Quincecare in the Hong Kong Court of Final Appeal

This note addresses issues of agency law that arise out of the judgment of Lord Sumption for the Hong Kong Court of Final Appeal in PT Asuransi Tugu Pratama Indonesia TBK v Citibank NA [2023] HKCFA 3 (‘Tugu’). The facts of the case involved what has come to be called a Quincecare claim, in typical form a claim against a bank for allowing funds to be withdrawn from an account by a dishonest mandatary of the customer when the bank ought to have detected that dishonesty.

An important issue of principle arises, almost incidentally, from Lord Sumption’s judgment. Assume that an agent, X, has a contract with its principal, P, that confers direct actual authority from P to make payments on P’s behalf, and a right to make book entries to reimburse itself for them, but is only to exercise that authority when directed to do so by another agent, Y. What is the position if Y, in giving a direction to X, is acting dishonestly in relation to P? Is X’s actual authority to follow the directions of Y automatically removed by the fact that Y is acting dishonestly?

In Tugu, a positive answer to that question appears to have been simply assumed by both parties (see at [15]). In those circumstances, it is perhaps unsurprising that Lord Sumption assumed the answer was correct. Hence, the defendant bank (the party in the position of X) accepted that it could rely only on the apparent authority of the relevant mandataries (Y) to give the directions in question. This in turn led to the further concession that it would be difficult for the bank to rely on apparent authority given the circumstances of which it was aware at the time of the payments. Instead, the bank fell back on the fact that the relevant payments had been made between 1994 and 1998, whereupon, the bank asserted, the account had been closed. Since the claim was not commenced until 2007, the bank argued that the proceedings were barred by limitation. It is not apparent why the case then took so long to get to trial.

Given the concessions of the bank, and the degree of actual knowledge the bank appears to have had as to the way in which the account was being operated (the mandataries appear to have made little attempt to disguise the fact that they were paying the principal’s funds directly to themselves), Tugu can fairly be said to be a weak precedent for concluding that X’s actual authority is automatically tainted by the dishonesty of Y in the above hypothetical. But the point of principle is potentially very important for other Quincecare cases, and to the extent that Lord Sumption appears to have endorsed that conclusion, Tugu might prove influential unless its weaknesses are addressed.

It is submitted here that there is no principle of the law of agency, nor public policy, that would lead to the conclusion that X, on whom P has directly conferred actual authority, automatically loses that authority because of the dishonesty of another agent, Y, merely on the basis that X has been instructed by P to follow, and has followed, the directions of Y. The position is likely to be different where X is appointed by Y, and is therefore a sub-agent of Y. None of the grounds for termination of actual authority listed in Bowstead & Reynolds on Agency (22nd ed) Article 117 support the assumption in Tugu. Modern dicta support the view that even the supervening mental incapacity of the principal does not automatically terminate an agent’s actual authority: see Blankley v Central Manchester University Hospitals [2015] EWCA Civ 18, [2015] 1 WLR 4307 [36].

This leaves the possibility that X’s mandate contains an implied term that X’s actual authority automatically falls with that of Y. It is submitted that such a term would be unreasonable, and to the extent that a contract exists between X and P would demonstrably fail the test for implied terms. Why would a bank, in particular, agree that its mandate should fall over merely because of the dishonesty of the mandatary?

It does not follow that no restriction should be implied into the mandate between X and P as to when X should not follow Y’s directions. Plainly, X should not do so when it knows that Y is being dishonest. X would itself normally be dishonest in such circumstances. A lower threshold for abstaining from following Y’s directions could be envisaged. Much will turn on the type of agent in question in that regard.

I have argued elsewhere that with a bank, the test should be actual knowledge or wilful blindness, given that banks are not expected, and indeed are generally not permitted, any discretion in whether to make the payments directed: see P Watts, ‘The Quincecare Duty: Misconceived an Misdelivered’ [2020] JBL 402; P Watts, ‘Playing the Quincecare Card’ (2022) 138 LQR 530. That might still leave room for something short of dishonesty, if, for example, some party in the bank knows all the facts but inexcusably (but honestly) fails to intervene to prevent the payments being made. It is certainly arguable that the bank in Tugu had actual knowledge that the company’s assets were being misappropriated.

A further possible implication is that X, including a bank, should not follow Y’s direction when X is going to benefit substantially from the payment being made and a reasonable party in the position of X would realise that Y is acting dishonestly. In such cases, X may be liable independently of its contract with P, in equity’s doctrine of knowing receipt. An example is Bank of New South Wales v Goulburn Valley Butter Co Proprietary Ltd [1902] AC 543 (PC), where a managing director transferred funds, by way of cheques, from the company’s bank account to his overdrawn private account at the same bank. In fact, the bank was found not liable because it was entitled to assume that the director was legitimately reimbursing himself in respect of debts owed him by the company and there was no one else of whom the bank could make inquiries. In the standard Quincecare case the bank will not have benefited from the payments.

What does follow from the foregoing argumentation is that in the usual Quincecare case, the concept of apparent authority, including its sub-rule as to being put on inquiry, is not relevant to the bank’s position. It follows also that cases such as East Asia Co Ltd v PT Satria Tirtatama Energindo [2019] UKPC 30, [2020] 2 All ER 294 and Thanakharn Kasikorn Thai Chamkat v Akai Holdings Ltd (No 2) (2010) 13 HKCFAR 479 are not pertinent to whether X should second-guess Y’s instructions. These cases cannot be deployed, as Lord Sumption did in Tugu, in order to rationalise the reasonableness test applied in Barclays Bank Plc v Quincecare Ltd [1992] 4 All ER 363 and those cases that have followed it. To reiterate, questions of apparent authority arise only when one does not have existing rights against the principal of the sort one needs. That was the case in both PT Satria and Akai Holdings.

Note that it is not being suggested here that issues of apparent authority arise only between principal and third party, and never between principal and agent. So, when an agent in obtaining initial authority deals with other agents of the principal, then issues of apparent authority can be relevant. Similarly, if the agent’s powers are being altered in some way through the exercise of the authority of other agents, then questions of apparent authority can be in play. Indeed, Lord Sumption was correct in Tugu (at [21]) to treat questions of actual and apparent authority as being relevant to the attempt in 1998 by the rogue mandataries formally to close the relevant bank account. The bank could not treat the purported closure of the account as effective to start limitation periods running unless those closing it had actual or apparent authority to do so. They did not.

Finally, it should also be observed that an honest bank’s contractual right to carry out the mandatary’s instructions is not determinative of the payee’s right to keep any payment as against the customer (see Reckitt v Barnett, Pembroke and Slater Ltd [1929] AC 176 (HL) 184–185). A bank’s payment contains no representation of entitlement to it.

Peter Watts is a Senior Research Fellow at Harris Manchester College, Oxford.