve breaches of a pre-existing trust (although breaches of a director’s fiduciary duties were alleged, which it was noted were capable of giving rise to a constructive trust (at [75])).

101    Woodland-Ferrari is the only one of these decisions that, as a matter of ratio decidendi, is in conflict with the construction of s 153(2)(b) that we prefer. (Armitage addresses a similarly worded provision in a different statutory context.) As noted above, much of the reasoning in Woodland-Ferrari proceeds from a view about the broad themes of bankruptcy and insolvency legislation rather than a focus on the actual words of the provision. That reasoning makes no attempt to give the words “fraudulent breach of trust” any separate work to do. Instead, an argument to the effect that “fraudulent breach of trust” must mean something different from “fraud” was dismissed as having “no substance” (at [45]). We respectfully disagree.

102    For these reasons a “fraudulent breach of trust”, for the purposes of s 153(2)(b), is to be taken to include a breach of those duties of the trustee that have a fiduciary character, which include at least the duty to exercise the trustee’s powers solely in the interests of the beneficiaries and the duty not to engage (without informed consent) in any dealing involving a conflict of interest. Those duties can in some circumstances be breached without any particular state of mind being present. Actual dishonesty, or “fraud” in the common law sense, is not necessarily an element of such a breach.

The breaches of duty in the present case

103    The primary judge described Anthony’s position as follows (at [157]).

In the present case, there can be no doubt that Mr Lewis, as an executor de son tort, held both the Montefiore Sum and the Additional Sum as a trustee who owed duties to the Estate. As such, he was precluded from making his position a means of obtaining a personal profit or benefit or conferring such a benefit on a third party, including not to engage in transactions in relation to which there was a conflict between his personal interest and his duty to the Estate.

104    His Honour then made the following findings in relation to Anthony’s conduct.

(a)    Anthony breached his duties as a trustee by dealing with the Montefiore sum and the additional sum in depositing them into an account with Lewis Securities—a company that was effectively his. It was irrelevant that he was not aware he was breaching his duties or that he hoped the estate would benefit (at [158]).

(b)    Anthony committed a further breach, in the nature of a deceit, by not informing his fellow executor of what he was doing (at [158] and [168]).

(c)    “In short, the Court has no difficulty in concluding that [Anthony], at the very least, speculated with money that was not his” (at [159]).

(d)    The investment was made “in circumstances in which there was a clear conflict of interest” (at [165]-[166]).

(e)    The impropriety of Anthony’s actions “goes beyond mere innocent error or negligence”. On any ordinary view, “when Anthony transferred the money to his company this was tantamount to putting the money in his own pocket” (at [167]).

105    These findings were alleged by ground 6 not to be open to the primary judge on the footing that they were not put to Anthony in cross-examination, and his evidence as to why he had taken certain steps was not challenged or addressed. As developed in submissions, the point appeared to be that these characterisations of Anthony’s conduct were not open to his Honour in the light of the limited grant of leave to raise s 153(2)(b). The argument focused particularly on his Honour’s observation that the impropriety of what occurred went “beyond mere innocent error”, which (as we understood the argument) imported a conclusion as to a state of mind.

106    We reject this submission. In this part of the reasons his Honour was characterising Anthony’s conduct, not attributing motivations to it. Once it was accepted that Anthony had made himself a trustee of the Montefiore sum and the additional sum, it was clearly improper for him to deal with those funds by depositing them with a company that was in substance his. There was plainly a conflict of interest, whether Anthony realised it or not, that made this conduct a breach of his fiduciary duty. This characterisation applied, as his Honour noted, even if Anthony did not perceive his conduct to be wrong or if he was hoping the estate would benefit from it. There was a further breach of duty in not disclosing what he was doing to his fellow trustee.

107    We do not perceive in his Honour’s reasons any finding (explicit or implicit) that Anthony’s conduct went “beyond mere innocent error” on the basis that he had any particular state of mind. His Honour was careful to emphasise that the “impropriety” lay in what was done and not why it was done. The conduct was not “innocent” because it amounted to a failure to adhere to the standards of disinterestedness and conscientious pursuit of beneficiaries’ interests that equity expects of a trustee.

108    The primary facts, upon which these characterisations were based, were not in contest in the appeal. His Honour’s characterisation of those facts was correct. That characterisation was sufficient to make Anthony’s conduct a “fraudulent breach of trust” within the meaning of s 153(2)(b) as correctly construed. For this additional reason, Geoffrey is entitled to succeed.

Disposition

109    The appeal must be dismissed.

110    Ordinarily, having dismissed the appeal we would order that HBSY pay Geoffrey’s costs of the appeal. However, in HBSY HCA the High Court ordered Geoffrey to pay HBSY’s “costs in the Full Court of the Federal Court of Australia and in this Court”. That order appears to cover the whole of the costs of the proceeding in this Court, even though a substantial part of the argument was directed at the substantive issues in the case (which had not been decided and did not reach the High Court). It is not for us to decide that an order of the High Court was made per incuriam. We will therefore make no order as to costs.

Associate:

Dated:    3 April 2025