Citadel General Assurance Co. v. Lloyds Bank Canada, [1997] 3 S.C.R. 805
The Citadel General Assurance Company and
the Citadel Life Assurance Company Appellants
v.
Lloyds Bank Canada and Hongkong Bank of Canada Respondents
Indexed as: Citadel General Assurance Co. v. Lloyds Bank Canada
File No.: 25189.
1997: May 20; 1997: October 30.
Present: La Forest, Sopinka, Gonthier, Cory, McLachlin, Iacobucci and Major JJ.
on appeal from the court of appeal for alberta
Trusts and trustees ‑‑ Breach of trust ‑‑ Liability of strangers to trust ‑‑ Knowing assistance ‑‑ Knowing receipt ‑‑ Insurance agent depositing premiums collected on insurer’s behalf into bank account ‑‑ Bank transferring funds to account of insurance agent’s parent company to reduce overdraft ‑‑ Whether bank liable for breach of trust on basis of knowing assistance or knowing receipt.
D sold insurance to auto dealers. After collecting the premiums, D paid commissions and settled any current claims under the policies. The balance of the premiums was remitted on a monthly basis to the appellant insurance companies, the underwriters of the policies. In December 1986, D and its parent company started banking with the respondents (the “bank”). D used one bank account for all its transactions. Through its senior officers, the bank was aware that insurance premiums were being deposited into that account. In May 1987, a “trip report” by one of the appellants’ employees indicated that D was reluctant to establish a trust account for the premiums but would do so if necessary. From June 1, D no longer settled claims under the insurance policies, with the result that the monthly premiums payable to the appellants increased significantly. In June the bank received instructions from the parent company’s signing officers, who were identical to D’s signing officers, to transfer all funds in D’s account to the parent company’s account at the end of each business day. In July and August, the transfer of funds between the accounts resulted in an overall reduction in the parent company’s overdraft. In late August D advised the appellants that the July and August premiums could not be remitted. It agreed to pay these outstanding receipts by way of promissory note. After D and its parent company ceased carrying on business, the appellants brought an action against the bank for the outstanding insurance premiums. They were successful at trial and judgment was entered against the bank. The Court of Appeal allowed the bank’s appeal and dismissed the appellants’ claim.
Held: The appeal should be allowed.
Per La Forest, Gonthier, Cory, McLachlin, Iacobucci and Major JJ.: The relationship between the appellant insurers and D was clearly one of trust. Under s. 124(1) of the Alberta Insurance Act, an agent who receives any money as a premium for an insurance contract from the insured is deemed to hold the premium in trust for the insurer. The promissory note was merely confirmation of the amount owed by D to the appellants and did not amount to a revocation of the trust. As well, the arrangement between them meets the three characteristics of a trust, namely certainty of intent, certainty of subject‑matter, and certainty of object. The fact that the trust funds in D’s account were commingled with other funds does not undermine the relationship of trust between the parties. Also, D’s actions in failing to remit to the appellants the insurance premiums collected on their behalf in July and August 1987 were clearly in breach of trust. Moreover, the appellants did not acquiesce in the breach of trust by asking for and receiving the promissory note from D.
There are three ways in which a stranger to a trust can be held liable as a constructive trustee for breach of trust: as a trustee de son tort; for “knowing assistance”; and for “knowing receipt”. The first type of liability is inapplicable to the present case since the bank never assumed the office or function of trustee. A stranger to a trust can be liable for breach of trust by knowingly assisting in a fraudulent and dishonest design on the part of the trustees. Assuming the present case falls under this “knowing assistance” category, it is clear that only actual knowledge, recklessness, or wilful blindness will render the bank liable for participating in the breach of trust. Since the bank had only constructive knowledge, it cannot be liable under the “knowing assistance” category of constructive trusteeship.
Liability on the basis of “knowing receipt” requires that strangers to the trust receive or apply trust property for their own use and benefit. By applying the deposit of insurance premiums as a set‑off against the parent company’s overdraft, the bank received a benefit and thus received the trust funds for its own use and benefit. The bank cannot avoid the “property” issue by characterizing the deposit of trust monies in D’s account as a debt obligation. A debt obligation is a chose in action and, therefore, property over which one can impose a trust. The receipt requirement in “knowing receipt” cases is best characterized in restitutionary terms. In this case the bank has been enriched at the appellants’ expense and thus, in restitutionary terms, there can be no doubt that the bank received trust property for its own use and benefit.
The second requirement for establishing liability on the basis of “knowing receipt” relates to the degree of knowledge required of the bank in relation to the breach of trust. While constructive knowledge is excluded as the basis for liability in “knowing assistance” cases, in “knowing receipt” cases, which are concerned with the receipt of trust property for one’s own benefit, there should be a lower threshold of knowledge required of the stranger to the trust. More is expected of the recipient, who, unlike the accessory, is necessarily enriched at the plaintiff’s expense. Because the recipient is held to this higher standard, constructive knowledge (that is, knowledge of facts sufficient to put a reasonable person on notice or inquiry) will suffice as the basis for restitutionary liability. This lower threshold of knowledge is sufficient to establish the “unjust” or “unjustified” nature of the recipient’s enrichment, thereby entitling the plaintiff to a restitutionary remedy. More specifically, relief will be granted where a stranger to the trust, having received trust property for his or her own benefit and having knowledge of facts which would put a reasonable person on inquiry, actually fails to inquire as to the possible misapplication of trust property.
On the issue of knowledge, it is clear from the trial judge’s findings that the bank was aware of the nature of the funds being deposited into, and transferred out of, D’s account. The bank knew that D’s sole source of revenue was the sale of insurance policies and that premiums collected by D were payable to the appellants. In light of the bank’s knowledge of the nature of the funds, the daily emptying of D’s account was in the trial judge’s view “very suspicious”. In these circumstances, a reasonable person would have been put on inquiry as to the possible misapplication of the trust funds. The bank should have inquired whether the use of the premiums to reduce the account overdrafts constituted a breach of trust. By failing to make the appropriate inquiries, the bank had constructive knowledge of D’s breach of trust. The bank’s enrichment was thus clearly unjust, rendering it liable to the appellants as a constructive trustee.
Per Sopinka J.: Subject to what was said in Gold, issued concurrently, La Forest J.’s reasons were agreed with.
Cases Cited
By La Forest J.
Referred to: Air Canada v. M & L Travel Ltd., [1993] 3 S.C.R. 787; Gold v. Rosenberg, [1997] 3 S.C.R. 767; Canadian Pacific Air Lines, Ltd. v. Canadian Imperial Bank of Commerce (1987), 61 O.R. (2d) 233; R. v. Lowden (1981), 27 A.R. 91; Bank of N.S. v. Soc. Gen. (Can.), [1988] 4 W.W.R. 232; Fletcher v. Collis, [1905] 2 Ch. 24; Selangor United Rubber Estates, Ltd. v. Cradock (No. 3), [1968] 2 All E.R. 1073; Agip (Africa) Ltd. v. Jackson, [1990] 1 Ch. 265, aff’d [1992] 4 All E.R. 451; Foley v. Hill (1848), [1843‑60] All E.R. Rep. 16; Fonthill Lumber Ltd. v. Bank of Montreal (1959), 19 D.L.R. (2d) 618; Lipkin Gorman v. Karpnale Ltd., [1991] 3 W.L.R. 10; Lac Minerals Ltd. v. International Corona Resources Ltd., [1989] 2 S.C.R. 574; Air Canada v. British Columbia, [1989] 1 S.C.R. 1161; In re Montagu’s Settlement Trusts, [1987] 1 Ch. 264; Polly Peck International plc v. Nadir (No. 2), [1992] 4 All E.R. 769; C.I.B.C. v. Valley Credit Union Ltd., [1990] 1 W.W.R. 736; Bullock v. Key Property Management Inc. (1997), 33 O.R. (3d) 1; Groves‑Raffin Construction Ltd. v. Bank of Nova Scotia (1975), 64 D.L.R. (3d) 78; Carl B. Potter Ltd. v. Mercantile Bank of Canada, [1980] 2 S.C.R. 343; Arthur Andersen Inc. v. Toronto‑Dominion Bank (1994), 17 O.R. (3d) 363, leave to appeal refused, [1994] 3 S.C.R. v; Glenko Enterprises Ltd. v. Ernie Keller Contractors Ltd. (1996), 134 D.L.R. (4th) 161; Cowan de Groot Properties Ltd. v. Eagle Trust plc, [1992] 4 All E.R. 700; El Ajou v. Dollar Land Holdings plc, [1993] 3 All E.R. 717; Royal Brunei Airlines Sdn. Bhd. v. Tan, [1995] 3 W.L.R. 64.
By Sopinka J.
Referred to: Gold v. Rosenberg, [1997] 3 S.C.R. 767.
Statutes and Regulations Cited
Bank Act, R.S.C., 1985, c. B‑1 [now repealed], s. 206(1).
Insurance Act, R.S.A. 1980, c. I‑5, s. 124(1).
Authors Cited
Birks, Peter. “Misdirected Funds: Restitution from the Recipient”, [1989] L.M.C.L.Q. 296.
Birks, Peter. “Overview: Tracing, Claiming and Defences”. In Peter Birks, ed., Laundering and Tracing. Oxford: Clarendon Press, 1995, 289.
Gardner, Simon. “Knowing Assistance and Knowing Receipt: Taking Stock” (1996), 112 L.Q.R. 56.
Halsbury’s Laws of England, vol. 48, 4th ed. London: Butterworths, 1995 (reissue).
Harpum, Charles. “The Stranger as Constructive Trustee” (1986), 102 L.Q.R. 114.
Millett, Sir Peter. “Tracing the Proceeds of Fraud” (1991), 107 L.Q.R. 71.
Pettit, Philip H. Equity and the Law of Trusts, 7th ed. London: Butterworths, 1993.
Underhill and Hayton, Law Relating to Trusts and Trustees, 14th ed. By David J. Hayton. London: Butterworths, 1987.
APPEAL from a judgment of the Alberta Court of Appeal (1996), 181 A.R. 76, 116 W.A.C. 76, 37 Alta. L.R. (3d) 293, [1996] 5 W.W.R. 9, 33 C.C.L.I. (2d) 241, [1996] A.J. No. 59 (QL), reversing a judgment of the Court of Queen’s Bench, [1993] A.J. No. 680 (QL), imposing a constructive trust on the defendant banks. Appeal allowed.
Eric F. Macklin, Q.C., and W. Scott Schlosser, for the appellants.
Donald R. Cranston and Stacy M. Neufeld, for the respondents.
//La Forest J.//
The judgment of La Forest, Gonthier, Cory, McLachlin, Iacobucci and Major JJ. was delivered by
1 La Forest J. -- This appeal concerns the liability of a bank for its customer’s breach of trust. The appellants, as beneficiaries of the trust, seek recovery for unpaid insurance premiums collected by the trustee and deposited with the respondent banks. The principal question in this appeal is this: Are the respondent banks liable as constructive trustees for the breach of trust committed by one of their clients? This question deals with the liability of strangers who participate in a breach of trust and, in particular, the degree of knowledge required for the imposition of liability as a constructive trustee.
I. Factual Background
2 The Citadel General Assurance Company and the Citadel Life Assurance Company (“Citadel”) are insurance companies which carried on business in Alberta. Beginning in 1979, Citadel’s business operations involved another Alberta corporation, Drive On Guaranteed Vehicle Payment Plan (1982) Limited (“Drive On”). As a wholly owned subsidiary of International Warranty Company Limited (“International Warranty”), Drive On sold consumer life, casualty, and unemployment insurance to auto dealers. The insurance premiums were collected by auto dealers at the time vehicles were sold and then remitted to Drive On. After collecting the premiums, Drive On paid commissions and settled any current claims under the policies. The balance of the premiums was remitted on a monthly basis to Citadel, the underwriter of the insurance policies. In early 1987, the premiums received during a calendar month were normally forwarded to Citadel at the end of the following month. This arrangement continued satisfactorily until August 1987 when Drive On defaulted on its payments to Citadel. As well, from 1979 until late August 1987, there was no written agreement between Citadel and Drive On.
3 In December 1986, the International Warranty Group of Companies, including Drive On and International Warranty, started banking with the now-amalgamated Lloyds Bank Canada and Hongkong Bank of Canada (hereinafter collectively called “the Bank”). During 1987, the only banker of Drive On was the Bank. Drive On used one bank account for all its transactions. The only deposits to that account were either insurance premiums collected from auto dealers or transfers of funds from International Warranty. Through its senior officers, the Bank was aware that insurance premiums were being deposited into Drive On’s account. During the period after April 1, 1987, Drive On’s account was usually in an overdraft position. On April 8, the Bank received instructions from International Warranty’s signing officers (who were identical to Drive On’s signing officers) to transfer funds between the International Warranty and Drive On accounts to cover overdrafts on either account.
4 Also in April 1987, the Presidents of Drive On and Citadel met to discuss their business relationship. As a result of the meeting, Citadel and Drive On agreed that a written agreement would be prepared to formalize their business relationship. It was also agreed that Drive On would no longer settle claims under the insurance policies. From June 1, 1987, Citadel assumed the adjudication and payment of all claims. As a result, the monthly premiums payable to Citadel were increased significantly. Also as a result of the presidents’ meeting, Citadel ordered a detailed examination of Drive On’s current procedures. A “trip report” was delivered by one of Citadel’s employees on May 14. The report found that Drive On was depositing insurance premiums in a general bank account which was not set up as a trust account. The report also indicated that Drive On was somewhat reluctant to establish a trust account but would do so if necessary.
5 On June 5, 1987, the Bank received instructions from International Warranty’s signing officers to transfer all funds in the Drive On account to the International Warranty account at the end of each business day. In July and August, the transfer of funds between the International Warranty and Drive On accounts resulted in an overall reduction in International Warranty’s overdraft.
6 On August 7, 1987, Drive On forwarded the June premiums to Citadel. In late August, Citadel first learned of Drive On’s financial difficulties. The President of Drive On advised Citadel that the July and August premiums could not be remitted. A new arrangement, effective September 1, was set in place whereby all premium monies would be forwarded directly to Citadel. With regard to the outstanding July and August receipts, Drive On agreed to pay Citadel by way of promissory note dated September 21, 1987. The note provided for monthly payments of $100,000. Drive On made a number of payments until it, and the other members of the International Warranty Group of Companies, ceased carrying on business in December. Citadel sued Drive On and the guarantor on the promissory note but has been unsuccessful in collecting anything. The parties agree that the outstanding amount payable to Citadel is $633,622.84.
7 Citadel brought an action against the Bank for the outstanding insurance premiums. At trial, Citadel was successful and judgment was entered against the Bank for $633,622.84: [1993] A.J. No. 680 (QL). The Court of Appeal allowed the Bank’s appeal and dismissed Citadel’s claim: (1996), 181 A.R. 76, 116 W.A.C. 76, 37 Alta. L.R. (3d) 293, [1996] 5 W.W.R. 9, 33 C.C.L.I. (2d) 241, [1996] A.J. No. 59 (QL).
II. Decisions Below
A. Court of Queen’s Bench of Alberta
8 The trial judge, Marshall J., found that a relationship of trust existed between Citadel and Drive On. This finding was based in part on s. 124(1) of the Insurance Act, R.S.A. 1980, c. I-5, which provides that an agent or broker who negotiates a contract of insurance with an insurer and receives insurance premiums for that contract is deemed to hold the premiums in trust for the insurer. As well, the trial judge found that the arrangement between Citadel and Drive On had the three certainties of a trust, namely, certainty of subject-matter, certainty of intent, and certainty of object. Further, the trial judge held that a breach of trust occurred when Drive On failed to remit the insurance premiums collected on Citadel’s behalf.
9 The trial judge stated that in order for the Bank to be liable as a stranger to the trust, the Bank must have had knowledge that the funds were trust monies or the circumstances must have required it to inquire before dealing with the money. In scrutinizing the evidence, the trial judge found that Drive On’s instructions to empty its account daily were “very suspicious” (para. 26), given the Bank’s knowledge that insurance premiums were being deposited in the account. As well, the trial judge concluded that the Bank, having received a benefit from Drive On’s breach of trust, was under a greater duty to explain its actions. Liability was imposed on the following basis (at para. 33):
The Bank had actual knowledge of the nature of the funds in the Drive On account and had an obligation to inquire about their position in the circumstances. The Bank shut its eyes in circumstances which should have caused it to inquire of its customer at least. It did not do so. The Bank had constructive knowledge and is a constructive trustee within the cases . . . .
The plaintiffs received judgment for $633,622.84.
B. Alberta Court of Appeal (1996), 181 A.R. 76
10 On appeal to the Alberta Court of Appeal, Kerans J.A., speaking for the court, began by noting, at p. 77, that there was “no serious difficulty with the finding of creation of the trust made by the trial judge”. As such, Kerans J.A. was willing to assume that a trust existed between Citadel and Drive On and that a breach of trust occurred.
11 In Kerans J.A.’s view, the real difficulty with the case was the imposition of liability on the Bank. More specifically, he disagreed with the trial judge’s conclusion that constructive knowledge was sufficient to render the Bank liable as a constructive trustee. Relying on this Court’s decision in Air Canada v. M & L Travel Ltd., [1993] 3 S.C.R. 787, Kerans J.A. concluded that only actual knowledge, recklessness, or wilful blindness could render the Bank liable for a breach of trust from which it received a benefit. This test was not met in the present case because, although the Bank had “shut its eyes” in the circumstances, the trial judge refused to find that the Bank was actually aware it was taking money in breach of trust. However, since Air Canada v. M & L Travel Ltd. was decided after the trial judge rendered judgment in the present case, the parties were invited to re-argue the facts before the Court of Appeal. Nonetheless, even after reconsidering the facts, Kerans J.A. concluded for the court at p. 78:
. . . we have come to the conclusion that we cannot say on the balance of probabilities that this bank, when it honoured the direction to pay, was aware that it was moving out money in breach of any trust between the defaulting company and the insurance company. We do not have any difficulty with the trial finding that the bank had some warning that this was the case. But that is not enough. Nor do we think this is an appropriate case in which to rely on wilful blindness.
The Court of Appeal accordingly allowed the Bank’s appeal and dismissed the claim.
III. Issues
12 As mentioned, one main question is raised on this appeal: Under what circumstances can the respondent banks be held liable as constructive trustees for the breach of trust committed by one of their customers?
IV. Analysis
A. The Nature of the Relationship between Citadel and Drive On
13 Before beginning my analysis regarding the liability of the Bank as a constructive trustee, I note that I have read the reasons of Iacobucci J. in Gold v. Rosenberg, [1997] 3 S.C.R. 767, a case also dealing with the liability of a bank for a breach of trust committed by one of its customers. I generally agree with Iacobucci J.’s approach in Gold and, indeed, consider it similar to my own in the present appeal.
14 There can be no doubt that the relationship between Citadel and Drive On was one of trust. In this Court, the parties did not dispute the existence of a trust relationship. Relevant to the arrangement between Citadel and Drive On is s. 124(1) of the Insurance Act, which provides:
124(1) An agent or broker who acts in negotiating, renewing or continuing a contract of insurance with an insurer licensed under this Act, and who receives any money or substitute for money as a premium for such a contract from the insured, shall be deemed to hold the premium in trust for the insurer.
From 1979 to 1987, Drive On, the insurance agent, was in the business of selling insurance policies underwritten by Citadel, the insurer. In negotiating insurance policies on Citadel’s behalf, Drive On collected insurance premiums from auto dealers, paid commissions, and settled current claims under the policies. However, on June 1, 1987, the arrangement was changed and the premiums collected by Drive On were to be forwarded without deductions for policy claims. However, even when Citadel assumed the adjudication of all claims, Drive On still acted as agent or trustee and Citadel remained the principal or beneficiary of the insurance premiums. Moreover, I agree with the trial judge that the repayment arrangements between Citadel and Drive On did not amount to a revocation of the trust. The promissory note dated September 21, 1987, was merely confirmation of the amount owed by Drive On to Citadel. By agreeing to have the promissory note prepared in its favour, Citadel did not revoke its beneficial interest in the insurance premiums.
15 As well, the arrangement between Citadel and Drive On meets the three characteristics of a trust, namely certainty of intent, certainty of subject-matter, and certainty of object; see Air Canada v. M & L Travel Ltd., supra, at pp. 803-4; Canadian Pacific Air Lines, Ltd. v. Canadian Imperial Bank of Commerce (1987), 61 O.R. (2d) 233 (H.C.), at p. 237. The arrangement in the present case was based on the collection of insurance premiums by the insurance agent, Drive On, and the remittance of these premiums, subject to adjustments, to the insurer, Citadel. The intent to create a trust clearly follows from this principal-agent relationship. The object of the trust is the insurer, Citadel. Finally, the insurance premiums constitute the subject-matter of the trust.
16 The fact that the trust funds in Drive On’s account were commingled with other funds does not undermine the relationship of trust between the parties. As Iacobucci J. wrote for the majority of this Court in Air Canada v. M & L Travel Ltd., supra, at p. 804, “[w]hile the presence or absence of a prohibition on the commingling of funds is a factor to be considered in favour of a debt relationship, it is not necessarily determinative”; see also R. v. Lowden (1981), 27 A.R. 91 (C.A.), at pp. 101-2; Bank of N.S. v. Soc. Gen. (Can.), [1988] 4 W.W.R. 232 (Alta. C.A.), at p. 238.
17 The intention of the parties in the present case was to create a trust relationship. That Drive On deposited the funds in a general bank account, as opposed to a special trust account, does not alter this intention. In May 1987, Citadel prepared a report of Drive On’s procedures. This report found that a trust account had not been set up by Drive On. However, the report also noted that Drive On would establish a trust account if required by Citadel. The report indicates, therefore, that the parties had turned their minds to the possibility of setting up a trust account to prohibit the commingling of funds. Even though the trust account was never established, the fact that the parties considered this possibility confirms that the relationship was viewed by Citadel and Drive On as one of trust.
B. The Liability of the Bank as a Stranger to the Trust
1. General Principles
18 Having found that the relationship between Citadel and Drive On was one of trust, it is clear that Drive On’s actions were in breach of trust. Quite simply, Drive On failed to remit to Citadel the insurance premiums collected on Citadel’s behalf in July and August 1987. Moreover, I agree with the trial judge that Citadel did not acquiesce in the breach of trust by asking for and receiving the promissory note from Drive On. By accepting the note, Citadel did not represent that it was acquiescing in the use of the funds by the Bank. Consequently, Citadel is not barred from bringing an action against the bank for breach of trust; see Fletcher v. Collis, [1905] 2 Ch. 24 (C.A.); P. H. Pettit, Equity and the Law of Trusts (7th ed. 1993), at p. 491. The question remains whether the Bank, as a stranger to the trust between Citadel and Drive On, can be liable as a constructive trustee.
19 There are three ways in which a stranger to a trust can be held liable as a constructive trustee for breach of trust. First, a stranger to the trust can be liable as a trustee de son tort. Secondly, a stranger to the trust can be liable for breach of trust by knowingly assisting in a fraudulent and dishonest design on the part of the trustees (“knowing assistance”). Thirdly, liability may be imposed on a stranger to the trust who is in receipt and chargeable with trust property (“knowing receipt”; see Air Canada v. M & L Travel Ltd., supra, at pp. 809-11).
20 To be liable as trustees de son tort, strangers to the trust must commit a breach of trust while acting as trustees. Such persons are not appointed trustees but “take on themselves to act as such and to possess and administer trust property”; see Selangor United Rubber Estates, Ltd. v. Cradock (No. 3), [1968] 2 All E.R. 1073 (Ch.), at p. 1095. This type of liability is inapplicable to the present case. The Bank never assumed the office or function of trustee; nor did it administer the trust funds on behalf of the beneficiary Citadel.
21 The two remaining categories of liability, namely “knowing assistance” and “knowing receipt”, relate to strangers to the trust who knowingly participate in a breach of trust. In Air Canada v. M & L Travel Ltd., supra, this Court considered the requirements for bringing a case within the “knowing assistance” category. In that case, the defendant travel agency collected funds from the sale of Air Canada tickets and held them in trust to be remitted to Air Canada. The funds were kept in the agency’s general bank account. The individual directors of the travel agency, who had personally guaranteed a demand loan, authorized the bank to withdraw funds from the general account to cover monies owing on the loan. A dispute arose between the directors with regard to misappropriation of funds. The bank sent demand notices to the directors and withdrew the full amount owing under the loan from the agency’s general account. As a result, Air Canada did not receive monies owed to it for ticket sales. The issue arose whether the appellant director, as stranger to the trust, was liable to Air Canada for the travel agency’s breach of trust.
22 This Court found the director liable for knowingly assisting in a breach of trust. The liability of the director was based on his knowledge of, and assistance in, a fraudulent and dishonest breach of trust on the part of the trustees. With regard to the knowledge requirement, Iacobucci J. wrote for the majority, at p. 811: “The knowledge requirement for this type of liability is actual knowledge; recklessness or wilful blindness will also suffice.” He expressly excluded constructive knowledge from this test. Iacobucci J. defined constructive knowledge, at p. 812, as “knowledge of circumstances which would indicate the facts to an honest person, or knowledge of facts which would put an honest person on inquiry”.
23 The Court of Appeal in the present case relied on the knowledge requirement set out in Air Canada v. M & L Travel Ltd., supra. Assuming the present case falls under the “knowing assistance” category, it is clear that only actual knowledge, recklessness, or wilful blindness will render the Bank liable for participating in the breach of trust. Constructive knowledge will not suffice. The trial judge’s conclusions regarding the knowledge of the Bank were as follows (at para. 33):
The Bank had actual knowledge of the nature of the funds in the Drive On account and had an obligation to inquire about their position in the circumstances. The Bank shut its eyes in circumstances which should have caused it to inquire of its customer at least. It did not do so. The Bank had constructive knowledge and is a constructive trustee. . . .
Kerans J.A. considered this passage from the trial judge’s reasons and concluded, at p. 77:
It is true that the judge used the words “shut its eyes”, but reading the passage in its entirety, it seems clear that the judge is refusing to say that the bank was actually aware that it was taking money in breach of trust, as opposed to what it should have known or what its duty was.
I agree with Kerans J.A. that the trial judge refused to make a finding of actual knowledge by the Bank. Rather, the trial judge restricted his findings to constructive knowledge, based on the Bank’s duty to inquire of its customer in the circumstances. Moreover, there was no finding of recklessness or wilful blindness as such. It follows from the trial judge’s findings that the Bank does not meet the knowledge requirement set out in Air Canada v. M & L Travel Ltd., supra. Since the Bank had only constructive knowledge, it cannot be liable under the “knowing assistance” category of constructive trusteeship.
24 The only basis upon which the Bank may be held liable as a constructive trustee is under the “knowing receipt” or “knowing receipt and dealing” head of liability. Under this category of constructive trusteeship it is generally recognized that there are two types of cases. First, although inapplicable to the present case, there are strangers to the trust, usually agents of the trustees, who receive trust property lawfully and not for their own benefit but then deal with the property in a manner inconsistent with the trust. These cases may be grouped under the heading “knowing dealing”. Secondly, there are strangers to the trust who receive trust property for their own benefit and with knowledge that the property was transferred to them in breach of trust. In all cases it is immaterial whether the breach of trust was fraudulent; see Halsbury’s Laws of England (4th ed. 1995), vol. 48, at para. 595; Pettit, supra, at p. 168; Underhill and Hayton, Law Relating to Trusts and Trustees (14th ed. 1987), at p. 357. The second type of case, which is relevant to the present appeal, raises two main issues: the nature of the receipt of trust property and the degree of knowledge required of the stranger to the trust.
2. Liability for Knowing Receipt
(a) The Receipt Requirement
25 Liability on the basis of “knowing receipt” requires that strangers to the trust receive or apply trust property for their own use and benefit; see Agip (Africa) Ltd. v. Jackson, [1990] 1 Ch. 265, aff’d [1992] 4 All E.R. 451 (C.A.); Halsbury’s Laws of England, supra, at paras. 595-96; Pettit, supra, at p. 168. As Iacobucci J. wrote in Air Canada v. M & L Travel Ltd., supra, at pp. 810-11, the “knowing receipt” category of liability “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”. In the banking context, which is directly applicable to the present case, the definition of receipt has been applied as follows:
The essential characteristic of a recipient . . . is that he should have received the property for his own use and benefit. That is why neither the paying nor the collecting bank can normally be made liable as recipient. In paying or collecting money for a customer the bank acts only as his agent. It sets up no title of its own. It is otherwise, however, if the collecting bank uses the money to reduce or discharge the customer’s overdraft. In doing so it receives the money for its own benefit. . . . [Footnotes omitted.]
P. J. Millett, “Tracing the Proceeds of Fraud” (1991), 107 L.Q.R. 71, at pp. 82-83.
26 Thus, a distinction is traditionally made between a bank receiving trust funds for its own benefit, in order to pay off a bank overdraft (“knowing receipt”), and a bank receiving and paying out trust funds merely as agent of the trustee (“knowing assistance”); see Underhill and Hayton, supra, at p. 361.
27 In the present case, we saw, Drive On deposited trust funds, namely insurance premiums collected on Citadel’s behalf, in an operating account at the Bank. Drive On’s parent company, International Warranty, also had an account at the Bank. In April 1987, the Bank transferred funds between the Drive On and International Warranty accounts to cover overdrafts in either account. As well, in June, the Bank transferred the balance of any funds in the Drive On account to the International Warranty account on a regular basis. As a result of the transfers between the accounts in July and August, a net amount was transferred to the International Warranty account. This amount, which was in excess of the July and August premiums deposited by Drive On, was used to reduce International Warranty’s overdraft. Although the Bank was instructed by Drive On’s signing officers to make the transfers, the Bank did not act as mere agent in the circumstances. The Bank’s actions went beyond the mere collection of funds and payment of bills on Drive On’s behalf. The Bank, by applying the deposit of insurance premiums as a set-off against International Warranty’s overdraft, received a benefit. This benefit, of course, was the reduction in the amount owed to the Bank by one of its customers. It follows that the Bank received the trust funds for its own use and benefit.
28 In this Court, the respondents argued that they could not be liable on the basis of “knowing receipt” because they had not received the trust property. The respondents took the position, accepted by the authorities, that a bank deposit is simply a loan to the bank; see Foley v. Hill (1848), [1843-60] All E.R. Rep. 16 (H.L.); Fonthill Lumber Ltd. v. Bank of Montreal (1959), 19 D.L.R. (2d) 618 (Ont. C.A.), at p. 628. Accordingly, the deposit of money in Drive On’s account was characterized as a debt obligation owed by the Bank to Drive On. This debt obligation gave rise to a credit in Drive On’s favour. On instruction from its customer, the Bank simply transferred “credits” from Drive On’s account to International Warranty’s account. The transfer of credits had the incidental effect of reducing an overdraft in the International Warranty account. In other words, the transfers between the accounts in July and August simply amounted to an off-setting of debt obligations. In the respondents’ view, the Bank was not receiving trust property but simply transferring credits from one account to another.
29 The respondents’ arguments are not convincing. A debt obligation is a chose in action and, therefore, property over which one can impose a trust. This conclusion is supported by the House of Lords’ decision in Lipkin Gorman v. Karpnale Ltd., [1991] 3 W.L.R. 10. In that case, a firm of solicitors was authorized to operate a client’s bank account. One of the firm’s partners subsequently stole funds from the account and used them for casino gambling. Considering whether the solicitors could trace their client’s funds at common law, Lord Goff of Chieveley wrote, at pp. 28-29:
The relationship of the bank with the solicitors was essentially that of debtor and creditor; and since the client account was at all material times in credit, the bank was the debtor and the solicitors were its creditors. Such a debt constitutes a chose in action, which is a species of property; and since the debt was enforceable at common law, the chose in action was legal property belonging to the solicitors at common law.
The respondents cannot avoid the “property” issue by characterizing the deposit of trust monies in Drive On’s account as a debt obligation. The chose in action, constituted by the indebtedness of the Bank to Drive On, was subject to a statutory trust in Citadel’s favour. That same chose in action can also be the subject of a constructive trust in Citadel’s favour.
30 Nonetheless, the respondents’ arguments reflect a difficulty with the traditional conception of “receipt” in “knowing receipt” cases. In my view, the receipt requirement for this type of liability is best characterized in restitutionary terms. In Lac Minerals Ltd. v. International Corona Resources Ltd., [1989] 2 S.C.R. 574, at p. 669, I stated that a restitutionary claim, or a claim for unjust enrichment, is concerned with giving back to someone something that has been taken from them (a restitutionary proprietary award) or its equivalent value (a personal restitutionary award). As well, in Air Canada v. British Columbia, [1989] 1 S.C.R. 1161, at pp. 1202-3, I stated that the function of the law of restitution “is to ensure that where a plaintiff has been deprived of wealth that is either in his possession or would have accrued for his benefit, it is restored to him. The measure of restitutionary recovery is the gain the [defendant] made at the [plaintiff’s] expense.” In the present case, the Bank was clearly enriched by the off-setting of debt obligations, or transferring of credits between the Drive On and International Warranty accounts. That is, the amount due to the Bank was reduced. As well, the Bank’s enrichment deprived Citadel of the insurance premiums collected on its behalf. Moreover, the fact that the insurance premiums were never in Citadel’s possession does not preclude Citadel from pursuing a restitutionary claim. After all, the insurance premiums would have accrued to Citadel’s benefit. The Bank has been enriched at Citadel’s expense. Thus, in restitutionary terms, there can be no doubt that the Bank received trust property for its own use and benefit.
(b) The Knowledge Requirement
31 The first requirement for establishing liability on the basis of “knowing receipt” has been satisfied. The Bank received the trust property for its own benefit and, in doing so, was enriched at the beneficiary’s expense. The second requirement relates to the degree of knowledge required of the Bank in relation to the breach of trust. With regard to this knowledge requirement, there are two lines of authorities. According to one line of jurisprudence, the knowledge requirement for both “knowing assistance” and “knowing receipt” cases should be the same. More specifically, constructive knowledge should not be the basis for liability in either type of case. A second line of authority suggests that a different standard should apply in “knowing assistance” and “knowing receipt” cases. More specifically, the authorities favour a lower threshold of knowledge in “knowing receipt” cases.
32 A leading case in relation to the first line of authority is In re Montagu’s Settlement Trusts, [1987] 1 Ch. 264. That case involved a dispute arising out of a 1923 settlement in which the future tenth Duke of Manchester had made an assignment of certain chattels to a number of trustees. The trustees were under a fiduciary duty to select and make an inventory of the chattels after the ninth Duke of Manchester died. The selection and inventory did not occur and the tenth Duke took absolutely whatever chattels he wanted. Megarry V.-C. held that the Duke was not liable as a constructive trustee because he did not know that the chattels were subject to a trust. In discussing the degree of knowledge required of the Duke, Megarry V.-C. emphasized that liability in “knowing receipt” cases is personal in nature and arises only if the stranger’s conscience is sufficiently affected to justify imposing a constructive trust. Although cases involving actual knowledge, recklessness, and wilful blindness justify imposing a constructive trust, Megarry V.-C. doubted, at p. 285, whether the carelessness associated with constructive knowledge cases could sufficiently bind the stranger’s conscience.
33 In re Montagu’s Settlement Trusts was followed in Polly Peck International plc v. Nadir (No. 2), [1992] 4 All E.R. 769 (C.A.). There, the plaintiff company sought to impose liability on a bank for assisting in the misapplication of trust funds and for receiving and dealing in some way with trust property. The bank had been instructed by the trustee to transfer substantial trust funds into offshore accounts. Scott L.J. dealt with the case on the basis of both “knowing assistance” and “knowing receipt” because some of the transfers were made by the banker as agent while others were received for the bank’s own benefit. With regard to the “knowing receipt” claim, the plaintiff beneficiary argued that, in the circumstances, the bank should have been put on inquiry as to whether there were improprieties in the transfers. Addressing the “knowing receipt” claim, Scott L.J. commented, at p. 777:
Liability as constructive trustee in a ‘knowing receipt’ case does not require that the misapplication of the trust funds should be fraudulent. It does require that the defendant should have knowledge that the funds were trust funds and that they were being misapplied. Actual knowledge obviously will suffice. Mr. Potts [lawyer for the plaintiff] has submitted that it will suffice if the defendant can be shown to have had knowledge of facts which would have put an honest and reasonable man on inquiry, or, at least, if the defendant can be shown to have wilfully and recklessly failed to make such inquiries as an honest and reasonable man would have made. . . . I do not think there is any doubt that, if the latter of the two criteria can be established against the Central Bank, that will suffice. I have some doubts about the sufficiency of the former criterion but do not think that the present appeal is the right occasion for settling the issue.
It should be noted that Scott L.J. went on to apply the test for constructive knowledge, but found that the bank was not liable because it did not have cause to suspect improprieties and was not put on inquiry.
34 The English approach favouring exclusion of constructive knowledge received the approval of the Manitoba Court of Appeal in C.I.B.C. v. Valley Credit Union Ltd., [1990] 1 W.W.R. 736. In that case, a business obtained a line of credit from the plaintiff bank. Under the bank’s general security agreement, the customer became trustee of monies paid to it with respect to accounts receivable or sales of inventory. The customer subsequently opened an account with the defendant credit union and used this account to deposit trust monies. The bank became aware of the other account, eventually called the customer’s loans, and brought an action against the credit union to recover the funds in the customer’s account. Philp J.A. refused to find the credit union liable as a constructive trustee. Without distinguishing between the categories of “knowing assistance” and “knowing receipt”, Philp J.A. doubted whether the carelessness associated with constructive knowledge was sufficient to impose liability on the bank as a constructive trustee. Relying in part on In re Montagu’s Settlement Trusts, supra, he stated, at p. 747:
I do not think that it can be said that it has been authoritatively decided in Canada that carelessness or negligence is sufficient to impute constructive knowledge to a stranger, and to impose upon him liability as a constructive trustee. I think that it is a doubtful test, particularly in the case of a bank. The relationship between a bank and its customer is contractual and a principal obligation of the bank is to pay out as directed the moneys its customer has deposited. It seems to me that that obligation should be a paramount one, save in special factual circumstances sufficient to hold the bank privy to its customer’s breach.
It should be noted, however, that later in his reasons Philp J.A. applied the test for constructive knowledge, but found that it had not been met in the circumstances.
35 That constructive knowledge should be excluded as a basis for liability in “knowing receipt” cases is also supported by the Ontario Court of Appeal’s decision in Bullock v. Key Property Management Inc. (1997), 33 O.R. (3d) 1. There, a trustee had deposited trust funds in a bank account. The funds were then used to service the trustee’s own interests, including the reduction of the trustee’s indebtedness to the bank. The court dismissed the action against the bank on the grounds that it did not have the requisite degree of knowledge of the breach of trust. The court did not deal with the categories of “knowing assistance” and “knowing receipt” and apparently found it unnecessary to distinguish between the various heads of liability. Apparently assuming that there was only one category of liability, the court concluded, at p. 4:
As the law presently stands, a stranger to a trust will be held liable for a breach of that trust by the trustee only where the stranger has actual knowledge or is reckless or wilfully blind as to both the existence of the trust and the dishonest conduct of the trustee in connection with the trust. The inquiry must be directed to what the stranger to the trust actually knew or suspected and not to what the stranger would have known had reasonable inquiries been made. Failure to make reasonable inquiries may have evidentiary value in determining what the stranger to the trust in fact knew or suspected, but it is not a basis for the imposition of liability as a constructive trustee.
36 As well, the Court of Appeal in the present case concluded, without restricting its comments to any particular head of liability, that constructive knowledge should be excluded as a basis for imposing liability on a stranger to the trust. Relying on Iacobucci J.’s reasons in Air Canada v. M & L Travel Ltd., supra, Kerans J.A. wrote, at p. 77, that “[a] stranger to a trust is not liable for a breach of trust from which it received a benefit unless it had both actual knowledge of the trust and participated in the breach”.
37 According to a second line of authority, however, constructive knowledge is sufficient to find a stranger to the trust liable on the basis of “knowing receipt”. A leading English authority, in terms of formulating the test for constructive knowledge in breach of trust cases, is Selangor, supra. There, a company director carried out a fraudulent takeover bid by using the company’s funds to purchase its own shares. Two banks were involved in the takeover. One bank acted on behalf of the director by paying, for a fee, those shareholders who had agreed to sell. The bank’s fee was paid for by way of an advance from a second bank, where the company’s account had been transferred. The second bank was repaid with trust funds drawn from the company’s account. In addressing the banks’ liability, Ungoed-Thomas J. did not distinguish between receipt and assistance cases. He presumed, at p. 1095, that there was only one category of liability for strangers to the trust who, unlike trustees de son tort, “act in their own right and not for beneficiaries”. Relying on this single category of liability, Ungoed-Thomas J. held, at p. 1104:
The knowledge required to hold a stranger liable as constructive trustee in a dishonest and fraudulent design, is knowledge of circumstances which would indicate to an honest, reasonable man that such a design was being committed or would put him on enquiry, which the stranger failed to make, whether it was being committed.
Ungoed-Thomas J. found both banks liable as constructive trustees.
38 The Selangor decision was followed by the British Columbia Court of Appeal in Groves-Raffin Construction Ltd. v. Bank of Nova Scotia (1975), 64 D.L.R. (3d) 78. There, a construction company deposited trust funds collected from building contracts with the defendant bank. These trust funds were then used to repay the company’s indebtedness to the bank and to reduce personal overdrafts belonging to a director of the company. As well, the director stole trust monies from the company’s account and transferred them to a personal account at a second bank. Addressing the liability of the second bank as a constructive trustee, Robertson J.A. wrote, at p. 136, that he was dealing with a “collecting bank” and not a “paying bank”, thereby suggesting that the case fell under the “knowing receipt” category. Relying in part on Selangor, supra, Robertson J.A. found, at p. 138:
Under what I think is the proper test no necessity to take care arises until either it is clear that a breach of trust is being, or is intended to be, committed, or until there has come to the attention of the person something that should arouse suspicion in an honest, reasonable man and put him on inquiry. The person, for his own protection, in the first event should have nothing to do with the improper transaction, and in the second event should not continue to be involved in the suspected transaction until his inquiry shows him — or, more correctly, would show a reasonable man — that the suspicion is unfounded.
39 The Selangor decision was also applied by this Court in Carl B. Potter Ltd. v. Mercantile Bank of Canada, [1980] 2 S.C.R. 343. In that case, the plaintiff bid for the construction of a waste treatment plant. The bid was accompanied by a tender deposit cheque, to be held in trust by the owner of the project. The proceeds of the tender cheque eventually found their way into the owner’s collateral account where they were drawn upon to meet the owner’s obligations to the defendant bank. In these circumstances, it appears that the trust funds were received by the bank for its own use and benefit, thereby meeting the first requirement under the “knowing receipt” head of liability. Ritchie J., at p. 347, approved of the following test regarding the bank’s knowledge of the breach of trust:
The position of a banker who has been placed “on inquiry” in the manner aforesaid is summarized in the following brief paragraph from Halsbury’s Laws of England (4th ed.) vol. III, para. 60:
A banker may be a constructive trustee of money in his customer’s account and in breach of that trust if he pays the money away, even on the customer’s mandate, in circumstances which put him upon inquiry.
The footnote references for this passage, although not referred to by Ritchie J., included Selangor, supra. Ritchie J. went on to apply this test to the facts and found the bank liable for breach of trust.
40 A similar test of constructive knowledge was applied by the Ontario Court of Appeal in Arthur Andersen Inc. v. Toronto-Dominion Bank (1994), 17 O.R. (3d) 363, leave to appeal refused, [1994] 3 S.C.R. v. In that case, the defendant bank was sued by a trustee appointed under the Construction Lien Act, R.S.O. 1990, c. C.30. The bank had agreed to administer the accounts of a number of associated construction companies, in accordance with a “mirror accounting system”. Among other things, this system eliminated the need to monitor overdrafts in individual accounts and permitted the informal transfer of debits and credits between all of the operating companies’ accounts. The trial judge’s findings implied that the funds in the accounts were transferred in breach of the trust requirements under the Construction Lien Act. Considering the liability of the bank as a constructive trustee, Grange and McKinlay JJ.A. thus wrote in their joint reasons, at pp. 381-82:
We consider that the law on this point can be summarized thus: in the absence of sufficient facts or circumstances indicating that there is a good possibility of trust beneficiaries being unpaid there is no duty of inquiry on a bank to determine whether the trades have been paid or will be able to be paid.
. . .
Only if a bank is aware of facts which would indicate that trades would not be paid in the normal course of business should it be charged with a duty of special inquiry.
It should be noted that Grange and McKinlay JJ.A. formulated this test without distinguishing between receipt and assistance cases. However, their comment, at p. 385, that the “Bank can only be liable for a breach of trust, and that breach would have to involve making use for its own benefit of money held on a trust for trade creditors”, suggests that the case fell under the “knowing receipt” head of liability.
41 This analysis was endorsed by the Manitoba Court of Appeal in Glenko Enterprises Ltd. v. Ernie Keller Contractors Ltd. (1996), 134 D.L.R. (4th) 161. On facts similar to Arthur Andersen, supra, the court, at pp. 164-65, considered “whether a bank which has applied trust funds received from a building contractor to reduce an account overdraft has participated in a breach of trust and must therefore account to the beneficiaries of that trust for those funds”. The funds in question, which were misappropriated by the defendant contractor, were subject to trust requirements under the Builders’ Liens Act, R.S.M. 1987, c. B91. Considering whether the bank was a party to a breach of trust by its customer, Scott C.J.M. held, at p. 167:
. . . the Bank is not liable for the builder’s breach of trust if the Bank, in the ordinary course of business, accepted deposits and allowed cheques to be written thereon — or for that matter if it applied the funds on the overdraft — unless it had or clearly should have had knowledge of the breach of trust by the contractor or of facts to put it on notice.
The court went on to agree with the trial judge, at p. 176, that the bank, although it apparently received the trust funds for its own benefit, was not liable because “the inquiries and arrangements for further information which were made by the bank . . . were reasonable in all the circumstances”.
42 There are also a number of recent English authorities supporting the view that constructive knowledge is sufficient to impose liability on the basis of “knowing receipt”. In Agip (Africa) Ltd. (Ch.), supra, Millett J. made a number of comments regarding “knowing receipt” cases, even though the case before him was of the “knowing assistance” category. With regard to the degree of knowledge required in “knowing receipt” cases, he wrote, at p. 291:
The first [category of “knowing receipt” cases] is concerned with the person who receives for his own benefit trust property transferred to him in breach of trust. He is liable as a constructive trustee if he received it with notice, actual or constructive, that it was trust property and that the transfer to him was a breach of trust; or if he received it without such notice but subsequently discovered the facts.
However, Millett J.’s comments must be read in light of a later passage, at p. 293, where he refused to express an opinion as to whether constructive knowledge sufficed in “knowing receipt” cases.
43 Millett J.’s comments were subsequently referred to by Knox J. in Cowan de Groot Properties Ltd. v. Eagle Trust plc, [1992] 4 All E.R. 700 (Ch.). In this “knowing receipt” case the issue arose whether a purchaser company had knowledge of a breach of duty arising out of the sale of another company’s property. The purchaser’s liability as a constructive trustee turned on whether or not it had knowledge that the directors of the vendor company were deliberately selling at a gross undervalue. Knox J. noted, at p. 758, that there was a “substantial body of authority in favour of the proposition that constructive notice based on what a reasonable man would have concluded though falling short of want of probity on the part of the person charged as a constructive trustee may suffice in a knowing receipt case”. Despite this body of authority, Knox J. preferred a test based on actual knowledge, wilful blindness, or recklessness. However, he added that if, contrary to his view, constructive knowledge was sufficient, there would still have been no liability on the facts before him. As well, at p. 761, he suggested “that the underlying broad principle which runs through the authorities regarding commercial transactions is that the court will impute knowledge, on the basis of what a reasonable person would have learnt, to a person who is guilty of commercially unacceptable conduct in the particular context involved”.
44 Millett J. reiterated the views he expressed in Agip (Africa) Ltd., supra, in El Ajou v. Dollar Land Holdings plc, [1993] 3 All E.R. 717 (Ch.). That case involved a massive share fraud carried out by three Canadians in Amsterdam between 1984 and 1985. The plaintiff, the largest single victim of the fraud, claimed to be able to trace some of the proceeds of the fraud from Amsterdam through locations in Geneva, Gibraltar, Panama, back to Geneva, and then to London, where they were invested in a joint venture to carry out a property development project. In this “knowing receipt” case, Millett J. was prepared to assume that constructive knowledge was a sufficient basis for liability. At p. 739, he stated:
In the absence of full argument I am content to assume, without deciding, that dishonesty or want of probity involving actual knowledge (whether proved or inferred) is not a precondition of liability; but that a recipient is not expected to be unduly suspicious and is not to be held liable unless he went ahead without further inquiry in circumstances in which an honest and reasonable man would have realised that the money was probably trust money and was being misapplied.
45 According to the second line of authority, then, the degree of knowledge required of strangers to the trust should be different in assistance and receipt cases. Generally, there are good reasons for requiring different thresholds of knowledge under the two heads of liability. As Millett J. wrote in Agip (Africa) Ltd., supra, at pp. 292-93:
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. 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.
46 In other words, the distinction between the two categories of liability is fundamental: whereas the accessory’s liability is “fault-based”, the recipient’s liability is “receipt-based”. In an extrajudicial opinion, Millett J. described the distinction as follows:
. . . the liability of the accessory is limited to the case where the breach of trust in question was fraudulent and dishonest; the liability of the recipient is not so limited. In truth, however, the distinction is fundamental; there is no similarity between the two categories. The accessory is a person who either never received the property at all, or who received it in circumstances where his receipt was irrelevant. His liability cannot be receipt-based. It is necessarily fault-based, and is imposed on him not in the context of the law of competing priorities to property, but in the application of the law which is concerned with the furtherance of fraud. [Footnotes omitted.]
“Tracing the Proceeds of Fraud”, supra, at p. 83.
47 S. Gardner makes a similar point in “Knowing Assistance and Knowing Receipt: Taking Stock” (1996), 112 L.Q.R. 56, at p. 85:
. . . it is questionable whether knowing receipt is about wrongfully causing loss at all. There may be more than one other thing that it could be about, but most modern opinion takes it to be a restitutionary liability, based on the fact that the defendant has acquired the plaintiff’s property.
The same view was expressed by the Privy Council in Royal Brunei Airlines Sdn. Bhd. v. Tan, [1995] 3 W.L.R. 64, at p. 70: “Different considerations apply to the two heads of liability. Recipient liability is restitution-based; accessory liability is not.” These comments are also cited with approval by Iacobucci J. in Gold, supra, at para. 41.
48 Given the fundamental distinction between the nature of liability in assistance and receipt cases, it makes sense to require a different threshold of knowledge for each category of liability. In “knowing assistance” cases, which are concerned with the furtherance of fraud, there is a higher threshold of knowledge required of the stranger to the trust. Constructive knowledge is excluded as the basis for liability in “knowing assistance” cases; see Air Canada v. M & L Travel Ltd., supra, at pp. 811-13. However, in “knowing receipt” cases, which are concerned with the receipt of trust property for one’s own benefit, there should be a lower threshold of knowledge required of the stranger to the trust. More is expected of the recipient, who, unlike the accessory, is necessarily enriched at the plaintiff’s expense. Because the recipient is held to this higher standard, constructive knowledge (that is, knowledge of facts sufficient to put a reasonable person on notice or inquiry) will suffice as the basis for restitutionary liability. Iacobucci J. reaches the same conclusion in Gold, supra, where he finds, at para. 46, that a stranger in receipt of trust property “need not have actual knowledge of the equity [in favour of the plaintiff]; notice will suffice”.
49 This lower threshold of knowledge is sufficient to establish the “unjust” or “unjustified” nature of the recipient’s enrichment, thereby entitling the plaintiff to a restitutionary remedy. As I wrote in Lac Minerals, supra, at p. 670, “[t]he determination that the enrichment is ‘unjust’ does not refer to abstract notions of morality and justice, but flows directly from the finding that there was a breach of a legally recognized duty for which the courts will grant relief”. In “knowing receipt” cases, relief flows from the breach of a legally recognized duty of inquiry. More specifically, relief will be granted where a stranger to the trust, having received trust property for his or her own benefit and having knowledge of facts which would put a reasonable person on inquiry, actually fails to inquire as to the possible misapplication of trust property. It is this lack of inquiry that renders the recipient’s enrichment unjust.
50 Some commentators go further and argue that a recipient may be unjustly enriched regardless of either a duty of inquiry or constructive knowledge of a breach of trust. According to Professor Birks, a recipient of misdirected funds should be liable on a strict, restitutionary basis. In his article “Misdirected Funds: Restitution from the Recipient”, [1989] L.M.C.L.Q. 296, he argues that a recipient’s enrichment is unjust because the plaintiff did not consent to it, not because the defendant knew that the funds were being misdirected. In particular, he writes, at p. 341, that “[t]he ‘unjust’ factor can be named ‘ignorance’, signifying that the plaintiff, at the time of the enrichment, was absolutely unaware of the transfer from himself to the defendant”. Birks, however, lessens the strictness of his approach by allowing a defendant to take advantage of special defences, including a defence arising out of a bona fide purchase for value. (See also P. Birks, “Overview: Tracing, Claiming and Defences”, in P. Birks, ed., Laundering and Tracing (1995), 289, at pp. 322 et seq.)
51 In my view, the test formulated by Professor Birks, while not entirely incompatible with my own, may establish an unjust deprivation, but not an unjust enrichment. It is recalled that a plaintiff is entitled to a restitutionary remedy not because he or she has been unjustly deprived but, rather, because the defendant has been unjustly enriched, at the plaintiff’s expense. To show that the defendant’s enrichment is unjustified, one must necessarily focus on the defendant’s state of mind not the plaintiff’s knowledge, or lack thereof. Indeed, without constructive or actual knowledge of the breach of trust, the recipient may very well have a lawful claim to the trust property. It would be unfair to require a recipient to disgorge a benefit that has been lawfully received. In those circumstances, the recipient will not be unjustly enriched and the plaintiff will not be entitled to a restitutionary remedy.
52 In the banking context of the present case, it is true that s. 206(1) of the Bank Act, R.S.C., 1985, c. B-1, negates any duty on the part of a bank to see to the execution of any trust, whether express, implied or constructive, to which a deposit is subject. In accordance with this provision, a bank is not under a duty to regularly monitor the activities of its clients simply because the funds deposited by those clients are impressed with a statutory trust. Nonetheless, this provision does not render a bank immune from liability as a constructive trustee or prevent the recognition of a duty of inquiry on the part of a bank. Indeed, in certain circumstances, a bank’s knowledge of its customer’s affairs will require the bank to make inquiries as to possible misapplication of trust funds. As discussed earlier, the degree of knowledge required is constructive knowledge of a possible breach of trust. It follows that a bank which is enriched by the receipt of trust property and has knowledge of facts that would put a reasonable person on inquiry is under a duty to make inquiries of its customer regarding a possible breach of trust. If the bank fails to make the appropriate inquiries, it will have constructive knowledge of the breach of trust. In these circumstances, the bank will be unjustly enriched and, therefore, required to disgorge the benefit it received at the plaintiff’s expense.
53 The respondents argued that imposing liability on a banker who merely has constructive notice of a breach of trust will place too great a burden on banks, thereby interfering with the proper functioning of the banking system. While this may be true in assistance cases where a banker merely pays out and transfers funds as the trustee’s agent, the same argument does not apply to receipt cases where a banker receives the trust funds for his or her own benefit. Professor Harpum addresses this point in “The Stranger as Constructive Trustee” (1986), 102 L.Q.R. 114, at p. 138:
Although there should be a reluctance to allow the unnecessary intrusion of “the intricacies and doctrines connected with trusts” into ordinary commercial transactions, considerations of speed and the importance of possession which normally justify the exclusion of these doctrines, are less applicable to a banker who chooses to exercise his right of set-off than they are to other commercial dealings. Where a banker combines accounts, he alone stands to gain from the transaction. Because of that benefit, more should be expected of him than if he gained nothing. [Footnotes omitted.]
In “knowing receipt” cases, therefore, it is justifiable to impose liability on a banker who only has constructive knowledge of a breach of trust.
54 In the present case, it has already been established that the Bank was enriched at Citadel’s expense by the receipt of insurance premiums collected by Drive On and subject to a statutory trust in favour of Citadel. The only remaining question is whether the Bank had the requisite degree of knowledge to render the enrichment unjust, thereby entitling the plaintiff insurer to a remedy.
55 On this issue, it is clear from the trial judge’s findings that the Bank was aware of the nature of the funds being deposited into, and transferred out of, Drive On’s account. On discovery, two of the Bank’s employees stated that they knew Drive On’s sole source of revenue was the sale of insurance policies. The Bank also knew that premiums collected by Drive On were payable to the plaintiff insurer. The Bank’s knowledge of the nature of Drive On’s deposits must also be considered in conjunction with the activities in Drive On’s account. It is recalled that in April 1987 the Bank began transferring funds between the Drive On and International Warranty accounts to cover overdrafts in either account. As well, in June 1987, the Bank was directed to empty the Drive On account on a daily basis, again to facilitate the transfer of funds to the International Warranty account.
56 In light of the Bank’s knowledge of the nature of the funds, the daily emptying of the account was in the trial judge’s view “very suspicious”. In these circumstances, a reasonable person would have been put on inquiry as to the possible misapplication of the trust funds. Notwithstanding the fact that the exact terms of the trust relationship between Citadel and Drive On may have been unknown to the Bank, the Bank should have taken steps, in the form of reasonable inquiries, to determine whether the insurance premiums were being misapplied. More specifically, the Bank should have inquired whether the use of the premiums to reduce the account overdrafts constituted a breach of trust. By failing to make the appropriate inquiries, the Bank had constructive knowledge of Drive On’s breach of trust. In these circumstances, the Bank’s enrichment was clearly unjust, thereby rendering it liable to Citadel as a constructive trustee.
57 I make one additional point regarding the nature of the Bank’s liability in the present case. As already established, recipient liability is restitution-based. The imposition of liability as a constructive trustee on the basis of “knowing receipt” is a restitutionary remedy and should not be confused with the right to trace assets at common law or in equity. The principles relating to tracing at law and in equity were thus set out by the English Court of Appeal in Agip (Africa) Ltd., supra, at pp. 463-64 and 466:
Tracing at law does not depend upon the establishment of an initial fiduciary relationship. Liability depends upon receipt by the defendant of the plaintiff’s money and the extent of the liability depends on the amount received. Since liability depends upon receipt the fact that a recipient has not retained the asset is irrelevant. For the same reason dishonesty or lack of inquiry on the part of the recipient are irrelevant. Identification in the defendant’s hands of the plaintiff’s asset is, however, necessary. It must be shown that the money received by the defendant was the money of the plaintiff. Further, the very limited common law remedies make it difficult to follow at law into mixed funds.
. . .
Both common law and equity accepted the right of the true owner to trace his property into the hands of others while it was in an identifiable form. The common law treated property as identified if it had not been mixed with other property. Equity, on the other hand, will follow money into a mixed fund and charge the fund.
58 In my view, a distinction should be made between the imposition of liability in “knowing receipt” cases and the availability of tracing orders at common law and in equity. Liability at common law is strict, flowing from the fact of receipt. Liability in “knowing receipt” cases is not strict; it depends not only on the fact of enrichment (i.e. receipt of trust property) but also on the unjust nature of that enrichment (i.e. the stranger’s knowledge of the breach of trust). A tracing order at common law, unlike a restitutionary remedy, is only available in respect of funds which have not lost their identity by becoming part of a mixed fund. Further, the imposition of liability as a constructive trustee is wider than a tracing order in equity. The former is not limited to the defence of purchaser without notice and “does not depend upon the recipient still having the property or its traceable proceeds”; see In re Montagu’s Settlement Trusts, supra, at p. 276.
59 Despite these distinctions, there appears to be a common thread running through both “knowing receipt” and tracing cases. That is, constructive knowledge will suffice as the basis for imposing liability on the recipient of misdirected trust funds. Notwithstanding this, it is neither necessary nor desirable to confuse the traditional rules of tracing with the restitutionary principles now applicable to “knowing receipt” cases. This does not mean, however, that a restitutionary remedy and a tracing order are mutually exclusive. Where more than one remedy is available on the facts, the plaintiff should be able to choose the one that is most advantageous. In the present case, the plaintiff did not seek a tracing order. It is therefore unnecessary for me to decide whether such a remedy would have been available on the facts of the present appeal, and I have not explored the issue.
V. Disposition
60 For these reasons, I would allow Citadel’s appeal with costs and restore the judgment rendered at trial.
//Sopinka J.//
The following are the reasons delivered by
61 Sopinka J. -- Subject to my reasons in Gold v. Rosenberg, [1997] 3 S.C.R. 767, I agree with Justice La Forest.
Appeal allowed with costs.
Solicitors for the appellants: Duncan & Craig, Edmonton.
Solicitors for the respondents: Cruickshank Karvellas, Edmonton.