Iacobucci J. (dissenting)
Iacobucci J. (dissenting) -- This appeal involves the question of when a person will be liable for the knowing receipt of trust property.
1. Facts
Abraham Rosenberg (“the testator”) died on March 30, 1985. He named his son, Maurice Rosenberg (“Rosenberg”), and his grandson, the appellant Jeffrey Gold, as executors and equal beneficiaries of the residue of his estate. The assets of the estate consisted primarily of commercial real estate held by two companies: Primary Developments Ltd. and Existing Enterprises Ltd. (the “estate companies”).
Both during the testator’s life and after his death, Rosenberg was closely involved in the running of the estate companies. Gold, however, has had no involvement in the estate’s business.
In 1985, shortly after the testator’s death, Rosenberg asked Gold to sign a general power of attorney which would permit Rosenberg to continue his management of the estate companies. Gold agreed.
Rosenberg had other commercial interests besides the estate companies. Together with his wife, he owned all of the shares of Trojan Self-Storage Mini-Warehouse Ltd. (“Trojan”).
The testator, the estate companies, Rosenberg and Trojan all banked at the respondent Toronto-Dominion Bank. Overseeing all of these accounts was a single account manager, Kenneth Slack. Mr. Slack was familiar with the details of the testator’s will and had a copy of Gold’s power of attorney.
In July of 1989, the Bank was trying to obtain repayment of a $300,000 loan made to Rosenberg personally and of a US $130,000 loan made to one of Rosenberg’s companies. At the same time, Rosenberg wanted to secure substantial new loans, totaling approximately $3.9 million. He and Mr. Slack eventually settled upon a mutually satisfactory arrangement which would use the assets of the two estate companies to repay Rosenberg’s personal indebtedness and to secure the new loan. The agreement included the following particulars, relevant to the present appeal:
The Bank agreed to make the $3.9 million loan to Trojan on the condition that the Bank receive a $1.2 million guarantee from one of the estate companies. This guarantee would be supported by the following:
(i) a $1.2 million second collateral mortgage over property owned by Primary, at 156 Columbia Street, Waterloo; and
(ii) a postponement of a $200,000 mortgage held by Existing over a property owned by Trojan, in favour of a new $4 million mortgage to be given by Trojan to the Bank.
A Kitchener law firm, Sills, Madorin, which was a defendant at trial (and which did not appeal the judgment against it), provided legal counsel to the following persons: the estate; Primary; Existing; Rosenberg; Trojan; and, on certain matters, the respondent Bank.
Putting the loan agreement into effect required Gold’s signature on certain documents, most notably on a Primary directors’ resolution, authorizing the guarantee. On July 28, 1989, the law firm prepared a Resolution of the Directors of Primary and also drew up the form for a $1.2 million guarantee. Rosenberg signed both documents on July 28, 1989. Gold signed the Directors’ resolution sometime between September 13, 1989 and October 20, 1989.
In connection with the guarantee, the law firm sent an opinion letter, dated August 1, 1989, to the Bank. The letter said:
The authorization, execution, issuance and delivery of the said Guarantee by the Corporation does not conflict with or contravene any terms, conditions or provisions of any law or agreement to which the Corporation is subject or to which the Corporation is a party.
In a cover letter to the opinion letter the law firm indicated that the Directors’ resolution had to be signed by both Gold and Rosenberg and that Gold had not yet signed it.
Nonetheless, the Bank proceeded to advance its new loans to Trojan. Gold’s signature was subsequently obtained on the Directors’ resolution.
In November 1989, Gold revoked the power of attorney executed in favour of Rosenberg. On January 14, 1993, Gold issued a statement of claim against Rosenberg, Primary, the Bank and the law firm of Sills, Madorin. Gold sought a declaration that the $1.2 million guarantee given to the Bank by Primary was invalid and unenforceable. In the alternative, Gold claimed indemnity for any loss which he might suffer if the guarantee were enforced. The Bank cross-claimed against Primary, seeking enforcement of the guarantee.
2. Judgments Below
A. Ontario Court of Justice (General Division), [1993] O.J. No. 2994 (QL)
In the view of Haley J., the Bank’s liability turned on whether or not it had knowingly assisted in a breach of trust. Citing MacDonald v. Hauer, [1977] 1 W.W.R. 51 (Sask. C.A.), the trial judge said that, in order to recover damages for knowing assistance, the plaintiff must prove the following:
(1) assistance by the defendant of a nominated trustee;
(2) with knowledge;
(3) in a dishonest and fraudulent design on the part of the nominated trustee.
With regard to the third element above, Haley J. explained that the “dishonest and fraudulent design” need not amount to a crime. Rather, the test is whether equity would regard the plan as “morally reprehensible” (para. 12).
Turning to the facts of the case, Haley J. found that the testator’s will created a trust and that Rosenberg was one of the trustees. She also found that the relationship between Rosenberg and Gold was such as to impose upon Rosenberg fiduciary duties (at para. 13):
Rosenberg, acting under a Power of Attorney given to him by Gold for his executorship, owed a fiduciary duty to Gold under the Power of Attorney and also to him as a 50% beneficiary of the estate.
The trial judge went on to hold that Rosenberg had breached his fiduciary duty in three ways:
(1) By causing Primary to give a $1.2 million guarantee in favour of the Bank;
(2) By causing Primary to secure its guarantee by a second collateral mortgage against its property at 156 Columbia Street, Waterloo; and
(3) By causing Existing Developments to postpone its mortgage, in the amount of $200,000, on a Trojan property, in favour of a collateral second mortgage held by the Bank.
In the opinion of the trial judge, these transactions were “fraudulent and dishonest” and, therefore, the case fell within the law of knowing assistance.
Turning next to the issue of the Bank’s liability, Haley J. held that, by making the necessary arrangements for the guarantee-related transactions, the Bank assisted Rosenberg in his fraudulent and dishonest scheme. Thus, all that remained to be determined was whether the Bank had offered this assistance with the requisite degree of knowledge so as to give rise to liability. The trial judge said (at para. 14):
The crucial element to be considered is whether the bank at the time it assisted had knowledge, either actual or imputed that:
(a) the scheme was dishonest and fraudulent;
(b) it was assisting Rosenberg in perpetrating the dishonest and fraudulent scheme.
In carrying out this inquiry, the trial judge focused on Kenneth Slack, the Toronto-Dominion account manager. In particular, Haley J. highlighted the following evidence:
(1) Slack was familiar with the details of the testator’s will.
(2) He knew that Gold played no role in the management of the estate.
(3) He knew that Gold had executed a Power of Attorney in favour of Rosenberg.
(4) He knew that the guarantee and its related transactions could not possibly benefit Gold.
Based on all of the above, the trial judge concluded that the Bank had actual knowledge of Rosenberg’s fraudulent and dishonest breach of trust. She said (at para. 66):
On the basis of all of the knowledge that Slack had in July 1989, when the guarantee was given, and hence that the bank had, I find that the bank knew that what Rosenberg as Trustee was doing, was dishonest and fraudulent. . . .
Accordingly, Haley J. imposed a constructive trust on the Bank in favour of Gold, saying (at para. 75):
. . . I find that equity should fix the bank with a constructive trust in favour of Jeffrey Gold of those assets of the estate representing his 50% beneficial interest by declaring that the $1.2 million guarantee given to the Bank by Primary is unenforceable as are the collateral mortgage on 156 Columbia Street in favour of the bank and the postponement of the Existing third collateral mortgage on 555 Fairway Drive which were given as part of the guarantee transaction.
B. Ontario Court of Appeal (1995), 25 O.R. (3d) 601
B. Ontario Court of Appeal (1995), 25 O.R. (3d) 601
Like the trial judge, the Court of Appeal proceeded on the basis that liability turned solely on whether or not the defendant Bank had knowingly assisted Rosenberg in a fraudulent and dishonest breach of trust. Writing for a unanimous court, Laskin J.A. applied the test for knowing assistance laid down in Air Canada v. M & L Travel Ltd., [1993] 3 S.C.R. 787. According to Air Canada, in order to impose liability on the Toronto-Dominion Bank for knowing assistance, the plaintiff would have to prove the following:
(1) That Rosenberg was a trustee of the property of Primary;
(2) That, in causing Primary to give the guarantee, Rosenberg committed a fraudulent and dishonest breach of trust;
(3) That the Bank participated in the giving of the guarantee; and
(4) That the Bank knew of or was wilfully blind to Rosenberg's fraud and dishonesty.
Laskin J.A. agreed with the trial judge that the plaintiff had proved the first and third elements of its case. However, with regard to whether or not Rosenberg had committed a fraudulent and dishonest breach of trust, the Court of Appeal disagreed with the holding of the trial judge. In reaching this conclusion, the Court of Appeal relied on the fact that Gold had eventually signed the directors’ resolution which authorized the guarantee. By signing the resolution, Gold effectively agreed to give the guarantee, thereby consenting to the breach of trust. This consent, if validly given, would negate any finding of a dishonest breach of trust. Laskin J.A. wrote (at p. 606):
If a beneficiary validly consents to a breach of trust before it is carried out, the beneficiary cannot claim compensation from the trustee for any subsequent loss. . . . This principle logically implies that a beneficiary’s consent will negate a finding of a dishonest breach of trust.
The Court of Appeal examined the facts surrounding the signing of the directors’ resolution and concluded that Gold understood the nature of the guarantee and was aware of the risk it posed to his share of the estate. Laskin J.A. said (at pp. 607-8):
[T]he record demonstrates that Gold knew what a guarantee was, he knew the reason for this guarantee and he knew the possible consequences of authorizing it. He was not misled about the purpose, the effect or the risk of giving his approval. Therefore Gold’s consent was valid.
Gold’s consent, since valid, precluded a finding that Rosenberg had committed a fraudulent and dishonest breach of trust. Therefore, the case did not fall within the scope of knowing assistance.
The above holding (i.e. that Rosenberg had not committed a fraudulent and dishonest breach of trust) was sufficient to dispose of Gold’s claim against the Bank. Nevertheless, Laskin J.A. proceeded to address the issue of whether or not the Bank had knowledge of the breach of trust.
In the opinion of the Court of Appeal, the trial judge had manifestly erred in concluding that the Bank had actual knowledge of the breach of trust. Specifically, Laskin J.A. held that the trial judge had failed to give sufficient weight to the Sills, Madorin opinion letter, which stated that the guarantee was valid. Laskin J.A. said (at p. 610):
[The opinion letter] undoubtedly provided comfort to the bank that the guarantee was not tainted by fraud. However, the trial judge did not refer to the opinion letter.
Given the content of the opinion letter, the Court of Appeal concluded that, contrary to the finding of the trial judge, the Bank neither knew of nor was wilfully blind to Rosenberg’s fraudulent and dishonest breach of trust. Laskin J.A. said (at p. 610):
Even if Rosenberg had misled Gold about the effect of the guarantee, the bank was not privy to their discussion and there is no other evidence that the bank knew Gold had been misled. Absent such knowledge the bank can divorce itself from Rosenberg’s dishonesty.
Accordingly, the Court of Appeal allowed the appeal and dismissed Gold’s claim against the Bank with costs.
3. Issues
(2) Is the respondent Bank liable under the doctrine of knowing receipt?
(1) Is the respondent Bank liable under the doctrine of knowing assistance?
(2) Is the respondent Bank liable under the doctrine of knowing receipt?
4. Analysis
At the outset, I should note that I have read the reasons of my colleague La Forest J. in Citadel General Assurance Co. v. Lloyds Bank Canada, [1997] 3 S.C.R. 805, and I agree generally with his approach regarding liability for knowing receipt. Like him, I believe our reasons take a similar approach.
A person who has not been appointed as a trustee may, under certain circumstances, attract the liabilities of trusteeship. In Barnes v. Addy (1874), L.R. 9 Ch. App. 244, Lord Selborne L.C. explained that there are three situations in which a breach of trust may give rise to liability in a person who is a stranger to the trust. First, a person may be liable as a trustee de son tort. The facts of this case do not require consideration of this category of liability. Second, a person will be liable if he or she knowingly assisted in a fraudulent and dishonest breach of trust. This type of liability is referred to as “knowing assistance”. And third, depending upon considerations of notice, equity may impose liability if the defendant received, in his or her own right, property obtained through breach of trust. This last category of liability is referred to as “knowing receipt”.
This Court has expressed its approval of the Barnes v. Addy classification system, most recently in Air Canada, supra (at p. 810):
In addition to a trustee de son tort, there were traditionally therefore two ways in which a stranger to the trust could be held personally liable to the beneficiaries as a participant in a breach of trust: as one in receipt and chargeable with trust property and as one who knowingly assisted in a dishonest and fraudulent design on the part of the trustees. The former category of constructive trusteeship has been termed “knowing receipt” or “knowing receipt and dealing”, while the latter category has been termed “knowing assistance”.
Knowing assistance and knowing receipt share certain factual similarities; however, they are distinct heads of liability. Confusion has developed in the case law on account of a failure to distinguish between these separate forms of liability. The main source of confusion stems from disagreement over the degree of “knowledge” required for liability in each category. As Charles Harpum writes in his article “The Stranger as Constructive Trustee” (1986), 102 L.Q.R. 114, at p. 120, although a degree of knowledge is required for liability under both heads of liability:
[t]he cases reveal a sharp difference of opinion as to the degree of cognisance that is required before a person may be held liable as a constructive trustee and as to whether the degree of knowledge required under one head is necessarily applicable to another.
In the present case, the appellant argues that the respondent is liable under both knowing assistance and knowing receipt. I will consider each head of liability in turn. I will pay particular attention to the degree of knowledge required to justify liability in each category.
Is the respondent Bank liable under the doctrine of knowing assistance?
This Court reviewed the law of knowing assistance in Air Canada. In that case, we adopted the definition of “knowing assistance” given in Barnes v. Addy, where Lord Selborne L.C. stated that a stranger to the trust will be liable if he or she “assist[s] with knowledge in a dishonest and fraudulent design on the part of the trustees” (p. 252).
A “dishonest and fraudulent design” includes “the taking of a knowingly wrongful risk resulting in prejudice to the beneficiary”. As was said in Air Canada (at p. 826):
I would therefore “take as a relevant description of fraud ‘the taking of a risk to the prejudice of another’s rights, which risk is known to be one which there is no right to take’.”
As the name “knowing assistance” implies, the plaintiff must prove not only that the breach of trust was fraudulent and dishonest, but also that the defendant participated knowingly in that breach of trust. As stated in Air Canada (at p. 811):
The knowledge requirement for this type of liability is actual knowledge; recklessness or wilful blindness will also suffice.
As Millett J. explained in Agip (Africa) Ltd. v. Jackson, [1990] 1 Ch. 265, at p. 292, liability is imposed, in cases of knowing assistance, on the basis that the defendant has participated in a fraud and “participation”, in its relevant sense, implies actual knowledge of the fraud being perpetrated by the rogue trustee. La Forest J. reached a similar conclusion in Citadel where he described liability in knowing assistance as “fault-based” liability (at para. 46). Thus, the basis of liability, participation in a fraud, supports the test for liability which I set out in Air Canada, actual knowledge of the trust and its fraudulent breach.
In the present case, in order to recover damages for knowing assistance, the appellant must demonstrate the following:
(1) That there was a trust;
(2) That the named trustee, Rosenberg, perpetrated a dishonest and fraudulent breach of trust; and
(3) That the respondent Bank participated in and had actual knowledge of Rosenberg’s dishonest and fraudulent breach of trust.
I should note at the outset that both parties agree that there was, indeed, a trust.
Assuming without deciding that Rosenberg committed a dishonest and fraudulent breach of trust and assuming without deciding that the Bank participated in that breach of trust, in my opinion, the appellant’s claim in knowing assistance fails because of the failure to prove that the Bank had actual knowledge of Rosenberg’s fraud. As mentioned above, the Bank received an opinion letter from the law firm of Sills, Madorin. The closing paragraph of the letter stated:
The authorization, execution, issuance and delivery of the said Guarantee by the Corporation does not conflict with or contravene any terms, conditions or provisions of any law or agreement to which the Corporation is subject or to which the Corporation is a party.
In light of this letter, in my view, it cannot be said that the Bank had actual knowledge that the guarantee was obtained in breach of trust. As Laskin J.A. said (at p. 610), this letter “undoubtedly provided comfort to the bank that the guarantee was not tainted by fraud” and I agree with the Court of Appeal that the trial judge made a manifest error in holding to the contrary.
For this reason, in my opinion, the appellant’s claim in knowing assistance should be dismissed.
I will next consider whether the respondent is liable for the knowing receipt of trust property.
Is the respondent Bank liable under the doctrine of knowing receipt?
In a knowing receipt case, the plaintiff sues to recover his or her property which has come into the possession of the defendant, as a result of a breach of trust. As Lord Selborne L.C. said in Barnes v. Addy, the defendant has “receive[d] and become chargeable with some part of the trust property” (at pp. 251-52). However, the mere fact of receipt, of possession, is not a sufficient condition for liability. To be “chargeable” with trust property, a defendant must have received it in his or her own right, must have enjoyed the property beneficially. There is, thus, no cause of action in knowing receipt against a person who holds trust property merely as an agent for a third party. This was the rationale for dismissing the plaintiff's knowing receipt claim in Air Canada. As mentioned therein (at pp. 810-11):
The . . . category of “knowing receipt” of trust property is inapplicable to the present case because it requires the stranger to the trust to have received trust property in his or her personal capacity, rather than as an agent of the trustees.
See also Agip (Africa) Ltd., at p. 288.
The essence of a knowing receipt claim is that, by receiving the trust property, the defendant has been enriched. Because the property was subject to a trust in favour of the plaintiff, the defendant's enrichment was at the plaintiff's expense. The claim, accordingly, falls within the law of restitution. As Denning J. said in Nelson v. Larholt, [1948] 1 K.B. 339, at p. 343:
The right here is not peculiar to equity or contract or tort, but falls naturally within the important category of cases where the court orders restitution....
Similarly, in Royal Brunei Airlines Sdn. Bhd. v. Tan, [1995] 3 W.L.R. 64 (P.C.), at p. 70, Lord Nicholls of Birkenhead stated, “Recipient liability is restitution-based”. I note that La Forest J. reached a similar conclusion in Citadel, where he described liability in knowing receipt as “receipt-based” liability (at para. 46). Therein lies a fundamental difference between the categories of knowing assistance and knowing receipt. Participation in a fraud underlies liability in cases of knowing assistance; unjust enrichment is the essence of a claim in knowing receipt. In Agip (Africa) Ltd., Millett J. distinguished between the two heads of liability (at pp. 292-93):
Tracing claims and cases of “knowing receipt” are both concerned with rights of priority in relation to property taken by a legal owner for his own benefit; cases of “knowing assistance” are concerned with the furtherance of fraud.
As with knowing assistance, the plaintiff must prove a certain degree of knowledge on the part of the defendant to justify liability in knowing receipt. Unlike knowing assistance, where a clear test for the requisite level of knowledge has been set out by this Court in Air Canada, the case law does not clearly set out the degree of knowledge required to justify liability in cases of knowing receipt. Harpum, supra, states (at p. 267):
If a trustee transfers trust property [in breach of trust], the recipient must take that property subject to the trusts unless . . . the recipient is a bona fide purchaser without notice of that fact.
However, the exact meaning of “notice” has occasioned some disagreement in the jurisprudence.
Two main schools of thought emerge regarding the level of knowledge required to justify liability for knowing receipt. The first holds that a defendant will be liable under knowing receipt only if he or she received the property with actual knowledge (and this includes wilful blindness) of the breach of trust. For example, in Carl Zeiss Stiftung v. Herbert Smith & Co., [1969] 2 Ch. 276, Sachs L.J. said (at p. 298):
[The plaintiff must prove] both actual knowledge of the trust’s existence and actual knowledge that what is being done is improperly in breach of that trust. . . .
This view of the law appears to have found approval in In re Montagu's Settlement Trusts, [1987] 1 Ch. 264, where Megarry V.-C. held that actual knowledge, wilful blindness or recklessness were required for liability in knowing receipt cases.
An opposing school of thought holds that a court may impose liability for knowing receipt even if the defendant acted with something less than actual knowledge of the breach of trust. According to this approach, the defendant cannot retain the property in question if it was acquired in circumstances which would have alerted a reasonable person to the possibility of a breach of trust. In Nelson v. Larholt, Denning J. explained (at p. 343):
[I]f the circumstances were such as to put a reasonable man on inquiry, and he [i.e. the defendant] made none, or if he was put off by an answer that would not have satisfied a reasonable man, . . . then he is taken to have notice. . . .
To similar effect was the decision of this Court in Carl B. Potter Ltd. v. Mercantile Bank of Canada, [1980] 2 S.C.R. 343, where Ritchie J. quoted with approval from the decision of the Court of Appeal (at p. 347):
[The defendant] had sufficient notice of the unusual nature of the . . . funds to put [it] on its inquiry to determine the exact nature of these funds before dealing further with them.
And, in Cartwright v. Lyster, [1934] 2 D.L.R. 166 (Ont. C.A.), Middleton J.A., writing for the majority, held that the defendant was liable because it “had knowledge of the facts and circumstances that place them [sic] upon inquiry” (p. 169). Finally, in Canadian Pacific Air Lines Ltd. v. Canadian Imperial Bank of Commerce (1987), 27 E.T.R. 281 (Ont. H.C.), Maloney J. imposed liability on the grounds that the circumstances were such as to put the defendant “on alert”.
In my opinion, this latter approach is the preferable one as it best suits the restitutionary basis of a claim in knowing receipt. Harpum states (at p. 273):
Because the issue in cases of knowing receipt is essentially a proprietary one, a recipient of trust property may be liable as a constructive trustee if he failed to make the inquiries that he ought to have made, even though he acted in good faith. It is taken for granted in the cases that constructive notice of the impropriety of the transfer suffices for liability . . . .
A stranger in receipt of trust property is unjustly enriched at the expense of the trust beneficiary. Participation in a fraudulent breach is irrelevant to the plaintiff’s claim. Liability essentially turns on whether or not the defendant has taken property subject to an equity in favour of the plaintiff. The jurisprudence has long held that, in order to take subject to an equity, a person need not have actual knowledge of the equity; notice will suffice. In my view, the same standard applies to cases of knowing receipt.
In my view, the test as put forward in both Carl Zeiss Stiftung and Montagu’s Settlement is inappropriate to cases of knowing receipt. The basis of liability in both of these cases is a want of probity (see the judgment of Edmund Davies L.J. in Carl Zeiss Stiftung, at p. 301; see also Montagu’s Settlement, at p. 285). As such, these cases appear to conflate knowing receipt with knowing assistance.
The judgment of Sachs L.J. in Carl Zeiss Stiftung provides further evidence of this collapsing of the two distinct heads of liability where he describes the defendant in knowing receipt cases as a “party to a fraud” (at pp. 298-99). In my view, such a description is inaccurate. Unlike knowing assistance, knowing receipt does not require the plaintiff to show that the breach of trust was fraudulent. And unlike knowing assistance, the defendant in knowing receipt is in no way implicated in any wrongdoing perpetrated by the rogue trustee.
Rather, the cause of action in knowing receipt arises simply because the defendant has improperly received property which belongs to the plaintiff. The plaintiff's claim amounts to nothing more than, “You unjustly have my property. Give it back.” Unlike knowing assistance, there is no finding of fault, no legal wrong done by the defendant and no claim for damages. It is, at base, simply a question of who has a better claim to the disputed property.
In Air Canada, this Court appropriately applied the reasoning developed in Carl Zeiss Stiftung and Montagu’s Settlement in support of the actual knowledge requirement for liability in knowing assistance cases. I should add that, as discussed above, Carl Zeiss Stiftung and Montagu’s Settlement required actual knowledge on the part of the stranger to justify liability in circumstances of knowing receipt. In my view, actual knowledge is inappropriate as a test for liability in knowing receipt cases.
Given the differences between the two causes of action, I can see no good reason why the standard of knowledge which will give rise to liability ought necessarily to be the same. As Millett J. said in Agip (Africa) Ltd. (at p. 292):
The basis of liability in the two types of cases is quite different; there is no reason why the degree of knowledge required should be the same, and good reason why it should not.
Harpum discusses the policy considerations which support the application of a stricter standard on strangers in receipt of trust property than that applied to strangers who assist in a breach of trust (at pp. 126-27):
In such a case [of knowing receipt], the conflict between the beneficiary and the stranger is at its most acute, because the court has in effect to determine which of them has the better title to the trust property. The winner takes all and the loser is left with a claim against the trustee that is likely to be worthless. Because the beneficiary stands to lose outright his beneficial interest, equity is at her most demanding, and insists upon compliance with her most exacting standard.
In all other cases, a stranger should be liable only if he had actual knowledge of, or wilfully closed his eyes to, the terms of the trust, or as the case may be, to the dishonest and fraudulent design on the part of the trustee. No question as to the title of the trust property is in issue. . . . In cases of knowing inducement and assistance, the stranger may never have received any part of the trust property. . . . In cases of knowing assistance, the emphasis on participation by the stranger in the fraud of the trustee necessarily implies that the stranger will be liable only is he acts in bad faith. [Emphasis added.]
Therefore, to conclude my discussion of the applicable legal principles, in order to recover the disputed property, the plaintiff must prove the following:
(1) That the property was subject to a trust in favour of the plaintiff;
(2) That the property, which the defendant received, was taken from the plaintiff in breach of trust; and
(3) That the defendant did not take the property as a bona fide purchaser for value without notice. The defendant will be taken to have notice if the circumstances were such as to put a reasonable person on inquiry, and the defendant made none, or if the defendant was put off by an answer which would not have satisfied a reasonable person.
Before turning to consider whether the guarantee in question was subject to a trust in favour of the plaintiff, I must first address the respondent's argument that the principles of knowing receipt are inapplicable to the present case, because the Bank never received any trust assets. Specifically, the Bank contended that the guarantee itself is not property and that, accordingly, in receiving the guarantee, the Bank did not acquire any property which could be the subject of a knowing receipt claim. I do not agree with this argument. There is some commentary which suggests that a guarantee does not constitute security of a proprietary nature since it is contract security and unlike a mortgage or pledge (K. P. McGuinness, The Law of Guarantee (2nd ed. 1996), at p. 5). However, on the facts of this case, when the Bank obtained the guarantee from Primary, it also acquired, as found by Haley J., a $1.2 million collateral mortgage on real estate belonging to Primary in support of the guarantee. As such, the Bank received both a contractual undertaking to assume the obligations of Trojan in the event of its default, and security of a proprietary nature in support of that undertaking. The mortgage, as security for the guarantee, conferred on the Bank a proprietary interest in the trust property. The guarantee provided by Primary, supported by a collateral mortgage over property owned by Primary, in my view, constitutes property which can be made the subject of a knowing receipt claim. Even if one takes the position that the guarantee does not constitute trust property, the giving of the guarantee confers a valuable benefit on the Bank and correspondingly encumbers the estate and detracts from its value. The benefit conferred on the Bank and the resulting loss in value suffered by the estate are sufficient, in my view, to bring the guarantee within the knowing receipt category of liability.
Furthermore, in the present action, the Bank has attempted to enforce the guarantee against Primary. If the guarantee is enforced, then the Bank will clearly receive property. For these reasons and on policy grounds, in my opinion, the disputed guarantee can be the subject of a claim in knowing receipt.
I will now turn to the question of whether the guarantee was subject to a trust in favour of Gold. The trial judge found that a trust existed under the testator's will. Gold was a beneficiary of that trust. The respondent disputes neither of these holdings. As discussed above, the guarantee deals with rights in the trust corpus. Therefore, in my view, it constitutes property of the trust. Thus, the first element of knowing receipt has been made out: the disputed property was subject to a trust in favour of the appellant.
The second part of the knowing receipt test requires the plaintiff to show that the defendant received property, in its own right, which was taken from him without authority. In this case, it is clear that the Bank took possession of the guarantee in its own right and not as an agent for a third party. With regard to whether or not the guarantee was given to the Bank in breach of trust, it is necessary to examine the circumstances surrounding the agreement. As noted above, Gold was a beneficiary of his grandfather’s estate. The assets of the estate included the shares of Primary. Rosenberg was an executor of the estate. The trial judge found that Rosenberg owed a fiduciary duty to Gold under the Power of Attorney and also to him as a 50 percent beneficiary of the estate. The trial judge further found that, in giving the guarantee to the Bank, Rosenberg acted in breach of trust. In my view, deference is due to all of these findings of fact. Furthermore, I fully agree with Haley J.'s conclusions.
Rosenberg placed an encumbrance on the assets of the estate in order to secure a loan to himself. The transactions surrounding the guarantee benefited only Rosenberg and the Bank; they did not offer any benefit to the appellant. Indeed, the guarantee represented a subtraction from the value of the estate. This kind of self-dealing is a clear breach of trust. Accordingly, in my opinion, the respondent Bank received property which was taken from Gold in breach of trust.
Lastly, the appellant must show that there is no juristic reason for the respondent’s enrichment. As the parties did not argue whether or not the Bank was a mere volunteer, I will not address that issue, assuming instead that the Bank did give value for the guarantee. Consequently, in the facts of this case, the appellant must demonstrate that the Bank did not acquire the property as a bona fide purchaser for value without notice. The question then becomes whether the circumstances were such as to place the Bank on inquiry.
In my view, the circumstances were sufficiently suspicious to give rise to an obligation on the part of the Bank to make reasonable inquiries to ensure that Rosenberg was not acting in breach of trust. In Barclays Bank plc v. O’Brien, [1993] 4 All E.R. 417, the House of Lords considered the circumstances in which a bank will have a duty to inquire when receiving a guarantee from the debtor’s spouse. The House of Lords held that, where a woman enters into a manifestly disadvantageous transaction, and where there is a substantial risk that the husband has committed some equitable or legal wrong (i.e., undue influence or misrepresentation) in order to secure the woman’s consent to the guarantee, the Bank is placed on its inquiry. It then must take reasonable steps to ensure that the wife’s agreement to stand as surety has been properly obtained.
I note that Barclays Bank dealt specifically with circumstances where the debtor and the guarantor were husband and wife and cohabitees, as described by the court. However, the House of Lords indicated that a duty to inquire may also arise in the context of parent-child guarantees (at p. 431). The following comment by Lord Browne-Wilkinson suggests that a bank may have a duty to inquire in a broader range of circumstances (at p. 431):
in a case where the creditor is aware that the surety reposes trust and confidence in the principal debtor in relation to his financial affairs, the creditor is put on inquiry in just the same way as it is in relation to husband and wife.
Further, a mortgage obtained from an employee to secure the debt of the employer and the business of the employer was held to fall within the rule as developed in Barclays Bank even though the parties were not cohabitees or family members (Credit Lyonnais Bank Nederland NV v. Burch, [1997] 1 All E.R. 144, at p. 147).
In light of the foregoing, in my view, the present facts were sufficiently suspicious to put the Bank on inquiry. In reaching this conclusion, I rely particularly on the following facts:
(1) Kenneth Slack, the Toronto-Dominion account manager knew all of the relevant details of Abraham Rosenberg's will and of the estate, in which Rosenberg and Gold held equal interests.
(2) Slack knew that Gold left the management of the estate entirely to Rosenberg, and that Gold had in fact executed a Power of Attorney in favour of Rosenberg.
(3) As the account manager, Slack knew all of the details of the banking arrangements made by Rosenberg in July of 1989.
(4) As mentioned above, Slack knew that Gold had not yet signed the Directors’ resolution authorizing the guarantee at the time that the loans were advanced to Trojan.
(5) Slack must have known that these dealings were to the advantage of neither the estate nor Gold. As the trial judge said (at para. 37):
There can be no doubt that a banker with the experience of Mr. Slack . . . would have been very much aware that the financial position of the Estate Group had been altered to its disadvantage.... He knew that Gold's 50% interest in the estate was now at greater risk than it had been in November 1988 and he also knew that Rosenberg, acting under a Power of Attorney, was in a much better credit position than he had been before the consolidation.
These facts are sufficiently unusual to place the Bank on inquiry. It failed to take reasonable steps to ensure that the guarantee did not constitute a breach of trust. Accordingly, it is fixed with notice of the breach of trust.
The question which then arises is whether the Bank made sufficient inquiry or, alternatively, whether it received assurances which would have satisfied a reasonable person. In this respect, the respondent relies on the opinion letter sent by Sills, Madorin which stated that the guarantee complied with all legal requirements. However, in my opinion, this letter does not satisfy the Bank’s obligation to make reasonable inquiry. My conclusion on this point is supported by the holding in Bertolo v. Bank of Montreal (1986), 57 O.R. (2d) 577 (C.A.). Mrs. Bertolo provided a promissory note and a mortgage on her house to the bank to secure her son’s indebtedness. She was not fluent in English and she had little education. The bank required that she obtain independent legal advice and undertook to ensure that she received such advice. A partner to the lawyer who was acting for both the bank and Mrs. Bertolo’s son advised Mrs. Bertolo; this advice satisfied the bank. The court held that the bank could not enforce the promissory note against Mrs. Bertolo since she did not understand the nature of the transactions nor the extent of her liability. The court stated that there was no fiduciary relationship as between the debtor and the bank, but that the bank (at p. 587):
was aware, or ought to have been aware, that this woman had not had the benefit of independent legal advice with respect to a transaction which, from a business viewpoint, was manifestly disadvantageous to her.
In the present case, the Bank knew that the law firm was acting on behalf of all parties: it represented Rosenberg, Trojan, the estate and the estate companies. Therefore, it knew that the firm could not have given Gold independent legal advice with regard to signing the directors’ resolution which authorized the guarantee. In my view, the opinion letter does not satisfy the Bank's “duty” to inquire. Accordingly, the Bank is fixed with notice of the breach of trust and, therefore, takes the guarantee subject to Gold's equity. For these reasons, in my view, the Bank cannot enforce the guarantee against Primary.
Before this Court, counsel for the respondent argued that imposing liability on the Bank in the present case would put an unreasonable burden on banks to ensure that every transaction which they facilitated did not involve a breach of trust. I do not agree. This responsibility will only arise in those circumstances where the Bank beneficially receives property. In this respect, the Bank will be in the same position as that occupied by any person who receives property, a state of the law which seems eminently fair, in my view.
For all of these reasons, I would allow the appeal, set aside the judgment of the Court of Appeal, and substitute therefor an order to the effect that the $1.2 million guarantee shall be unenforceable as against Primary. I would make the foregoing order with costs here and in the Court of Appeal.
//Sopinka J.//
The judgment of Sopinka, McLachlin and Major JJ. was delivered by
Sopinka J.
Sopinka J. -- I have had the benefit of reading the reasons of my colleague, Iacobucci J., in this appeal as well as those of La Forest J. in Citadel General Assurance Co. v. Lloyds Bank Canada, [1997] 3 S.C.R. 805, released concurrently, with which I concur subject to my comments in these reasons. While I agree with Iacobucci J.’s approach to liability for knowing assistance, I have serious reservations with his conclusion that this is a knowing receipt case but, assuming it is, I cannot agree that the bank is liable for breach of its duty to inquire.
The case was presented in both the trial court and the court of appeal as a knowing assistance case. It was dealt with in those courts on that basis. The theory that it was a knowing receipt was raised in this Court. While, no doubt, we have the jurisdiction to do so, this Court is reluctant to allow a party to depart from the theory of liability as presented in the courts from which an appeal is taken. Assuming the issue should be entertained, I have difficulty believing that when a bank receives a guarantee supported by a collateral mortgage on trust property, it has received the trust property to its own use and benefit. The Oxford English Dictionary (2nd ed. 1989) provides the following principal definition of the verb to receive:
To take in one’s hand, or into one’s possession (something held out or offered by another); to take delivery of (a thing) from another, either for oneself or for a third party.
In the context of knowing receipt cases, I would say that to receive trust property means, at a minimum, to take the trust property into one’s possession. Possession here does not imply any form of ownership. It implies only physical control.
Does the bank receive the trust property into its possession simply because it holds a guarantee supported by a collateral mortgage on that property? I believe it does not. A guarantee is a contract whose performance is contingent on the default of the principal debtor: K. P. McGuinness, The Law of Guarantee (2nd ed. 1996), at pp. 318-19. See also Halsbury’s Laws of England (4th ed. 1993), vol. 20, at pp. 56-65, 115-16, 123-24. The guarantor is liable to make good the debts of the principal debtor. If the guarantor supports the guarantee with a mortgage on real property, the creditor only enjoys, at best, a contingent interest in that property. The guarantee supported by the mortgage is no more than a contractual undertaking by the guarantor that, if the principal debtor defaults and the guarantor cannot make good the debt from his or her other assets, the creditor will receive the trust property. The mortgage is security for the performance of the contractual provision embodied in the guarantee.
The Bank Acted Reasonably
Even if this is properly viewed as a knowing receipt case, in my view the bank, knowing what it knew, acted reasonably in the circumstances. It therefore cannot be found liable and the appeal should be dismissed.
In order to establish that the respondent was in “knowing” receipt, the appellant must establish one of the following:
(i) actual knowledge; (ii) wilfully shutting one’s eyes to the obvious; (iii) wilfully and recklessly failing to make such inquiries as an honest and reasonable man would make; (iv) knowledge of circumstances which would indicate the facts to an honest and reasonable man; (v) knowledge of circumstances which would put an honest and reasonable man on inquiry.
Baden v. Société Générale pour Favoriser le Développement du Commerce et de l’Industrie en France SA, [1992] 4 All E.R. 161 (Ch.), at p. 235.
My colleague finds that the bank had knowledge on the basis of item (v), that is, knowledge of circumstances which would put an honest and reasonable person upon inquiry. What then was the knowledge that the bank had? The bank had knowledge of the following circumstances:
(1) Gold was content to leave the management of the estate to Rosenberg and had signed a general power of attorney in favour of Rosenberg to this effect.
(2) A guarantee along with a collateral mortgage was duly executed by Primary.
(3) The solicitor for the estate (selected by Gold and Rosenberg) and for the guarantor acted on the guarantee and formally attested to its validity by letter.
(4) The accountants for the estate and the guarantor gave financial advice with respect to the guarantee.
(5) Gold had signed a resolution as one of the directors of Primary in the office of Mr. Sills, a senior partner of the law firm representing the estate and the guarantor.
(6) Rosenberg’s share of the estate was worth far in excess of the amount of the guarantee.
In my opinion, an honest person with knowledge of these facts would not have made further inquiries. Indeed, it is difficult to determine to whom such inquiries would be directed. As stated by Laskin J.A. (25 O.R. (3d) 601, at p. 610):
Gold argues that the bank had a duty to go behind the resolution and the opinion letter to inquire about the fairness of the transaction. But inquire of whom? Presumably the lawyers and the accountants who acted on the transaction would have affirmed its fairness if asked. And the bank could reasonably assume that Gold would give the guarantee to help his uncle.
I would add that if Gold had been asked about the guarantee, he would hardly have questioned it in view of the fact that he signed the resolution. I agree with Laskin J.A. (at pp. 607-8) that at the time:
. . . Gold knew what a guarantee was, he knew the reason for this guarantee and he knew the possible consequences of authorizing it. He was not misled about the purpose, the effect or the risk of giving his approval.
It was only later when circumstances changed that he questioned the guarantee. Inquiries to Gold would not, therefore, have alerted the bank to any facts which would have raised a doubt about the transaction.
As I read my colleague’s reasons, he does not answer the question what would an inquiry have revealed but suggests that the bank should have insisted that Gold receive independent legal advice.
In certain circumstances, a third party in the position of the bank will not have discharged its duty to inquire unless the guarantor has been advised to obtain independent legal advice. In certain cases, the law imposes on a creditor a duty to inquire when the transaction is clearly detrimental to the person offering security and the relationship between that person and the principal debtor is particularly close. In such circumstances, the law presumes undue influence on the part of the principal debtor. The clearest type of relationship giving rise to this presumption is that of husband and wife. Iacobucci J. cites Barclays Bank plc v. O’Brien, [1993] 4 All E.R. 417, in which the House of Lords extended this presumption to include cohabitees. Lord Browne-Wilkinson held that when a creditor is approached by cohabitees, one the principal debtor and the other the surety, and the proposed transaction is clearly to the disadvantage of the surety, it will be under a duty to inquire. A creditor can discharge this duty by explaining to the surety in a meeting not attended by the principal debtor the amount of her potential liability and the risks involved and advising her to take independent advice: Barclays, supra, at pp. 431-32.
At one point in his reasons, Lord Browne-Wilkinson appeared to extend the duty beyond cohabiting couples when he characterized the kinds of relationships that will trigger the duty to inquire as follows, at p. 431:
[I]n a case where the creditor is aware that the surety reposes trust and confidence in the principal debtor in relation to his financial affairs, the creditor is put on inquiry in just the same way as it is in relation to husband and wife.
When setting out the strict requirements of a separate meeting with the surety, however, Lord Browne-Wilkinson spoke of “the emotional pressure of cohabitation” (p. 431). Elsewhere, he spoke of how “the sexual and emotional ties between the [married] parties provide a ready weapon for undue influence” (p. 424). When a bank is presented with such a relationship, it should recognize the risk of undue influence (assuming that the transaction is on its face detrimental to the party offering security). But by the same logic, a relationship that is more distant will raise less suspicion of undue influence, even if the transaction is apparently unfavourable to the guarantor. Consequently, less may be required to satisfy an honest and reasonable person that the surety or guarantor is aware of the legal implications of the proposed transaction.
At the time that Gold signed the resolution he had three years of university education in which he had taken courses in business, economics and accounting. He was employed by the University of Western Ontario Student Council full-time as Manager of Entertainment and Productions. The purpose of the guarantee was explained to him by Rosenberg. The following sworn statement by Rosenberg was read in as part of Gold’s case:
In July of 1989 I advised Jeffrey that I was required by the Toronto-Dominion Bank to provide further security for the indebtedness of Trojan and that as I was unable to, quote, access, end of quote, my half interest in the estate at that time I requested that Primary guarantee Trojan’s indebtedness in the amount of $1,200,000 which was to be secured by a collateral mortgage on 186 Columbia Street, a property owned by Primary. The reasons for the guarantee were fully disclosed to Jeffrey before he signed a resolution as director of Primary authorizing same.
Since my fifty percent share of the estate was worth far in excess of $1,200,000, there was virtually no chance that the guarantee could possibly put Jeffrey’s half interest in the estate at risk.
The statement that Rosenberg’s share in the estate was worth far more than the guarantee was accurate at the time.
The guarantor, it must be remembered, was not Gold but Primary, a corporation whose shares formed part of the estate.
In these circumstances, I cannot accept that further explanation or legal advice was required and if it had been offered it would not have made any difference. I am confident that Gold would have found a “chat” with Mr. Slack quite superfluous. As for more legal advice, I am sure that if Gold thought independent legal advice was needed, he would have obtained it. He was aware that the solicitors for the estate also acted for his uncle and it was scarcely necessary for the bank to advise him that he could consult another lawyer if he wished.
A bank is only required to act reasonably in the circumstances. Corporate guarantees in situations in which the director of the corporation may be a beneficiary of a trust in relation to the shares of the corporation are common transactions. Is the bank obligated to advise each director, whose consent was necessary to obtain the guarantee, to obtain independent legal advice? I am of the opinion that, in the circumstances, advising Gold to obtain independent legal advice may be a counsel of perfection but goes beyond what is expected of an honest and reasonable banker. To quote Gibson J. in Baden, supra, at p. 268, this would impose “an impractically extensive duty of inquiry” on a bank which is otherwise acting reasonably.
My colleague refers to the decision of the Ontario Court of Appeal in Bertolo v. Bank of Montreal (1986), 57 O.R. (2d) 577. In that case, Mrs. Bertolo provided the bank with a promissory note and a mortgage on her house to secure a loan taken out by her son. The bank’s policy required that she receive independent legal advice, but this, for various reasons, did not occur. When the son’s business failed, the bank sought to enforce the guarantee. At trial and on appeal, Mrs. Bertolo succeeded in having the promissory note and mortgage declared void because she did not fully understand the consequences of her actions. The bank could not enforce the guarantee because it had failed to ensure that she received independent legal advice.
In my opinion, Bertolo is of little assistance. One can explain the necessity for independent legal advice in that case by noting that Mrs. Bertolo was incapable without such advice of understanding any aspect of the transaction. I need only quote from the Court of Appeal’s description of her (at p. 579):
She has no business experience and little formal education. She is not fluent in English and . . . is unable to read and discern such documents as promissory notes, collateral mortgages and financial statements.
Whether or not someone requires independent legal advice will depend on two principal concerns: whether they understand what is proposed to them and whether they are free to decide according to their own will. The first is a function of information and intellect, while the second will depend, among other things, on whether there is undue influence. Leaving aside entirely the possibility of undue influence by her son, something which was never alleged, it is clear that, without independent legal advice, Mrs. Bertolo could not possibly have understood what she was agreeing to. The Ontario Court of Appeal held that Mrs. Bertolo ought to have had independent legal advice and that she did not get it, first, because the lawyer she met worked for the same firm that represented her son and the bank, and, second, because the nature and consequences of the transaction were not explained to her.
There is no comparison between Gold and Mrs. Bertolo and the circumstances of the two cases. Indeed, the contrast serves to illustrate the type of case in which independent legal advice is not a prerequisite.
I would dismiss the appeal with costs.
//Gonthier J.//
The following are the reasons delivered by
While I agree with Iacobucci J. that this case is one of knowing receipt of trust property, I share the view of Sopinka J. that the Bank, knowing what it knew, acted reasonably in the circumstances. I concur with his reasons in this respect and his disposition of the appeal.
Cases cited
- Citing MacDonald v. Hauer 1 W.W.R. 51 ¶ 13
- 25 O.R. (3d) 601 25 O.R. (3d) 601 ¶ 18
- Air Canada v. M & L Travel Ltd. 22416-1993-10-21 para 37 ¶ 19, ¶ 27, ¶ 30, ¶ 31, ¶ 32, ¶ 33, ¶ 40, ¶ 42, ¶ 50, ¶ 63
- Citadel General Assurance Co. v. Lloyds Bank Canada 25189-1997-10-30 ¶ 25, ¶ 69
- Harpum Harpum, supra ¶ 42
- Potter (Carl B.) Ltd. v. Mercantile Bank of Canada item-5727-1980-07-18 ¶ 44
- Cartwright v. Lyster 2 D.L.R. 166 ¶ 44
- Bertolo v. Bank of Montreal 57 O.R. (2d) 577 ¶ 65, ¶ 84, ¶ 85, ¶ 86
- Barclays Barclays, supra, at pp. 431-32 pp. 431-32 ¶ 78
- Baden Baden, supra, at p. 268 p. 268 ¶ 83