• The Supreme Court has reined in the law on restitution for indirect enrichment (used in cases as an alternative to equitable tracing).
  • The general rule is that there must be direct dealings between claimant and defendant.
  • The test for recognising exceptions used since Henderson J’s decision in ITC and the general causation test recently endorsed in Menelaou v Bank of Cyprus are no longer good currency.
  • Equitable subrogation appears to be the only acknowledged exception to the general rule.
  • Such apparent exceptions as exist apart from this are in reality dictated by the general law such as agency.

​Both ITC (see Lord Reed at [39]) and Lowick (see Lord Sumption at [22]) emphasised that ‘legal rights arising from unjust enrichment should be determined by rules of law which are ascertainable and consistently applied’. Restitution, lawyers were reminded, should not as a general rule be a matter of discretion (see Lord Goff in Lipkin Gorman v Karpnale [1991] 2 AC 548 at 578) but should be a predictable and certain part of the law of obligations.
The starting point is Lord Steyn’s oft-cited series of questions in Banque Financière de la Cité v Parc (Battersea) Ltd [1999] 1 AC 221 (“Banque Financière”) at p 227:

  1. Has the defendant been enriched?
  2. Was the enrichment at the claimant’s expense?
  3. Was the enrichment unjust?
  4. Are there any defences?

ITC was concerned solely with Lord Steyn’s second question. Lowick was concerned with both (ii) and (iii) or, more precisely, the interplay between those questions.

ITC is plainly the more significant of the two decisions. The ITCs were investment trusts that sought to reclaim refunds of mistakenly paid VAT. The ITCs had been charged a notional £100 VAT on the services provided by their investment managers (“the IMs”). The IMs in turn had passed on to the commissioners £75, being the net of VAT collected (“output tax”) and the VAT paid (“input tax”). Although the IMs had been able to recover some of the VAT for the ITCs the IM could not recover more than £75 actually paid by them and the statutory recovery mechanism imposed a limitation period. The ITCs sought the £25 shortfall from the commissioners on the grounds of unjust enrichment.

There was no direct transfer of value from the ITCs to the commissioners. The case therefore raised the question whether enrichment had to be direct. Henderson J held that the commissioners had been unjustly enriched for the full extent of the £100 but that the ITCs could not recover the lost years. The Court of Appeal held that the ITCs could recover in respect of the lost years but could not recover more than the £75 which the commissioners had received. Both Henderson J and the Court of Appeal considered that the “at the expense” requirement could be satisfied without the need for a direct transfer. The Supreme Court disagreed. In his careful speech Lord Reed explained why the balance needed to be restored.


Lord Steyn’s second question is of particular importance if the law of restitution is to remain true to its purpose. Restitution is concerned with reversal of “normatively defective transfers of value” (see Goff & Jones, The Law of Unjust Enrichment, 9th edn (“Goff & Jones”) [6-01]; Burrows, The Law of Restitution, 3rd edn, Oxford University Press, 2011, p 66; Burrows, A Restatement of the English Law of Unjust Enrichment p 45 and ITC [42]). This was, as Lord Reed observed, the restoration of the Aristotelean equilibrium (see ITC [42]), the implication appearing to be that the equation should be a simple one.

The clear and general rule, expressed by Morritt LJ in Kleinwort Benson Ltd v Birmingham City Council [1997] QB 380 at 400F, was that the words “at the expense of the claimant”, traced back to the US Restatement of the Law of Restitution (1937), ‘do no more than point to the requirement that the immediate source of the unjust enrichment must be the Plaintiff’. If that rule is not observed, slowly but surely other areas of the law (such as the principle of separate corporate personality and tracing) are subverted.

Unfortunately, perhaps because the remarks in Kleinwort Benson were obiter, it did not take long for this otherwise clear rule to be challenged. Professor Birks in Unjust Enrichment, Oxford (2005), chapter 4, at pp 89–98 suggested that not only should it be possible to have indirect enrichment but any causal link would do. If there was unfairness it could be cured by the defences. Professor Burrows’ Restatement of The Law of Restitution, 3rd edn (2011), suggested a general rule of direct enrichment subject to extensive exceptions. In Filby v Mortgage Express (No2) [2004] EWCA Civ. 759 May LJ appeared to recognise a test of “commercial or economic reality”.

Faced with this patchwork, Henderson J in ITC [2012] S.T.C. 1150 opted for a general rule heavily diluted by a broad approach for recognising exceptions. He could hardly have done much else at that point. His explanation of the circumstances in which exceptions should be recognised attracted widespread praise and was followed in successive appeals (see TFL Management Services v Lloyds Bank Plc [2014] 1 WLR 2006 (TFL), Relfo v Varsani (No 2) [2015] BCLC 14 (Relfo), Menelaou v Bank of Cyprus UK Ltd [2014] 1 WLR 854 and Investment Trust Companies v HMRC [2015] STC 1280).

The unintended consequence of Henderson J’s approach was that in every subsequent case the inquiry was liable to focus on building exceptions with vague concepts of “fairness” and “economic reality”, rather than following the primary rule. A spate of such cases seemed to follow. It may well be that this was precisely because restitution for indirect enrichment now seemed to be worth a try as an alternative to equitable tracing (see eg Relfo and ITC) or to reach dividends received by shareholders when it was not possible to pierce the corporate veil (see MacDonald Dickens & Macklin v Costello [2012] QB 244). TFL was the most extreme example. There appeared to be every prospect of a growing mountain of exceptions to the general rule.

When Menelaou [2016] AC 176 reached the Supreme Court, there was a clear opportunity to bring clarity to the analysis of Lord Steyn’s second question. Regrettably, the decision did not meet everybody’s expectations. Professor Virgo likened its arrival to ‘the opening of a much anticipated Christmas present and realising both that it is not what you wanted but also that the giver of the present never had any sensible idea as to what you wanted’ (see Restitution and Unjust Enrichment in the Supreme Court: Reflections on Bank of Cyprus UK Ltd v Menelaou, Cambridge, Feb. 2016). He described it as ‘arguably the worst decision in the history of the Supreme Court’. Goff & Jones also bemoaned the continued uncertainty after the Supreme Court decision in Menelaou (see Goff & Jones [6-02]).

If anything, the decision endorsed the loose causation reasoning that had gained favour in the Court of Appeal. It was enough for Lord Clarke (see Menelaou [27]) to ask ‘whether there is a sufficient causal connection in the sense of a sufficient nexus or link between the loss to the bank and the benefit received by the defendant’. As Lord Reed was to remark in ITC (see [37]), hiding behind phrases such as “sufficient causal nexus”, “link” or even “proximity” (which had also been used in Relfo and TFL) lurks the unanswered ‘critical question, namely, what connection or link is sufficient?’

Goff & Jones at least made it clear (see [6-02]) that they expected the Supreme Court in ITC to answer this question, an invitation which Lord Reed accepted, saying that although it was not wise to attempt to arrive at a definitive statement of principle, ‘in view of the uncertainty which has resulted from the use of vague and generalised language, this court has a responsibility to establish more precise criteria’ (see ITC [38]).


Lord Steyn’s formula looks deceptively simple until one remembers that the four questions do not amount to precise tests. They represent broad headings ‘intended to ensure a structured approach to the analysis of unjust enrichment’ (see ITC at [41]). When performing the detailed analysis the court should keep in mind the purpose of restitution, namely, to restore normatively defective transfers of value (see ITC [42] and Lowick [30]). The legal requirements embodied within the “at the expense of” question follow from this “principle of corrective justice” (see ITC [43]).

The analysis does not proceed as if, alongside this structured approach, the Court has been handed a clean slate. On the contrary, ‘there are centuries’ worth of relevant authorities whose value should not be underestimated’ (see ITC [40]) on which Lord Reed was able to rely. It was therefore unacceptable to set aside cases such as Ruabon Steamship v London Assurance [1906] AC 6, as Floyd LJ had done in TFL (see ITC [39]) on the basis that it ‘was not looking at the law through the eyes of the modern law of unjust enrichment’. In Ruabon Lord Halsbury rejected “a general principle of justice” whereby a man ought not to get an advantage unless he paid for it: that would imply, absurdly, that a man is to be compensated when he cuts down his wood and thereby gives his neighbour a better view.

Lord Reed expressly approved the general rule in Kleinwort Benson (albeit ostensibly only in the context of tax payments; see ITC [67]). Much else was implicitly jettisoned, although only TFL was expressly disapproved. The causation test used in TFL and Menelaou is no longer good currency. There is no encouragement to devise exceptions. The “economic reality” test in Filby is also discarded as it cannot be applied with rigour, certainty or consistency. Deciding where the economic burden of an unjust enrichment lies was virtually unworkable (see ITC [59]–[60]).

The “at the expense of” limb connotes a number of basic ideas, which emerge from the careful judgment in ITC. Certain of these will merit greater exposition:

  • First, the process of enriching must have involved a “transfer of value” to the defendant (see ITC [42–[43]).
  • Second, the transfer to the defendant must be of a “normatively defective” kind (see ITC [42]–[43]).
  • Third, the claimant must have suffered a “loss” or a diminution in wealth through the provision of that benefit [see ITC [43]–[45]).
  • Fourth, the general rule is that there must be direct dealings between claimant and defendant: where the benefit has not been received “directly” from the claimant it will be “difficult” to demonstrate that one has been enriched by the other (see ITC [51]).
  • Fifth, this rule is subject to limited “exceptions” primarily when the general law is otherwise prepared to treat certain transactions as “equivalent” to a direct transfer (see ITC [52]). Rather unhelpfully Lord Reed suggested that the door was never absolutely closed.


Lord Reed’s general rule means, as the Court of Appeal held in MacDonald Dickens, that it will not be possible to say that restitution should be claimed of dividends paid to a shareholder by a company that received a mistaken payment. Similarly, restitution should not therefore reach the beneficiary of a trust whose distribution was funded by a mistaken payment to trustees (see ITC [51]). What is much harder to know is whether there are any genuine exceptions at all, apart from equitable subrogation (see below).

Even when parties have not in fact dealt with each other directly the case may nevertheless be treated as one of direct enrichment. This occurs when, on closer examination, ‘the difference from the direct provision of benefit to the defendant is more apparent than real’ (see ITC [47]). Examples given were interposition of agents or where the claimant is an assignee of rights. The law of agency ensures that the principal is treated as the true party in interest. An assignee stands in the shoes of the assignor. These are legal common law fictions in which the law imposes a direct enforceable relationship between parties for purposes other than restitution.

There are also other cases where enrichment was “equivalent to a direct payment”, notwithstanding the interposition of a third party. By way of example, Lord Reed singled out the arrangement in Relfo which had been found to be a sham for the precise purpose of concealing the defendant’s involvement (see ITC [48]). ‘As a matter of substance, or economic reality, [the defendant] was a direct recipient’ (see Relfo [99]) because ‘the intermediate arrangements [had been] an elaborate façade to conceal what was in truth intended and arranged to be a payment for the benefit of [the defendant]’ (see Relfo [121]). Similarly, a court might look behind the corporate veil to identify the real recipient when companies have been interposed to conceal their identity or been established to evade liability (Lord Sumption’s “concealment” or “evasion principle” in Petrodel Resources v Prest [2013] 2 AC 415 p 484 [28]).

There is also a well-established tradition of placing substance over form and disregarding transactional links, for example in the context of fiscal planning. Lord Reed referred to the disregard of links in a chain of a pre-ordained or coordinated series of transactions in Banque Financière (see ITC [61ff]). In order to circumvent disclosure obligations in relation to a proposed loan, the bank lent money first to its general manager (Mr Herzig) who in turn lent the money to another company which in turn used it to discharge a loan from RTB, another bank. The House of Lords held that the interposition of Mr Herzig was purely a formal act and did not ‘prevent recognition of the reality’ (see Banque Financière, p 238–39). Similarly, in Kleinwort Benson Saville LJ suggested that the interposition of the immediate recipient might be disregarded if that person was acting as a “mere conduit” (see p 395).

These are all examples of techniques in the general law whereby a party or step in a chain of transactions is disregarded usually because either for good reason (agency) or bad (concealment or evasion) that party is not intended to play any personal role. These are not therefore true “exceptions” to the general rule driven only by the law of restitution.


There are indications that Lord Reed might have considered transfers of traceable property as an “exception” to the general rule (see ITC [51] and [70]). If so, it would not be entirely easy to see why. Where a defendant receives the traceable proceeds of property, the defendant will be subject to proprietary remedies for the recovery of the property. Those consequences follow not from the law of restitution but from equitable rules relating to personal property (see Foskett v McKeown [2001] 1 AC 102). It is not immediately obvious why unjust enrichment is necessary against a recipient of traceable property. In Relfo the claim for recovery of traceable property was simply an alternative to unjust enrichment if tracing did not work.

In contrast, the non-availability of tracing is an additional reason for denying indirect restitution at least against innocent parties. If unjust enrichment could supply a remedy against a third party whenever the law of property would refuse one, the strictures of tracing could simply be ignored. Why would it then be necessary, for example, for the claimant to establish a proprietary base at all? When Lord Reed refers to the absence of any right to trace funds he more probably had this negative consideration in mind as another basis for refusing to recognise exceptions. Indeed this certainly seems to have been a reason for rejecting the claims of the ITCs (see ITC [70]).


Even in a two party case, the requirement for loss will not be satisfied if the benefit to the claimant was merely an “incidental or collateral” result of the claimant’s expenditure. The claimant may well have received what he bargained for as the counterpart of any expenditure (see ITC [52]). Examples of collateral benefits can be found in the case law (referred to in ITC [53]–[58]):

  • The classic examples of incidental benefit, given by Lord Dunedin in Edinburgh and District Tramways Co Ltd v Courtenay (1909) SC 99 at 105, are the man who by felling the trees improves his neighbour’s house and Lord Halsbury’s house-owner whose coal fire heats his neighbour’s house (see Ruabon p 12).
  • In Ruabon itself a damaged ship was in a dry dock for repairs at the insurers’ expense. While it was being repaired the owners took advantage of its presence by having it surveyed. The owners were not liable to the insurers for the port fees because there was no increase in cost to the insurer who suffered no loss and obtained what it bargained for.
  • In TFL “A” sought restitution from “B” to recover A’s legal costs incurred in unsuccessful proceedings against “C”. A had lost those proceedings because C was held liable to B, not A. B had been enriched by having its rights clarified. Summary judgment was refused because the causal link of A’s mistake and B’s benefit was arguably sufficient. This was wrong. There had been no direct transfer from A to B. B’s benefit was merely incidental to A pursuing its own objective. Nor was this unjust: A had received the legal services it had bargained for.

When is a benefit merely incidental or collateral? Usually, the claimant will have been pursuing his own objectives and the defendant’s enrichment will be an accidental consequence of doing so. In neither of the examples given by Lords Halsbury or Dunedin is that because of the direct transfer rule. It might well be more convincing to answer this question not by reference to the “at the expense” issue but by looking at the “unjust” limb of Lord Steyn’s formula (see ITC [52]).

Lowick was also a case of incidental benefit. Holdco purchased shares in EMSL, based on HMT’s negligent accounting advice about its working capital. As a result Holdco had to lend money to EMSL, which was, as reasonable mitigation, also a potential head of damages for Holdco. Two years later Holdco was restructured by its sole shareholder (“H”). He lent money to EMSL which in turn discharged its liability to Holdco. H thereby cleaned up Holdco’s balance sheet and received Holdco’s security over EMSL as well as tax advantages. Without realising it, Holdco had recovered its “loss” from EMSL and HMT’s tort liability to Holdco had thereby been reduced, at H’s expense.

Had HMT been unjustly enriched? The analysis is more complex as this was a claim for equitable subrogation. Lords Mance, Neuberger and Clarke acknowledged that they had struggled to disentangle the “at the expense” question from the “unjust” limb (see Lowick [68] and [117]):

  • The majority view, provided by Lord Sumption, was that HMT’s enrichment was not unjust. There was no defeated expectation that H would obtain rights from the transaction: H got what he bargained for – a restructuring of Holdco and security. H had simply miscalculated the consequences of the transaction (see [32]–[34]).
  • Lord Mance also concluded that the “unjust” limb was not satisfied but did so on a slightly different basis (see [68]–[69]). H’s mistake was not the relevant mistake. The arrangements he made did not concern the relationship between Holdco and HMT but between EMSL and Holdco (see [86]–[87]).
  • Lords Neuberger and Clarke held that H had not shown that the enrichment of HMT was at H’s expense. H had not discharged HMT’s liability to Holdco. H had loaned funds to EMSL. It was EMSL that paid Holdco. There was no direct enrichment. According to the general rule the claim had to be dismissed (see [115] and [119]). What is telling is that the court followed the rule reminding itself that it had only just in ITC ruled out the idea that it would fashion exemptions.


An indirect transfer to a third party is not a bar to restitution in cases where the claimant discharges the debt owed by the defendant to a third party (see Burston Finance Ltd v Speirway Ltd [1974] 1 WLR 1648). In such cases the law reverses or prevents the enrichment by affording the remedy of equitable subrogation (see Banque Financière, p 231). Both Menelaou and Lowick were concerned with equitable subrogation.

Equitable subrogation normally results in the claimant receiving the benefit of some form of security (eg the unpaid vendor’s lien in Menelaou) but it can also allow the claimant to revive personal rights against the defendant. It operates on a fictionalised basis: rights are not in fact “kept alive” when a debt has been discharged; instead the rights between the parties are regulated as if the security or right had been assigned (see Banque Financière, p 236 and Lowick [26]).

This is a true exception to Lord Reed’s general rule. Although the third party receives a benefit, the enrichment is not conferred directly. Because the direct enrichment rule does not apply, the only possible safety value is the “unjust” factor. Lowick demonstrates how much more work has to be done by this limb when the direct enrichment rule is suspended.

The unjust element in a case of equitable subordination is usually expressed as being based on the claimant’s mistake about the transaction. The court was clearly troubled by the idea that a claimant could use equitable subordination as a “general escape route” whenever the consequences of a transaction had been misunderstood (see Lowick [34]–[35], [78]–[83] and [120]). Unfortunately, revisiting what amounts to mistake for the purposes of restitution is not an option. Mistake here is already broader than in other areas of law and includes mistake of law and unilateral mistake (see Lowick [28] to [30]).

The answer in Lowick was to identify a different “unjust” factor by looking at what equitable subordination sets out to achieve. Instead of restoring the pre-transfer position (see ITC [49]), restitution here protects a loss of bargain by ‘[replicating] as far as possible that element of the transaction whose absence made it defective’ (see Lowick [31]). The remedy specifically enforces a defeated expectation, normally the loss of a right of security (see Lowick [30]). The majority held that the claimant’s defeated intention must relate to the relevant defect (see Lowick [30] and also Menelaou [21]). Tailored to this remedy, the unjust element can be articulated as the “absence or failure of basis for the transfer”.

The only difficulty about this sleight of hand is that it is for the claimant to decide whether to rely on a mistake. If he is not to be allowed to do so that must be because of what is required to be shown for a valid restitutionary mistake. That in turn would involve rewriting a great deal of case-law.


A general rule and the balance in restitution has been restored. Outlandish claims of unjust enrichment circumventing principles of separate corporate personality or tracing will no longer work in the light of ITC. It is unclear to what extent there will be any scope for departing from this general rule. Apart from those “exceptions” canvassed by Lord Reed, no others were acknowledged and no basis for recognising new exceptions was suggested.

Thomas Lowe QC is a barrister practising from Wilberforce Chambers in England and independently in the Cayman Islands. Email: tlowe@wilberforce.co.uk and tlowe@wilberforce-cay.com