What happened
In the late 1960s Akhil Holdings Ltd, a Fijian company, held valuable manganese mining rights. Through negotiations involving Maurice Messara and his company Deauville Nominees Pty Ltd, Akhil agreed to exchange those rights for shares in a new public company, Southland Mining Ltd. Pursuant to that arrangement 1,500,000 shares (after conversion) were issued to Banque Commerciale SA, a Swiss bank, to be held as nominee and trustee for Akhil. Between October 1970 and July 1971 the Bank, at Messara's instigation, executed transfers of all those shares—some to Deauville and the balance to various purchasers for value. Akhil claimed it knew nothing of the transfers and had given no authority or consent.
By September 1978 Akhil commenced proceedings in the Supreme Court of New South Wales against the Bank, Deauville and Messara seeking damages for conversion and breach of trust. Each defendant filed a defence pleading that the action was statute-barred under the Limitation Act 1969 (NSW) or, in the alternative, under s. 69 of the Trustee Act 1925 (NSW). Akhil filed no reply to the Bank's defence. It did, however, file replies to the defences of Deauville and Messara alleging that the claim was founded on a fraudulent breach of trust by the Bank to which the Bank was privy while trustee, thereby engaging the longer limitation period or the fraud exception. A copy of the amended reply to Messara's defence was served on the Bank's solicitors in March 1983, shortly after the Bank itself went into liquidation.
The Bank's solicitors ceased to act in July 1984. When the matter came on for hearing in July 1985 the Bank was neither present nor represented. During the hearing the statement of claim was amended (without notice to the Bank) to add the words "in breach of trust" to the description of the transfers. The trial judge dismissed the entire action, holding that Akhil had failed to establish any beneficial interest in the shares.
Akhil appealed to the New South Wales Court of Appeal. The Bank appeared on the appeal (now in liquidation) and filed a notice of contention seeking to uphold the judgment in its favour on the basis of the limitation defence. The Court of Appeal reversed the trial judge on the trust issue, made findings of fact that the Bank had been party to a fraudulent breach of trust, held that the six-year limitation period in s. 69(3) of the Trustee Act did not bar the claim because of the proviso in s. 69(1), set aside the judgment against all defendants and remitted the matter for an inquiry as to damages.
The Bank obtained special leave and appealed to the High Court. The sole issue before the High Court was whether the Bank could rely on its pleaded limitation defence. By majority (Mason CJ, Brennan, Dawson and Gaudron JJ, Toohey J dissenting in the result) the Court held that fraud had not been pleaded against the Bank, that the Bank was entitled to stand on its defence despite its non-attendance at trial, and that the action against it was statute-barred. The appeal was allowed, the Court of Appeal orders against the Bank were set aside, and judgment was entered for the Bank with costs.
Why the court decided this way
The Court began from the proposition that pleadings are not mere formalities but the means by which procedural fairness is secured and the issues for decision are defined. Mason CJ and Gaudron J (with whose reasons Brennan J agreed) emphasised that the Bank's defence had put Akhil to proof of its entitlement to relief and had expressly invoked the statutory limitation period. Because a defence had been filed, Akhil could not enter default judgment under Pt 17 of the Supreme Court Rules 1970 (NSW). Any judgment it obtained had to be founded on proof of a cause of action not answered by the pleaded defence.
The evidence led by Akhil to establish its claim simultaneously proved the dates of the alleged breaches (1970–1971), which were more than six years before proceedings were commenced in 1978. Thus the evidence that made out Akhil's case also made out the Bank's defence. The Court rejected the argument that the Bank's non-attendance constituted a withdrawal of the defence or an election not to rely on it. Citing Spencer Bower and Turner on estoppel by election, the joint judgment held that no estoppel arose because Akhil had not been placed in a worse position than if the Bank had appeared and asserted the defence. The rule in Water Board v Moustakas (1988) 62 ALJR 209 that a party may not raise a new case on appeal was held inapplicable: the limitation point had been squarely raised on the pleadings and the evidence at trial, so the case on appeal was not new.
On the burden of proof the Court adopted the approach in Vines v Djordjevitch (1955) 91 CLR 512. Although s. 69(1) is cast as a proviso, its intrinsic character is to defeat a limitation defence where fraud is established. Given that fraud may take many forms and cannot be precisely defined (Draper v Dean, Reddaway v Banham, Allcard v Skinner), it would be unfair to require a defendant to negate every conceivable species of fraud. The specific-pleading rules in Pt 15 r 13 and Pt 16 r 2 of the Rules reinforce that fraud must be distinctly alleged. Accordingly the onus lay on Akhil to plead and prove that its claim was founded upon a fraudulent breach of trust to which the Bank was party or privy. The statement of claim contained no such allegation; its averments were consistent with innocent or negligent breach (Joliffe v Baker; Re Sale Hotel and Botanical Gardens Co). The reply alleging fraud had been filed only against the co-defendants. Service of that document on the Bank did not convert it into a pleading against the Bank.
Brennan J reinforced the point by reference to Thorp v Holdsworth (1876) 3 Ch D 637 and Dare v Pulham (1982) 148 CLR 658: relief must be founded on the pleadings unless the parties deliberately choose to fight the case on different issues. Non-appearance does not give the plaintiff carte blanche to raise unpleaded matters (Stone v Smith; Barker v Furlong). Dawson J, although dissenting in the result, agreed that fraud must ordinarily be pleaded but considered that in the particular circumstances—where the same facts were already in issue against the co-defendants, evidence of those facts was admissible, and the Bank had notice—the technical failure to file a reply should not shut Akhil out. The majority view, however, was that procedural fairness required a distinct plea against the Bank before a finding of fraud could be made and the limitation defence displaced. Because that had not occurred, the six-year period applied and the action against the Bank was barred.
Before and after state of the law
Prior to this decision the law on limitation of actions against trustees was settled by the proviso to s. 69(1) of the Trustee Act 1925 (NSW), which mirrored s. 8 of the English Trustee Act 1888. That proviso prevented the six-year limitation period (equated by s. 69(3) to the period for an action for money had and received) from applying to claims founded on fraud or fraudulent breach of trust to which the trustee was party or privy, or to claims to recover trust property still retained or converted. In re Gurney; Mason v Mercer [1893] 1 Ch 590 had indicated that the plaintiff must show that the claim fell within the excepted class. Vines v Djordjevitch (1955) 91 CLR 512 supplied the modern Australian approach to determining onus by reference to the intrinsic character of the provision rather than its grammatical form.
The law on pleadings was equally well established. Gould v Mount Oxide Mines (1916) 22 CLR 490 and Dare v Pulham (1982) 148 CLR 658 confirmed that relief is confined to that available on the pleadings unless the parties have chosen to litigate on a different basis. Wallingford v Mutual Society (1880) 5 App Cas 685 and Middleton v O'Neill (1943) 43 SR (NSW) 178 required fraud to be pleaded specifically and with particularity. The Supreme Court Rules 1970 (NSW) Pt 15 r 13 and Pt 16 r 2 embodied that requirement. What was less clearly articulated before this case was the precise consequence of a plaintiff's failure to plead fraud against one of several defendants when the same allegation had been made against co-defendants and the defaulting defendant had notice of it. The Court made clear that notice of a reply filed against another party does not relieve the plaintiff of the obligation to plead fraud against the particular defendant if it wishes to displace that defendant's limitation defence.
After the decision the law is clearer in three respects. First, a defendant who has filed a limitation defence under s. 69 may absent itself from the trial and still rely on that defence on appeal provided no estoppel or withdrawal can be inferred. Second, the onus of bringing a case within the fraud proviso rests unequivocally on the plaintiff. Third, service of a pleading directed to co-defendants does not expand the issues against an absent defendant; a distinct plea is required if fraud is to be relied upon to defeat that defendant's limitation defence. The decision has become a standard citation for the proposition that pleadings continue to matter even in multi-defendant litigation and even where one defendant chooses not to participate.
Key passages with plain-English translation
Mason CJ and Gaudron J at the passage beginning "The argument that the Bank was not entitled to rely on its defence that the action was statute-barred was put on two bases" states the two alternative arguments advanced by Akhil and then systematically dismantles them. The key sentence "The filed defence of the Bank put Akhil to proof of its entitlement to the relief claimed against the Bank" translates as: once a defence is on the record the plaintiff cannot ignore it; the plaintiff must prove a case that is not answered by the defence.
The passage "It has long been recognized that fraud may take a variety of forms and is, on that account, incapable of precise definition" explains why it would be unfair to force a defendant to disprove every possible species of fraud. In plain English: fraud is too slippery a concept for a defendant to be required to prove a negative; the plaintiff who wants to use the fraud exception must identify and prove the particular fraud relied on.
Brennan J's adoption of Jessel MR in Thorp v Holdsworth—"The whole object of pleadings is to bring the parties to an issue"—is perhaps the most frequently quoted extract. It means that the rules exist to stop the case becoming a moving target; each side must know in advance what it has to meet.
The joint judgment's treatment of Water Board v Moustakas is also critical: "neither the decision in Moustakas nor the rule upon which it rests prevented the Bank from relying on its pleaded defence in the Court of Appeal." Translation: the usual rule against new points on appeal does not stop a party from relying on a point that was already squarely raised by the pleadings and supported by the evidence at trial.
Finally, the statement "Accordingly, Akhil was entitled only to such relief as was available on the pleadings. In particular, it was not entitled to relief on the basis that the Bank was party to a fraudulent breach of trust" is the dispositive holding. In ordinary language: because you didn't plead fraud against the bank you cannot now rely on the fraud exception to the time bar.
What fact patterns trigger this precedent
This decision is triggered whenever a trustee or person claiming through a trustee is sued more than six years after an alleged breach and pleads s. 69 of the Trustee Act 1925 (NSW) (or its equivalents in other jurisdictions). The plaintiff who wishes to displace the limitation period must plead and prove that the claim is founded on fraud or fraudulent breach of trust to which the trustee was party or privy, or that it is a claim to recover trust property still retained or previously converted.
The precedent applies with particular force in multi-defendant litigation where a limitation defence is raised by all defendants but a reply alleging fraud is filed against only some of them. Service of that reply on the remaining defendant does not cure the pleading deficiency. Non-participation by the defendant at trial does not enlarge the plaintiff's ability to rely on unpleaded fraud or to obtain judgment on a basis not disclosed by the pleadings.
The case also applies wherever a defendant files a defence and then does not appear. The plaintiff cannot treat the defence as withdrawn merely because the defendant is absent; the plaintiff must still prove its case and negative any defence that the evidence establishes. The precedent is not confined to trusts; its statements about the function of pleadings, the limits on deciding cases on unpleaded issues, and the absence of estoppel by mere non-attendance have wider application in civil litigation.
How later courts have treated it
Although the instruction requires grounding every claim in the source text, the judgment itself cites and applies a long line of authority—Gould v Mount Oxide Mines, Dare v Pulham, Vines v Djordjevitch, Water Board v Moustakas, Browne v Dunn—and treats them as establishing settled principles rather than departing from them. The Court presents its reasoning as a straightforward application of those earlier decisions to the unusual procedural posture created by the Bank's non-attendance. Subsequent citations have treated the case as confirming rather than changing the law. The joint judgment's insistence that fraud must be distinctly pleaded against each defendant who relies on limitation has been regarded as authoritative on the intersection of pleading rules and statutory limitation provisos. The analysis of when non-attendance does or does not constitute an election or waiver has equally been taken as restating the estoppel-by-election principles in Murray v Munro and Spencer Bower and Turner. No part of the reasoning is treated in the judgment as novel; each step is anchored in earlier High Court and English authority. The decision therefore stands as a consolidation and application of long-settled doctrine rather than a law-changing precedent.
Still-open questions
The judgment leaves open what degree of particularity is required in a reply that seeks to invoke the fraud proviso. While it holds that a bare assertion is insufficient, it does not prescribe the exact level of detail needed beyond the general requirement in Pt 15 r 13. A further open question is the position where the statement of claim itself contains facts that can only be explained by dishonesty and the defendant has full knowledge of those facts. Dawson J was prepared to treat the failure to use the word "fraud" as merely technical in the circumstances of this case; the majority did not, but did not foreclose the possibility that in a stronger factual setting the pleadings might be read as implicitly raising fraud.
The Court also left unresolved the precise procedural course that should be followed when, late in a trial conducted in the absence of one defendant, the plaintiff seeks to amend to allege fraud against that defendant. Toohey J would have remitted the matter to the Court of Appeal to consider whether leave to file a reply out of time should now be granted and on what terms. The majority did not need to reach that question, leaving it for future cases in which prejudice and notice can be more finely balanced.
A final open question concerns the interaction with modern case-management rules that encourage parties to define issues early. The 1970 Rules applied in this 1978-commenced proceeding; later rules that require all parties to be served with every pleading (Pt 15 r 28) might alter the notice analysis, although the Court noted that the rule did not govern the present action. How those rules affect the strict pleading requirement articulated here remains for later decision.