Vadim Don Benyatov v Credit Suisse (Securities) Europe Ltd

The duty of care relied on by the Claimant is pleaded in para. 24 of his Amended Particulars of Claim (“the AMPOC”), which begins:

“… [T]he Defendant owed Mr Benyatov a duty of care not to expose him to criminal conviction in the performance of his duties for the Defendant. In particular, the Defendant had a continuing duty:

24.1.

Generally to take reasonable care to protect him from criminal conviction and the resulting losses by performing the duties enumerated in paragraphs 24.2 to 24.6 below;

…”

The particular duties pleaded at paras. 24.2-24.6 are very elaborately formulated and I need not reproduce them in full. Their essence is that the Bank was obliged to carry out an assessment of the risks to the Claimant of working on projects in Romania generally, but more particularly on projects which might place him “at a material risk of criminal conviction” or “otherwise expose [him] to criminal conviction in Romania” in consequence of properly carrying out his contractual duties; and that it should not have required him to carry out such work, or to visit Romania, without communicating to him the results of such a risk assessment and giving him any necessary training or advice in order to enable him to mitigate any risks identified. I will refer to these as “the risk assessment duties”.

It will be noted that the risk against which it is said that the Bank had a duty to take reasonable care to protect the Claimant is conviction tout court, rather than wrongful conviction. That is evidently deliberate. One of the risks which it is said that the Bank should have assessed is “the risk of Mr Benyatov being convicted for conducting business activities that are in accordance with international banking standards but that could be construed as unlawful as a matter of Romanian law”. I take that as a plea that there was a duty to assess the risk that Romanian law might idiosyncratically proscribe conduct which was elsewhere regarded as legitimate and which the Claimant would accordingly reasonably expect to be lawful unless warned otherwise. Mr Ciumei did, however, accept that there could be no duty to protect an employee against being convicted of an offence committed knowingly.

At paras. 25-29 of the AMPOC the Claimant pleads the facts and matters relied on as giving rise to both the implication of the contractual terms on which he relies and the duties of care pleaded at para. 24. For present purposes I need refer only to para. 25, which begins:

“Facts and matters giving rise to the implication of the terms referred to in paragraphs 19.1, 19.2 and 19.3 as a matter of fact and/or business efficacy, and giving rise to the duties at paragraph 24 above, include the following: …”

23 sub-paragraphs then follow, mostly setting out factual matters which are said to show that there was a foreseeable risk, against which he could reasonably expect to be protected, of the Claimant being the subject of criminal prosecution and therefore conviction. I need not reproduce them here. The substance of the case appears sufficiently from para. 25.22, which reads:

“By sending Mr Benyatov to emerging markets and/or to Romania, which were high risk in relation to political risk, deep issues of corruption and a poorly functioning judiciary, with a concomitant risk of political exposure in dealmaking, particularly in respect of privatisation, and in circumstances where individuals exploited the internal security and police forces and/or the legal system and processes and/or the judiciary to target competitors and political/business adversaries, with the risk that Mr Benyatov might be caught up in such matters, including the risks of being the subject of criminal investigation, criminal charges, prosecution, conviction, arbitrary detention, and/or imprisonment as a consequence of working and acting on behalf of the Defendant in Romania, the Defendant had such duties of care as pleaded at paragraph 24.”

In short, the Claimant’s work in Romania was “high-risk”.

Para. 28 addresses the question of the Bank’s knowledge of the risks which the Claimant says it should have assessed. It begins by averring that “[t]he risks set out in paragraph 24 were and/or should have been known to the Defendant because, inter alia, they were widely documented in the public domain at all material times or available to any well-resourced and competent banking institution operating globally”. It will be seen that the Claimant pleads that the risks either were known to the Bank (“actual knowledge”) or should have been known (“constructive knowledge”). It should be noted that particulars of knowledge of some particular risks are also pleaded in connection with para. 52 of the AMPOC, which I summarise next.

Para. 52 of the AMPOC pleads, under eight sub-paragraphs, the alleged breaches of the duties of care. The essence of the pleaded case is that the Bank did not carry out any risk assessment of the kind which was alleged in para. 24 to be necessary.

Para. 52.2 pleads a number of specific matters which it is said showed that “foreseeable risks were materialising” and/or which constituted “material developments”. These were referred to in the submissions before the Judge as “amber or red flags” which should have alerted the Bank to the need to take steps to “protect [the Claimant] from criminal conviction and the resulting losses” (see para. 52.8). Para. 52.2.4 reads:

“Adverse press coverage of the Petrom privatisation from May 2006 and the interest of the Romanian secret police in that transaction [emphasis supplied], which it is reasonably to be inferred (the Defendant’s disclosure on this point being inadequate to date) the Defendant knew about given that it was the subject of the adverse coverage and it was involved in the Petrom transaction and/or which should have been known to the Defendant if it had conducted adequate regular assessments of risk.”

The Petrom privatisation was an earlier privatisation in which the Bank and the Claimant had been involved. As will appear, the allegation that “the Romanian secret police” – a reference to Serviciul Român de Informații (“the SRI”), which is the Romanian state intelligence service – were interested in the Petrom privatisation is relevant to two of the grounds of appeal. It will be seen that the Claimant avers actual knowledge (based on inference) “and/or” constructive knowledge.

I need to say something more about the history of the pleading of para. 52.2.4. It was introduced by an amendment permitted by Mr ter Haar in the context of the Bank’s strike-out application. The history is summarised by Freedman J at paras. 196-197 of his judgment as follows:

“196.

… The Bank objected to [the proposed amendment] in that it did not state whether it was alleged that the Bank knew (in this case about the alleged interest of the secret police) and, if so, who, when and how they knew. The Deputy Judge held on 25 November 2020 that these were legitimate questions, and it was a condition of permission that these particulars were given: see his judgment at para. 192.

197.

When the matter next came before the Deputy Judge, particulars had not been provided, but it was stated on behalf of Mr Benyatov that the amendments provide the best particulars of knowledge available based on the inadequate disclosure provided by the Bank. The Deputy Judge gave permission at para. 28 of his judgment dated 8 December 2020 in the following terms, namely:

‘I have no way of assessing whether the disclosure given thus far has been inadequate, but I accept Mr Ciumei’s assurance that these are the best particulars which the Claimant can presently give. I give permission to amend the pleading as requested, subject, however, to an important caveat: I do so on the express understanding (which seems to me clear from the pleading) that at present there is no case of actual knowledge made against the Defendant. If proper grounds for such a serious allegation were to emerge, I would expect a further application to be made to amend to make such allegation expressly. Any such application would of course fall to be considered by the Court on its merits when made.’”

No application of the kind identified by Mr ter Haar was made prior to the trial.

We are not on this appeal concerned with any issues about remedy or quantum, but I should briefly identify how the Claimant’s case was put. Paras. 53-55 are headed “Indemnity/Loss and Damage”. Para. 53 begins:

“In consequence of the foregoing breaches of contract and/or duties of care, Mr Benyatov has suffered and continues to suffer loss and damage, in respect of which he is entitled to indemnification and/or damages, including the following”.

The loss and damage in question is pleaded as:

The prayer is for, so far as relevant:

“(1)

An indemnity and/or damages as pleaded at paragraphs 53 to 55 ...;

(2)

Declarations as pleaded at paragraph 56;

(3)-(5) …”

The Bank pleads to para. 24 of the AMPOC – the alleged duties of care – at para. 18 of its Amended Defence. This reads:

“As to paragraph 24:

18.1.

As a matter of law, an employer has no general or implied duty to protect the economic interests of its employees and, for the avoidance of doubt, at no time did the Defendant assume any such duty.

18.2.

It is admitted that, whilst he was its employee, the Defendant owed the Claimant a tortious duty co-extensive with its duty under the Claimant’s contract of employment, to take reasonable care not to expose the Claimant to risks of personal and psychiatric injury that were reasonably foreseeable and which the Defendant could guard against by proportionate measures.

18.3.

Paragraph 24 is otherwise denied.

18.4.

More particularly, the duties alleged in paragraph 24 are inconsistent with the term of the Contract pleaded at paragraph 13.2.1 above. It was the Claimant’s duty and responsibility to know and comply with the law in any place in which he worked for the Defendant.

18.5.

Moreover, the duties alleged are contrary to decided and/or analogous authority, alternatively have not been recognised in any established authority; recognition of such duties would not amount to an incremental step from previously recognised duties; and the imposition of such duties would not be fair, just or reasonable.”

(I have included para. 18.4 for completeness: for the purpose of the issues before us I need not elucidate the references in it.)

As regards para. 25 of the AMPOC, para. 19 of the Amended Defence begins with a general denial that the matters relied on give rise to the duties alleged. The 23 sub-paragraphs are then individually pleaded to: I need not summarise what the Bank says. Para. 22 denies para. 28 of the AMPOC in its entirety.

The Bank also pleaded a limitation defence. Para. 63 of the Amended Defence, which was added by amendment, reads (so far as material):

“If (which is denied) the Defendant breached any alleged duty of care it owed to the Claimant, and such breach caused recoverable loss to the Claimant, such a claim is time-barred by virtue of section 2 of the Limitation Act 1980:

63.1.

Insofar as the Claimant relies on a duty of care to prevent financial loss, such loss was suffered by the Claimant, and his cause of action in negligence accordingly accrued:

63.1.1when he was arrested in November 2006, alternatively when he was detained in Romania between November 2006 until August 2007, alternatively when he was charged in September 2007, and thereby suffered loss of earning capacity;

63.1.2

alternatively, when his compensation decreased in 2007 and/or in 2008 as a result of his arrest, detention, charge and/or prosecution.

63.2

…

63.3

The Claimant did not issue his claim until 22 January 2018, more than 6 years after his cause of action first accrued.”

The Bank raised other issues by way of defence but I need not refer to them here. I should, however, mention that in certain parts of the pleading raised by amendment it sought, inconsistently with its stance up to that point, to rely on the Claimant’s conviction by the Romanian court as establishing, or in any event as constituting evidence, that he was in fact guilty of criminal conduct. The Judge rejected that part of the Bank’s submissions, and I need say no more about them.

The Judge analyses the nature of the Claimant’s case on the existence of a duty of care in Part V of his judgment. After setting out at paras. 104-105 the case as pleaded, he continues:

“106.

Mr Benyatov in this case does not allege that the Bank owed him a general duty to protect him from economic loss. The duty of care is significantly narrower than that. It is based on, among others, the specific nature of the work that he did; the country in which he was required to work; the foreseeability of the harm that could befall him, including in the form of specific amber and red flags that allegedly came to the Bank’s attention; and the gravity of the consequences if the risks materialised.

107.

It is worth noting that the same facts and matters said to give rise to a duty of care are also relied upon in the AMPOC to give rise to the implication of the claim for contractual indemnities. In the claim as pleaded, both as regards the duty to exercise reasonable care and to indemnify for losses, there is a common factual substratum to justify both the scope of the duty to take care and the scope of the duty to indemnify.”

108.

It is therefore a fact intensive duty of care and duty to indemnify…”

In Part VI the Judge reviews in detail the case-law as to the existence of a duty of care of the kind relied on. At paras. 112-118 he identifies his basic approach as being that the present case involved a “novel situation” – that is, one where the existence of a duty of care had not been recognised in previous authority – with the result that he should proceed “incrementally and by analogy with established authority, identifying the legally significant features of the situations”. That approach derives from the decision of the House of Lords in Caparo Industries Plc v Dickman [1990] 2 AC 605, as explained in Robinson v Chief Constable of West Yorkshire Police [2018] UKSC 4, [2018] AC 736, which corrected some misunderstandings in the case-law about the effect of Caparo: see para. 46 below. At paras. 119-147 he considers the circumstances in which a duty may exist to protect against economic loss, noting that the law is generally reluctant to impose such a duty but that it has nevertheless been recognised in some circumstances in the context of the employment relationship. In that connection he discusses the decision of Kerr J in Rihan v Ernst & Young Global Ltd [2020] EWHC 901 (QB), to which also I will have to return: see paras. 57-61 below. At para. 148 he sets out various general propositions which I need not summarise.

At para. 149 the Judge identifies the approach which he believes is correct in the present case. After rejecting a submission by the Bank that the claim was inadmissible because it was purely for economic loss, he concludes, at paras. 154-156:

“154.

In a novel case, the question as to whether there is a duty of care is not to be decided subject to these general parameters. It is very fact sensitive in each case. That is recognised by the way in which the matter has been pleaded. Mr Benyatov’s legal team is not saying that there is a blanket duty in each case for an employer to be procuring risk assessments in respect of its employees. It is brought about by features in this case. They include that it is a large multi-national bank which sends its employees around the world to countries of high-risk and/or in connection with high-risk transactions which take place and/or where there are other high-risk features affecting their employees.

155.

There may be cases where there is in principle a duty of care to take reasonable care for the safety of the employee and the financial interests which go with that. In each case, it will depend upon a consideration of the precise facts, the foreseeability of the particular harm and whether in all the circumstances, it is fair, just and reasonable to impose a duty.

156.

It is for this reason that at the heart of the case is para. 25 of AMPOC in which Mr Benyatov set out the facts and matters giving rise to the duty of care not to expose Mr Benyatov to criminal conviction in the performance of his duties and the resulting losses. This was said to include the facts and matters identified in the 23 sub-paragraphs of para. 25. This is indeed fact sensitive, and it is to the facts that this judgment must now turn.”

In Part VII, which comprises paras. 157-227, the Judge carries out the detailed factual assessment which he has thus found to be necessary. This includes, at paras. 181-206, a careful consideration of each of the alleged warning flags, followed by his conclusions about them at paras. 207-216. In his submissions Mr Ciumei had identified seven such flags, of which the fourth to seventh were concerned with the problems about the Petrom privatisation as pleaded at para. 52.2.4: the fourth was the adverse press comment about the privatisation and the fifth to the seventh concerned alleged manifestations of the SRI’s interest in it. The Claimant’s case that the SRI’s interest was known to the Bank was based on notes of interviews with two of its employees – Mr Raj Rao, a former Director of Credit Suisse, and Mr Nathan Burkey, a Managing Director in its Fixed Income Division (“FID”), as part of the Bank’s internal investigation following the Claimant’s arrest: Mr Burkey was interviewed once, on 24 November 2006 and Mr Rao twice, on 24 and 29 November 2006. The notes were referred to at trial as “the Rao memorandum” and “the Burkey memorandum”. In view of the concerns expressed by Mr Ciumei to the Deputy Judge about the adequacy of the Bank’s disclosure at the time of the amendment application (see para. 16 above), I should make it clear that the memoranda were disclosed in 2019, long before that application was made.

I will have to say more later about some aspects of the Judge’s reasoning in Part VII, but at this stage I need only quote his overall conclusion on whether a duty of care arose (omitting one immaterial point):

“224.

The effect is that key features relied on by Mr Benyatov to establish the duty of care have not been established. … .

225.

I am not satisfied that any of the evidence gives rise to the alleged duty of care. Having reviewed the entirety of para. [25] of the AMPOC, Mr Benyatov has been unable to prove the basis of the pleaded duty of care. He has failed to prove that in the circumstances of this case it was reasonably foreseeable that Mr Benyatov would be exposed to a conviction in the performance of his duties for the Bank.

226.

Among the matters which have been taken into account in reaching the conclusion that the particular duty of care has not been established are the following factors, namely:

(1)

The facts showing that Romania was not regarded as a high-risk country during the Relevant Period;

(2)

The facts showing that the EMS transaction was not regarded as a high-risk transaction;

(3)

The fact that none of the alleged amber or red flags have been established as putting the Bank on notice of some special need for vigilance in the instant case, nor in respect of various of the alleged flags were they escalated by Mr Benyatov to the Bank on the basis that they were treated as worthy of consideration at a higher level;

(4)

The fact that, contrary to Mr Benyatov’s pleaded case, there was no standard practice in respect of commissioning a political or other detailed risk assessment whether as alleged by Mr Benyatov or at all. Indeed, the evidence is that international companies did not commission such an assessment;

(5)

There were no circumstances applying specifically to Mr Benyatov including his name, his origin and his experience which made it inappropriate for him to be appointed on the EMS transaction.

227.

The case as to the duty of care therefore fails by reference to the facts of the case. If it had been a case with radically different facts, there would arise for consideration whether such a novel duty of care could be found. These questions do not arise in a factual vacuum. In the instant case, the claim as to the particular duty of care must fail. For the purpose of completeness, I shall consider additional arguments of the Bank to seek to negate the duty of care. They do not arise for necessary consideration in view of the conclusion above.”

The final section under Part VIII of the judgment (which otherwise deals with additional arguments raised by the Bank which I need not consider here) is headed “Conclusion on duty of care”. It consists of a single paragraph, as follows:

“245.

It follows that major building blocks on which the duty of care is said to exist are not established. The case that Romania was or was perceived to be a ‘high-risk’ country fails. There also fails the case that the EMS transaction or the other privatisations in which Mr Benyatov was involved were ‘high-risk’ transactions or were perceived as such. Nor have any of the warning flags been established. Nor has it been shown that there were any factors specific to Mr Benyatov which needed to be brought into the equation. Absent all of this and having considered the evidence as a whole, in my judgment the conviction and the subsequent losses were not reasonably foreseeable. Further, in circumstances where there was so much business activity in Romania (in the run up to the imminent entry of Romania into the EU) and against the background of successful privatisation work undertaken by the Bank involving Mr Benyatov in the early years of the decade, it was not fair, just and reasonable to create the alleged or a related duty of care in the circumstances of this case. It follows that Mr Benyatov’s case on the duty of care must fail.”

Part IX of the judgment, which comprises paras. 246-301, is headed “Breach of duty of care”. The Judge’s conclusion that the Bank did not owe the Claimant the duty alleged meant that there was in fact no need to consider the issue of breach. However, as he observed at para. 246:

“… [T]he concepts of whether there is (a) a duty of care, and (b) a breach of duty in the circumstances of this case, are intimately connected. The effect may often be that the very points which would negative a duty of care may be the answers to the allegations that there was negligence and/or a breach of duty.”

The overlap is illustrated by the fact that the Claimant’s case on the alleged warning flags is pleaded in the part of the AMPOC dealing with breach (para. 52) rather than the part dealing with the existence of the duty (para. 25).

The Judge states his conclusion at the start of Part IX, at para. 248, where he says:

“I am satisfied, on the basis of the information reasonably available to the Bank at the time of the alleged negligence, that there was no reasonable probability or possibility that Mr Benyatov would be arrested and the subject of a criminal conviction in Romania as a result of working in Romania and/or by reason of the work undertaken in Romania up to 2005-2006.”

He gives his detailed reasons in the following paragraphs, which I need not summarise.

In Part X of the judgment the Judge upholds the Bank’s limitation case, as pleaded at para. 63.1.2 of the Amended Defence. I will return to his reasoning later. Strictly, his conclusion on this issue rendered his earlier conclusions in Parts VII-IX redundant, but he was clearly wise not to decide the negligence case on this basis alone, and there is in any event an overlap between those conclusions and the contractual indemnity claim: see paras. 147-150 below.

The Claimant challenges the dismissal of his negligence claim on three grounds. The first alleges that the Judge took the wrong approach in law to the question whether the alleged duties of care arose. The second challenges his decision on the facts relied on as establishing a duty of care, alleging that he failed to take into account relevant evidence and/or that he reached a perverse conclusion on the basis of the evidence adduced: in both respects the focus is on his finding about the alleged warning flags. The third challenges his decision on the limitation issue.

This ground is formulated as follows (omitting some immaterial references):

“The Judge erred in his approach in law to determining whether there was a duty of care in tort … in that he focussed his analysis on whether there was, in the subjective understanding of the defendant Bank, a reasonable foreseeability of Mr Benyatov being exposed to criminal conviction. This is the wrong test. The Supreme Court has repeatedly emphasised that the correct approach in law requires an analysis of whether in this type of situation the Bank had an implied assumption of responsibility, objectively determined, to its employee, the claimant Mr Benyatov, in respect of the said risk of harm. It is through that lens that the Court should have analysed whether it was appropriate to develop the law incrementally, and in this regard the Court erred … in failing to draw an analogy from the audit duty found in Rihan v Ernst & Young Global Ltd [2020] EWHC 901 (QB).”

There are three strands in that ground, though there may be some overlap between them.

Reliance on the Bank’s “Subjective Understanding”

First, it is said that the Judge adopted a subjective rather than an objective approach to the foreseeability of the risk to the Claimant – that is, he did not consider what the Bank should have foreseen but only what it actually foresaw. If that was indeed his approach, I agree that it would have been wrong, but I do not believe that it was. My reasons are as follows.

I start with the Judge’s self-direction as to the applicable law. The first of the propositions set out at para. 148 of his judgment reads (so far as material):

“The Court must consider the information reasonably available to the Bank at the time of the alleged negligence.”

The reference to the information “reasonably available” to the Bank plainly, as I understood Mr Ciumei to accept, recognises that the standard is objective – that is, it refers not only to what the Bank actually knew but to what it should have known.

That formal self-direction is consistent with how the Judge defines the issues at several points elsewhere in the judgment. By way of example only:

Mr Ciumei submitted that although those passages identified the correct approach the Judge did not in fact follow it. That submission is not developed with any particularity in his skeleton argument: mention is made in para. 27 of paras. 226-227 of the judgment proceeding by reference to “the Bank’s subjective appreciation of the risk”, but I cannot see that those paragraphs (which I have quoted at para. 28 above) support that contention. However, in his oral submissions he relied on the Judge’s treatment of the fifth of the warning flags, being one of those relating to the alleged interest of the SRI in the Petrom privatisation. The Judge set out the Claimant’s case on this aspect at paras. 196-201 of the judgment and stated his conclusion at para. 213, where he said (for reasons which he gave), that there was “no suggestion of what the Bank knew or that they knew that the Romanian secret police was involved”. That, Mr Ciumei submitted, is a finding only about what the Bank actually knew and does not address what it ought to have known.

I do not believe that that is a valid criticism. At para. 198 of the judgment the Judge characterises the Claimant’s case at the trial as being “that the Bank had actual knowledge of the alleged interest of the secret police in the Petrom transaction”. That case was advanced by reference to the contents of (so far as this flag is concerned) the Rao memorandum. It is clear that that – i.e. actual knowledge – is the only basis on which the Claimant put his case on this flag. That is not surprising: the pleading in the closing lines of para. 52.2.4 that the Bank ought to have known of the SRI’s interest as a result of “regular assessments of risk” is, to say the least, flimsy – as well as begging the question whether risk assessments of the relevant kind were required in any event.

For those reasons I do not believe that this strand of ground 1 is made out.

Assumption of Responsibility

The case advanced in the pleaded ground, and developed in Mr Ciumei’s skeleton argument and oral submissions, is that the Judge’s approach to the issue of whether a duty of care arose in the present case was fundamentally flawed because it did not approach it “through the lens” of the concept of assumption of responsibility. That concept was of course first recognised in Hedley Byrne & Co Ltd v Heller & Partners Ltd [1964] AC 465 in the context of negligent misrepresentation, but it has since been applied in other contexts – most notably the negligent provision of services (see Henderson v Merrett Syndicates Ltd [1995] 2 AC 145). Mr Ciumei’s submission appears to be that it should be treated as the fundamental principle underlying the recognition of a duty of care in tort, or in any event as a necessary part of the analysis in every case. Although as pleaded in ground 1 this submission appears to be related to the “subjective approach” point considered above, it seems to me quite distinct. (Footnote: ¹)

Mr Goulding submitted that it was not open to the Claimant to challenge the Judge’s reasoning on that basis. He pointed out that a case based on assumption of responsibility had been advanced in a response to a Request for Further Information of the original Particulars of Claim but had not been carried over into the eventual AMPOC, which was supposed to incorporate any averments contained in earlier pleadings. Perhaps more significantly, Mr Ciumei had not at any stage during the trial relied on the concept of assumption of responsibility and in his closing submissions had advanced his case on the basis of precisely the approach which the Judge adopted but which he now says is flawed.

Mr Ciumei did not dispute Mr Goulding’s summary of how the case had been pleaded or argued below, but he submitted that since the point was purely one of law he should not be prevented from advancing it in this Court. I am far from sure that he is entitled to do so, but I need not decide the question since in my view the challenge is in any event ill-founded. It is well established that in determining whether a duty of care should be recognised in a novel situation the correct overall approach is that most recently expounded in Robinson. In applying that approach the concept of assumption of responsibility may, depending on the particular circumstances, be a useful tool; but it is not apt or useful in every kind of case. I would analyse the relevant case-law as follows.

It is unnecessary to consider Caparo or Hedley Byrne themselves. The issue raised by this ground is not about the principles established by those cases but rather about how they relate to one another in a novel situation like the present. It is enough to record that Caparo introduced the well-known categorisation of the factors relevant to the recognition of a duty of care in a novel situation as foreseeability, proximity, and fairness, justice and reasonableness. These have sometimes been referred to as “the tripartite test”, but the Supreme Court in Robinson has now made it clear that that description is liable to lead to misunderstanding.

The starting-point for considering how the concept of assumption of responsibility relates to the approach in Caparo must be the decision of the House of Lords in Commissioners of Customs and Excise v Barclays Bank plc [2006] UKHL 28, [2007] 1 AC 181. The ultimate issue raised by the appeal was defined by Lord Bingham as being “whether a bank, notified by a third party of a freezing injunction granted to the third party against one of the bank’s customers, affecting an account held by the customer with the bank, owes a duty to the third party to take reasonable care to comply with the terms of the injunction”: see para. 1 of his opinion. The House held unanimously that it did not. All five members of the Appellate Committee delivered substantive opinions, and their approaches to the definition of the correct test for the recognition of a duty of care are not identical. However, what matters for our purposes is that none of them is consistent with the recognition of assumption of responsibility as the fundamental underlying concept. It is sufficient to refer to four passages:

I turn to Robinson v Chief Constable of West Yorkshire Police, which is the authority by reference to which the Judge principally directed himself in the present case. The issue was whether police officers owed a duty of care to a bystander who was injured in the course of an attempt to arrest a suspect. The case was one of physical injury, so the ultimate decision proceeded by reference to established criteria; but the judgments contain important general observations about the correct approach in novel situations. At paras. 21-30 of his judgment Lord Reed, with whom Lady Hale and Lord Hodge agreed, addressed the role of the “tripartite test” said to be derived from Caparo. He makes his fundamental point in para. 21, as follows:

“The proposition that there is a Caparo test which applies to all claims in the modern law of negligence, and that in consequence the court will only impose a duty of care where it considers it fair, just and reasonable to do so on the particular facts, is mistaken. As Lord Toulson pointed out in his landmark judgment in Michael v Chief Constable of South Wales Police [2015] UKSC 2; [2015] AC 1732, para 106, that understanding of the case mistakes the whole point of Caparo, which was to repudiate the idea that there is a single test which can be applied in all cases in order to determine whether a duty of care exists, and instead to adopt an approach based, in the manner characteristic of the common law, on precedent, and on the development of the law incrementally and by analogy with established authorities.

He makes that point good by an analysis of Caparo in the following paragraphs, and continues, at para. 27:

“It is normally only in a novel type of case, where established principles do not provide an answer, that the courts need to go beyond those principles in order to decide whether a duty of care should be recognised. Following Caparo, the characteristic approach of the common law in such situations is to develop incrementally and by analogy with established authority. The drawing of an analogy depends on identifying the legally significant features of the situations with which the earlier authorities were concerned. The courts also have to exercise judgement when deciding whether a duty of care should be recognised in a novel type of case. It is the exercise of judgement in those circumstances that involves consideration of what is ‘fair, just and reasonable’.”

(As para. 21 makes clear, Lord Reed was there concerned to scotch a particular heresy about the application of a test of what is “fair, just and reasonable”. His statement in para. 27 of the correct approach in a novel situation is not of course intended to exclude considerations of foreseeability or proximity so far as relevant in the particular case.)

Lord Mance delivered a concurring judgment, which helpfully glosses Lord Reed’s observations and expressly refers to the role of the concept of assumption of responsibility. At para. 83 he says:

“As Lord Reed demonstrates, it is unnecessary in every claim of negligence to resort to the three-stage analysis (foreseeability, proximity and fairness, justice and reasonableness) identified in [Caparo]. There are well-established categories, including (generally) liability for causing physical injury by positive act, where the latter two criteria are at least assumed. The concomitant is that there is, absent an assumption of responsibility, no liability for negligently omitting to prevent damage occurring to a potential victim. … Economic loss also falls outside the established category of liability for physical injury, but an assumption of responsibility for economic loss will, as discussed in [Hedley Byrne], likewise satisfy the latter two Caparo criteria. Outside any established category, the law will proceed incrementally, and all three stages of the Caparo analysis will be material.”

That passage makes clear that the role of “assumption of responsibility” is as a characterisation of why, in certain kinds of situation, a relationship of proximity should be recognised and/or why it is fair, just and reasonable that one party to a relationship should have a duty to take care to protect the other from the kind of damage in question.

In James-Bowen v Commissioner of Police for the Metropolis [2018] UKSC 40, [2018] ICR 1353, which concerned the existence of a duty of care for economic and reputational loss in another novel situation, the Supreme Court applied the approach identified by Lord Reed in Robinson: see para. 22 of the judgment of Lord Lloyd-Jones. I need only note that a case based on assumption of responsibility had been rejected in the Court of Appeal and was not pursued in the Supreme Court.

In his skeleton argument Mr Ciumei based his submission that the Judge was obliged to proceed by reference to the concept of assumption of responsibility on observations by Lord Wilson in NRAM Ltd v Steel [2018] UKSC 13, [2018] 1 WLR 1190, and by Lord Sumption in Playboy Club London Ltd v Banca Nazionale del Lavoro SpA [2018] UKSC 43, [2018] 1 WLR 4041. I take them in turn.

In NRAM v Steel the solicitor for a borrower had carelessly informed the lender that the entirety of the loan was being repaid and the lender in reliance on that statement discharged the security that it held and suffered loss as a result. It was thus a case of negligent misrepresentation, the issue being whether the relationship between the borrower’s solicitor and the lender was such as to give rise to a duty of care. The Supreme Court held that it was not. Lord Wilson gave the only judgment. At para. 24 he referred to Henderson v Merrett and Spring v Guardian Assurance plc [1995] 2 AC 296. He also quoted a statement by Lord Steyn in Williams v Natural Life Health Foods Ltd [1998] 1 WLR 830 that “there was no better rationalisation for liability in tort for negligent misrepresentation than the concept of an assumption of responsibility”. He continued:

“It has therefore become clear that, although it may require cautious incremental development in order to fit cases to which it does not readily apply, this concept remains the foundation of the liability.”

The natural reading of “the liability” is as a reference to liability for negligent misrepresentation, which is what was being referred to by Lord Steyn in the immediately preceding quotation and which was the kind of liability in issue in NRAM v Steel itself. I accept that Lord Wilson may also have had in mind the rather wider kinds of case exemplified by Henderson v Merrett and Spring v Guardian Assurance (Footnote: ²), to which he had referred earlier in the passage; but there is no reason to suppose that he was intending any wider proposition. Even if he was, however, his statement would be both obiter and wholly out of line with the clear and considered views of the members of the Appellate Committee in the Barclays Bank case (Footnote: ³), which post-dates Henderson v Merrett, Spring v Guardian Assurance and Williams v Natural Life and in which all three are considered.

In the Playboy Club case a bank carelessly gave a good financial reference for a customer who wished to gamble in the claimant’s casino. In reliance on the reference the casino cashed cheques for him which turned out to be counterfeit. This was thus also a case of negligent misrepresentation. The Supreme Court held that the bank did not owe the claimant a duty of care. Lord Sumption gave the only judgment. He referred to Hedley Byrne and the principle of “express or implied undertaking of responsibility” and continued, at para. 7:

“The principle thus established is capable of development. Indeed it has undergone considerable development since 1964, for example to cover omissions and the negligent performance of services. But these have been incremental changes within a consistent framework of principle. One area in which the courts have resisted expanding the scope of liability concerns the person or category of persons to whom the duty is owed. The defendant’s voluntary assumption of responsibility remains the foundation of this area of law, as this court recently confirmed after a full review of the later authorities in [NRAM], paras 18-24 (Lord Wilson JSC).”

Again, the natural reading of “this area of the law” is as a reference to negligent misrepresentation, though perhaps also as extended by the developments since 1964 to which Lord Sumption refers. I do not think it can be read, any more than Lord Wilson’s statement in NRAM v Steel, as endorsing the adoption of assumption of responsibility as the universal touchstone for the existence of a duty of care; but even if that were the correct reading it could not be authoritative, for the reasons which I have given.

I should add that, as Mr Goulding pointed out, both Robinson and James-Bowen were decided very close in point of time to NRAM v Steel and the Playboy Club case and there was substantial overlap between the constitutions; so that it would be very unlikely that there is any inconsistency between their reasoning.

Mr Ciumei also referred us in his oral submissions to the recent decision of the Privy Council in JP SPC 4 v Royal Bank of Scotland International Ltd [2022] UKPC 18, [2022] 3 WLR 261. In that case a bank paid away moneys, allegedly carelessly, from the account of a customer of which the claimant investment fund was the beneficial owner. The fund advanced a number of alternative bases for the existence of a duty of care owed to it, including “assumption of responsibility” and “incremental development”. Both were rejected: see paras. 59-68 and 69-84 respectively. In connection with the former, Lord Hamblen and Lord Burrows, with whom the other members of the Board agreed, said at para. 60:

“The principle of an assumption of responsibility has been criticised by some commentators as being elusive and tending to obscure the real reasoning: see, for example, Kit Barker, ‘Unreliable assumptions in the modern law of negligence, (1993) 109 LQR 461; and Christian Witting, Street on Torts, 16th ed (2021), pp 53-55. Other commentators consider it to be an important and distinctive source of legal obligation: see, for example, Donal Nolan, ‘Assumption of Responsibility: Four Questions’ (2019) 72 Current Legal Problems 123. The courts have continued to apply it and to find it useful.”

(I would add to that list of academic criticism the trenchant analysis by Professor Stapleton in Three Essays on Torts (2021), pp. 53-57.) Mr Ciumei relied on the observation by Lord Hamblen and Lord Burrows that the courts have continued to apply the concept of assumption of responsibility and to find it useful. That is incontrovertible, but it says nothing about in what circumstances, or for what purposes, it may be usefully applied. In fact the reference to “usefulness” (also Lord Hoffmann’s term in the Barclays Bank case) is inconsistent with it having any status as the foundation principle in this area.

I should add that in JP SPC 4 Lord Hamblen and Lord Burrows quoted a passage from the judgment of Lord Reed in N v Poole Borough Council [2019] UKSC 25, [2020] AC 780, which expounds the origins and nature of the concept of assumption of responsibility. I need not summarise that exposition here: it is enough to say that there is nothing in it which lends any support to Mr Ciumei’s case.

In the light of those authorities it seems to me that the position is as follows. The correct course for a Court which has to decide whether a duty of care should be recognised in a novel situation is to take the incremental approach endorsed in Robinson. That will in principle involve consideration of the three “Caparo factors” to the extent that they are in issue. It may be a useful analytical tool, particularly in considering the factors of proximity and/or “fairness, justice and reasonableness”, to ask whether the defendant can be regarded as having assumed a responsibility to take care to protect the claimant against a loss of the kind claimed; but its usefulness will depend on the issues in the particular case.

Applying that approach in the present case, it was quite unnecessary for the Judge to make any reference to the concept of assumption of responsibility. It could add nothing on the issue of proximity: the Claimant was the Bank’s employee and there is no need to look for some relationship “akin to contract” of the kind identified in Hedley Byrne. Nor can I see why it would be an apt or useful tool in the present case for considering whether it was fair, just and reasonable to impose a duty in this case. But the main point is that the decisive issue in the Judge’s reasoning was foreseeability: see para. 225 of the judgment (para. 28 above). As Singh LJ pointed out in the course of argument, reference to assumption of responsibility adds nothing on that question: the Bank could not have assumed responsibility for risks that were not reasonably foreseeable.

Rihan

The pleaded criticism, as we have seen, is that the Judge “erred … in failing to draw an analogy from the audit duty found in Rihan v Ernst & Young Global Ltd …”. The legal and factual issues in that case were complex and Kerr J’s judgment correspondingly lengthy. I gratefully adopt the summary given in the Westlaw digest:

“The claimant (R) claimed damages for negligence and conspiracy to injure from four UK-based entities that were part of a network of companies providing accountancy and related services to businesses worldwide (EY).

In 2013, while a partner in EY's Middle East and North Africa (MENA) entity, R conducted an assurance audit of a Dubai-based client (K). He claimed to have discovered that K was participating in irregular activities which suggested that it was involved in money laundering; that the local regulator (the DMCC) pressured him to cover up his findings; and that the DMCC and K required him to conduct the audit unethically and in a way that amounted to professional misconduct. He asserted that the defendants colluded with the DMCC in that regard, which led to his resigning, publicly disclosing the wrongdoing, and fleeing Dubai out of fear for his safety. He claimed that he was thereafter unable to secure alternative employment and his earning capacity was largely destroyed. He sought damages for economic loss, mainly in the form of loss of earnings. He claimed that the defendants had breached two duties of care: a duty to take reasonable steps to keep him safe by relocating him outside Dubai (the safety duty), and a duty to take reasonable steps to prevent him suffering financial loss by reason of their failure to conduct the audit ethically and without professional misconduct (the audit duty). He also claimed that they had conspired to injure him by driving him out of EY.

The issues were whether any such duties existed and, if they did, whether they had been breached; whether there was a conspiracy to injure R; and whether, in terms of any loss, the chain of causation had been broken because R's decision to make public disclosures was unreasonable.”

I should add that the defendants did not include the legal entity in which the claimant was a partner. Accordingly, the two duties on which he sought to rely were said to be owed in tort, albeit that he relied by analogy on the duties that would have been owed to him as an employee.

Kerr J dismissed the claim based on the alleged “safety duty” – that is, a duty to take reasonable steps to keep the claimant safe by relocating him. The Claimant does not rely on that aspect of his decision, but it is worth quoting Westlaw’s summary of his reasoning because it has some resonances for the present claim:

“The defendants did not owe the safety duty. It would not be legitimate to extend an employer’s duty to safeguard its employees against personal injury to a duty to safeguard them against pure economic losses arising from their need to cease working to avoid a threat to their safety. The duty to provide a safe place and system of work did not extend to protecting purely economic interests. Although an employee who was instructed to work in unsafe conditions might be entitled to resign and claim constructive dismissal, they could not recover the entire cost of a lost career (see paras 476-487 of judgment).”

Kerr J did, however, find that the “audit duty” – that is, a duty on EY to take reasonable steps to prevent the claimant suffering financial loss by reason of their failure to conduct the audit ethically and without professional misconduct – was established, and that it had been breached. As regards the existence of the duty, he directed himself, like Freedman J in the present case, in accordance with the guidance in Robinson. He found that the requirements of foreseeability of loss and of proximity were met (see paras. 587 and 595). He also held that it was “fair, just and reasonable” to impose such a duty. His reasoning sufficiently appears from para. 621:

“The duty is part of the obligation of the employer (or quasi-employer) to provide an acceptable work environment. The physical integrity of the employee is protected against injury by the classic duty of care to take reasonable steps to provide a safe place of work and a safe system of work. By parity of reasoning, I see no reason why, in certain circumstances, the moral and professional integrity of the employee (or quasi-employee) should not be protected by a duty to take reasonable steps to provide an ethically acceptable work environment, free of criminal conduct … and free of professionally unethical conduct.”

It is in my view clear from that summary that the reasoning in Rihan has no application to the circumstances of the present case. In the first place, Kerr J found that the loss complained of by the claimant was foreseeable, whereas Freedman J made the opposite finding here. Secondly, the essence of the duty found by him was that EY should conduct the audit ethically and without misconduct. The duty alleged in the present case has nothing to do with the Bank avoiding wrongdoing: it is a duty to protect him against the wrongdoing, or in any events the unjust acts, of others. That being so, the complaint that the Judge failed to “draw an analogy” with the finding of the audit duty in Rihan is misconceived.

I should as a matter of prudence make it clear that I express no view either way about the correctness of Kerr J’s decision about the audit duty. We were told that permission to appeal to this Court was given but that an appeal was not in the event pursued.

I would dismiss ground 1. The pleaded challenges do not show any error in the approach which the Judge took to the question of whether a duty of care arose.

Ground 2 reads:

“The Judge erred in failing to take into consideration relevant and material evidence as to whether the Bank owed such a duty of care and/or reached conclusions that were not reasonably open to him, in that:

A.

It was not reasonably open to the Judge to conclude at [216] in relation to the ‘red flags’ relied on by the Claimant as alerting the Bank to the real risk of conviction he faced, that ‘none of the flags have been proven’ or came to be in the knowledge of the Bank so as to require a response from the Bank. This was contradicted by the unanimous expert evidence, incorrect on the face of the ‘red flag’ documents themselves and contrary to the witness evidence, the credibility of which was not in doubt. None of these matters were considered.

B.

The Judge erred in deciding by reference to a pleading point at [198] and [213] that the Bank did not know about the critical fourth to seventh ‘red flags’.

C.

The Judge erred at [88]-[89] and [285] in failing to draw adverse inferences from the Bank’s failure to call key witnesses or give any disclosure as to risk assessment undertaken by the Bank (see also [258]-[261]) and, most importantly, to give disclosure or call evidence in relation to the pleaded allegations at paragraph 52.2 of the Amended Particulars of Claim that the Bank failed to respond to ‘red flags’.

D.

The Judge was wrong to disregard (or failed adequately to take into account) at [168] evidence of the risks of doing business in Romania, which is not referred to in the Judgment, and wrong to conclude that there was only one item of evidence to the contrary at [175].”

These points are developed at paras. 34-42 of Mr Ciumei’s skeleton argument, with considerable cross-reference to the evidence of the expert and other witnesses (both the written evidence and the transcripts of the oral evidence) and to the interview notes generated by the Bank’s internal investigation – specifically the Rao and Burkey memoranda (see para. 27 above). Mr Goulding’s skeleton argument responded in kind. This material is dense and allusive, and at the start of the hearing the Court told counsel that we would welcome particular assistance on these aspects. In the event, Mr Ciumei chose to address the Court on only one aspect of the evidence, relating to the Burkey memorandum. I am in no way critical of that choice: he evidently considered that the Judge’s treatment of that memorandum was the best example of the alleged errors in his fact-finding and that it made better sense to devote most of the available time to the other grounds. But it affects the detail in which I need, or indeed can, address the other matters raised by ground 2.

It is convenient at this stage to set out the relevant parts of the Burkey memorandum:

“1.

NB [i.e. Mr Burkey] said that, about a month or so ago, Frontier Consultants, FID’s Bulgarian advisers on the Trakiya financing in Bulgaria, told CS [Credit Suisse] that it had come to the attention of the Bulgarian Government, through information sharing between the Bulgarian and Romanian security services, that a file was being compiled in Romania on Stamen Stantchev and CS in the context of the privatisation of Petrom. Michal Susak (MS) was also mentioned. The suggestion was that there had been bribery or corruption surrounding the privatisation of Petrom. NB said that, in light of the information the Bulgarian Government had received, the Bulgarian Government was concerned that it had instructed CS on the Trakiya deal.

2.

NB spoke to Vadim Benyatov (VB) regarding the information NB had received. VB said that there was nothing more than gossip which was fuelled by a frustration at the sale of Petrom for a price lower than might have been achieved at a later date because of the post privatisation increases in oil prices. Frontier Consultants passed this back to the Bulgarian Minister to whom they had previously spoken.

3.

NB said that Emil Stava, a Director in FID, had met with MS a couple of nights previously and MS did not even know which deal was the subject of the criminal investigation in Romania.

4-7. …

(Mr Stantchev was a local consultant of the Bank in Romania.) It will be seen that the principal relevance of the memorandum is to head A, and specifically to the sixth warning flag, though it may have some tangential relevance to heads B and C.

I start by setting out para. 216 of the judgment, since it is the focus of the Claimant’s challenge:

“There is a further matter to add. In a case such as this with numerous interactions, it is possible with the advantage of hindsight to pull out a document and seek to say that if only the Bank had reacted, then the arrest and the conviction could have been prevented. That is not sufficient. They have to be real flags, that is to say matters that ought to have been noticed and acted upon at the time. They have to be matters which were, or ought reasonably to have been, in the knowledge of the Bank. They have to be matters which if they had registered at the time ought to have been acted upon in a manner which would have avoided a conviction or reduced the chance of a conviction. In this case, (a) without a specific pleading in advance of the parts of the documents dated 24 and 29 November 2006 now relied upon (which themselves were compiled only after the arrest of Mr Benyatov), (b) without proving relevant knowledge on the part of the Bank of any of these matters, and (c) without Mr Benyatov having escalated within the Bank such knowledge as he may have had, none of the fifth to the seventh flags have been established. Indeed, for all the reasons referred to by the Bank and for the reasons set out in this conclusory section, none of the flags have been proven let alone that they were warning signals on which the Bank ought to have acted.”

The challenge under this head is to the second half of that paragraph, and more particularly to the Judge’s points (b) and (c). Although the ground refers to the Judge having said that “none of the flags have been proven”, in fact he was referring specifically to the fifth to seventh flags, concerning the interest of the SRI in the Petrom privatisation, which were clearly the main focus of the case being advanced to him. He had indeed elsewhere in his judgment made negative findings about the first three flags, but they are not referred to in Mr Ciumei’s skeleton argument and it is evident that they are not challenged on this appeal. As to the fourth – adverse press comment about the Petrom privatisation – that is touched on in the skeleton but it is clearly of secondary importance compared with the allegation of SRI interest, and it did not feature in Mr Ciumei’s oral submissions.

The Judge’s points (b) and (c) are summaries of what the Judge had said more fully at paras. 213-215 (which in turn build on his review of the evidence and submissions at paras. 181-206). In view of Mr Ciumei’s focus on the Burkey memorandum, I need only quote para. 214. After referring to the pleading issue it continues:

“The note only reveals knowledge on the part of Mr Burkey and Mr Benyatov. Mr Burkey was not senior to Mr Benyatov: they were both at Managing Director level. Mr Benyatov was clearly unconcerned about the issue and hence did not escalate it internally. He was so unconcerned that he did not recall the conversation: see para. 80 of his third witness statement. Even if this had been proven, there is no pleaded allegation as to what the Bank could or should have done and bearing in mind that at this time the investigation had been started. In the case of Mr Flore, he was indicted even though he was not in the country. Here too, for these reasons and for any additional reasons relied upon by the Bank as set out above, the existence of this flag has not been proven.”

(The reference to “any additional reasons” is to the summary of Mr Goulding’s submissions in para. 203, but these do not appear to add anything of substance.)

In his oral submissions Mr Ciumei made three points about the Judge’s conclusions on the Burkey memorandum.

First, he relied on para. 38 of his skeleton argument, which sets out passages from the expert evidence which establish that knowledge that the SRI was investigating a transaction in which the Bank was concerned would be a “significant event which should have triggered a response by the Bank”. That is no doubt the case, but the Judge’s reasoning was directed to the separate question of what knowledge the Bank in fact had. Mr Goulding drew our attention to passages in the transcripts of the oral evidence of the witnesses that recognised that distinction, particularly in the cross-examination of Mr Worman, who said that if “someone senior” had been given the information an alarm bell should have been rung but accepted that that depended on who was told.

Second, he submitted that para. 1 of the memorandum was decisive of the issue of knowledge because it states that Mr Burkey had said that Frontier Consultants “told CS” that, in short, the Romanian authorities were investigating the Bank’s role in the privatisation of Petrom. But that says nothing about who within the Bank – and more particularly at what level of seniority – had been given that information. It is necessary, in an allegation of knowledge of this kind, to identify an individual: “CS” (even assuming Frontier Security’s information was correct) could be anyone. At para. 38.2.1 of his skeleton argument Mr Ciumei pointed out that para. 3 of the memorandum shows that Mr Susak and Mr Stava also knew of the allegation; but it does not show that they did so prior to Mr Benyatov’s arrest, and indeed the context suggests that the reference is to conversations consequent on it. He also contended that “it is to be inferred that others in the Bank knew, including Mr Kyriakos-Saad, who was senior to Mr Burkey as head of the fixed income division in Emerging Markets” (Footnote: ⁴). Mr Goulding submitted that there was no basis for such an inference. On what we have seen, I would agree; but in any event the question for us is whether the basis in the evidence for such an inference was so strong that the Judge should have drawn it, and I am satisfied that that is not the case. (If the Bank had been put on notice that the Claimant was relying on this statement these points might perhaps have been elucidated – one way or the other – by the calling of further evidence, but it was not: as to this, see further the discussion of heads B and C below.)

Third, Mr Ciumei submitted that it was irrelevant that Mr Benyatov himself was said by Mr Burkey to have made light of the information, because any robust risk management system should ensure that such decisions are not left to those on the front line, whose judgment was liable to be impaired by being too closely involved in the deals. That is a fair point as far as it goes, but it goes nowhere unless the information in question reaches those who may be in a position to make a cooler judgment.

For those reasons I do not believe that anything in Mr Ciumei’s submissions on the Burkey memorandum impugns the Judge’s approach to his fact-finding or his conclusion in para. 214.

This head challenges the Judge’s reliance on the pleading point referred to at (a) in para. 216. It does so specifically by reference to paras. 198 and 213 of the judgment. I have set out the relevant background at para. 17 above.

Para. 198 reads:

“[I]n order to make an allegation of actual knowledge, the documents evidencing the same ought to have been pleaded by no later than the time of the amendment before the Deputy Judge. Failing this, in the light of the above ruling, they should have been pleaded in advance of the time for witness statements and in any event, well in advance of the trial. The consequence is that the Bank did not have the opportunity to meet a case about actual knowledge and in particular to consider whether to call witnesses to deal with the documents referred to below [i.e. the Rao and Burkey memoranda].”

As for para. 213, which is concerned with the fifth warning flag, for which the Claimant relies on the Rao memorandum, this begins:

“As regards the SRI investigations into Petrom in June 2006, the point about no particularisation of the knowledge of the Bank that the Romanian secret police had an interest in the transaction is a good one.”

(The Judge puts the point succinctly because he had already set out the background and the parties’ submissions at paras. 196-201.)

As to this issue, Mr Ciumei adopted what he had said at para. 39 of the Claimant’s skeleton argument, as follows (omitting references and footnotes):

“[T] he Court erred in deciding by reference to a pleading point at [198] and [213] that the Bank did not know about the critical fourth to seventh red flags. The Bank knew in advance that these matters were in issue, since they were pleaded at AMPOC paragraph 52.2 on the basis of an inference of knowledge and matters were canvassed in further detail in witness and expert evidence in advance of trial. The Judge was wrong to find that this issue did not arise until trial at [214]. No objection was made at trial to any questioning on these documents (nor could there have been such an objection). The Judge was wrong to rely on a pleading point when actual knowledge was demonstrated at trial.”

The short answer to this submission is that the Judge did not reject the case that the Bank knew about the warning flags only on the basis of the pleading point. As we have seen, he made an explicit finding that the contents of the Burkey memorandum did not prove the requisite knowledge as regards the sixth flag; and there are equivalent findings as regards the Rao memorandum and the fifth and seventh flags, which are considered at paras. 213 and 215 respectively. Even if the pleading point were bad those findings would stand: contrary to what is said in the ground of appeal, this is not a case where “actual knowledge was demonstrated”. In fact, however, I do not believe that the Judge intended the pleading point to be a free-standing reason for rejecting the Claimant’s case on knowledge. Rather, I understand him to have been explaining why the Bank had called no witnesses – most obviously Mr Burkey or Mr Rao – to explain or amplify what was said in the memoranda: it had no reason to suppose that any reliance was being placed on them.

That being so, I need not address all the points made in para. 39 of the Claimant’s skeleton argument. I would only say that the fact that the Burkey and Rao memoranda were referred to in witness statements or the expert reports is not in itself an answer to the objection that they were not relied on in the pleading. As Mr Goulding reminded us by reference to the passages from various authorities set out at paras. 55-57 of the judgment of Cotter J in Charles Russell Speechlys LLP v Beneficial House (Birmingham) Regeneration LLP [2021] EWHC 3458 (QB), the principle that “the pleadings frame the limits of the action” remains of fundamental importance; and reliance on the pleadings in the present case is certainly not a mere technicality given the importance attached at the hearings before Mr ter Haar to identifying exactly what was being alleged and on what basis.

This head complains of the Judge’s failure to draw adverse inferences from what is said to be the Bank’s failure (1) to call key witnesses and (2) to give proper disclosure. I take those two aspects in turn.

As regards (1), the Judge records at para. 86 of his judgment that Mr Ciumei in his closing submissions identified no fewer than nine employees or former employees of the Bank who were not called as witnesses but who might have been expected to know more about the various issues than those who were in fact called. There is nothing in either Mr Ciumei’s skeleton argument or his oral submissions developing that point; but two of the “missing witnesses” are Mr Rao and Mr Burkey, and in their case, as I have said, the answer to the criticism is, as we have already seen, that there was no pleaded reliance on those documents. The Judge makes that point at para. 87, and it seems to me unimpeachable.

As regards (2), the Judge identifies at para. 258 of his judgment three failures of disclosure alleged by Mr Ciumei in his closing submissions. All concern aspects of the Bank’s risk assessment process, either generally or in relation to doing business in Romania or to the EMS transaction. It is not clear to what extent the complaint was that relevant documents had been withheld or that the absence of disclosure showed that documents of the relevant kind did not exist and thus that proper procedures were not followed. At paras. 259-261 the Judge considers each category and explains why the absence of any disclosure of the kind in question was of no significance. The Claimant’s skeleton argument makes no attempt to challenge the reasoning in those paragraphs and Mr Ciumei made no reference to the point in his oral submissions. In those circumstances I need say no more than that I can see nothing wrong in the Judge’s reasoning or conclusions.

This head challenges paras. 168 and 175 of the judgment. These form part of section (d) of Part VII, which is headed “Romania as a ‘high-risk’ country and the degree of perceived risk”. It is followed by section (e), comprising paras. 176-178, which is headed “High-risk transactions and the degree of perceived risk”. The Judge acknowledges in para. 176 (1) that there is an overlap between the two sections, and they should in my view be read together.

At paras. 168-173 of section (d) the Judge sets out a number of substantial pieces of evidence supporting the view that Romania was not at the relevant time regarded as a “high-risk” country, albeit that there were some areas of concern. I need not summarise the passage in detail, but I should note that the evidence to which he refers includes para. 7 (v) of the joint statement of Mr Worman and Dr Donald. That reads:

“Comprehensive political risk analyses were more frequently undertaken by companies for high-risk territories such as those emerging from conflict or prolonged political instability, or those with serious issues of organised criminality or a predatory state. Examples from the Relevant Period would include The Democratic Republic of the Congo, Iraq, and Russia. Lower risk territories, where one or more of these factors might be present but to a lesser degree, would be less likely to be the subject of comprehensive political risk analyses. Examples from the Relevant Period would include Nigeria, Pakistan, and Albania. Romania during the Relevant Period would fall under this category of lower risk territories. While a comprehensive political risk analysis would be less likely, due diligence (business intelligence) with possible elements of political risk analysis may have been employed for deals in problematic sectors or those with political dimensions.”

The Judge summarises that passage as follows:

“In that part of the joint report, ‘cases of high risk countries were territories emerging from conflict or prolonged instability or those with serious issues of organised criminality or a predatory state’. Examples at the relevant time included the Democratic Republic of Congo, Iraq and Russia. Lower risk countries where one or more of the above factors might be present, but to a lower degree, included Nigeria, Pakistan, Albania and Romania at the relevant time.”

At para. 175 the Judge considers the evidence of Mr Simon Menneer (a former Chief Operating Officer of the Bank) that Romania was indeed a high-risk country at the relevant time. He acknowledges that that evidence is entitled to considerable respect, but he gives reasons why he does not accept it. He concludes:

“In the end the Court prefers the evidence of the majority to the effect that Romania was not a high-risk country or perceived as such at the relevant time.”

As already noted, we heard no oral submissions in support of this head and we can proceed only by reference to the Claimant’s skeleton argument, where it is developed in para. 41. The points there made can be sufficiently summarised as follows:

I am not persuaded that any of the points there made undermines the Judge’s conclusion in section (d) of Part VII. It is unnecessary to recapitulate here the well-known cautions about an appellate court overturning the factual conclusions of the trial judge, and more particularly about the risks of “island-hopping” in a case like the present where there has been a sea of evidence (see, most recently, Volpi v Volpi [2022] EWCA Civ 464, [2022] 4 WLR 48, per Lewison LJ at paras. 2-5). As regards the three points summarised above:

I would dismiss ground 2. None of the points raised establishes that the Judge reached conclusions that were not open to him on the evidence.

I begin by setting out the Judge’s reasoning. At para. 291 of the judgment he directed himself as to the applicable law by reference to the following passage from para. 12 of the judgment of Clarke LJ in Hatton v Chafes [2003] EWCA Civ 341:

“The following principles are not in dispute and may be summarised in these propositions:

(i)

A cause of action in negligence does not arise until the claimant suffers damage as a result of the defendant’s negligent act or omission.

(ii)

The damage must be ‘real’ as distinct from minimal: Cartledge v Jopling [1963] AC 758 per Lord Reid at 771 and Lord Evershed MR at 773–4.

(iii)

Actual damage is any detriment, liability or loss capable of assessment in money terms and includes liability which may arise on a contingency: Forsted v Outred [1982] 1WLR 86 per Stephenson LJ at 94, approved by the House of Lords in Nykredit [Nykredit Mortgage Bank Plc v Edward Erdman Group Ltd (no. 2) [1997] 1 WLR 1627] per Lord Nicholls (with whom the other members of the appellate committee agreed) at 630F.

(iv)

The loss must be relevant in the sense that it falls within the measure of damages applicable to the wrong in question: Nykredit at 1630F. (Propositions (i) to (iv) were confirmed by Sir Murray Stuart-Smith in Khan v Falvey [[2002] EWCA Civ 400] at paragraphs 11 and 12.)

(v)

A claimant cannot defeat the statute of limitations by claiming only in respect of damage which occurs within the limitation period if he has suffered damage from the same wrongful act outside that period: Khan v Falvey at paragraph 23, following Knapp v Ecclesiastical Insurance Group Plc [1998] PNLR 172 per Hobhouse LJ at 184 and 187.”

That summary was not challenged by Mr Ciumei.

The Judge’s starting-point is that any breach of the pleaded duties must have been suffered prior to the Claimant’s arrest in November 2006. The issue accordingly is when the Claimant first suffered any damage of the relevant kind as a result of that negligence. He shows at paras. 293-297 that on the basis of the Claimant’s own evidence he suffered a loss of bonus of the order of US$500,000 in each of the years 2006 and 2007, principally because of the effect of his being imprisoned and then unable to leave Romania for the greater part of 2007. That damage was on any view more than minimal, with the result that that was when time started to run; the limitation period accordingly expired long before the proceedings were commenced in January 2018. It made no difference that the Claimant only pleads a claim for loss of earnings from the date of the termination of his employment on 13 June 2015: that is the point of Clarke LJ’s proposition (v).

The Claimant had sought to contend that, since the pleaded duty was only to protect him from “criminal conviction [my emphasis] and the resulting losses” (see para. 24.1 of AMPOC), any losses caused prior to his conviction in 2013 were irrelevant. The Judge rejected that submission at paras. 298-300. He said:

“298.

The duty is framed in a way that captures only loss of earnings consequent upon a conviction. The question is then whether that really is the duty, assuming that there was a duty. It was submitted on behalf of Mr Benyatov that the duty situation is limited to not exposing Mr Benyatov to the risk of conviction and loss of career earnings thereafter. It had not been formulated as a duty not to expose Mr Benyatov to the risk of arbitrary arrest, detention and restriction of movement, which would have been more akin to a physical safety duty. It was submitted that ‘there was no real loss in relation to those’ [the quote is from Mr Ciumei’s closing submissions]. In my judgment, there is an artificiality about capturing the duty in the way in which it is pleaded. It was not just a conviction and the subsequent loss of earnings which ensued from the failure to exercise reasonable care. It was an arbitrary process of arrest, imprisonment, house arrest, facing a criminal trial and conviction. The case is in reality predicated upon an abusive process from start to finish. In his opening skeleton (at para. 124), Mr Benyatov referred to the duty of care being one whereby the Bank was obliged to take reasonable steps to identify and avoid certain risk which Mr Benyatov faced in carrying out the EMS project in Romania ‘including in particular the risk of wrongful prosecution and conviction’. The prosecution took place over a period of years. Either a part of that prosecution was the arrest, imprisonment and incarceration in Romania, or it is so closely related to it that it must be a part of the risk for which the Bank must be responsible to exercise reasonable care to identify and avoid.

299.

Another way of looking at the matter is that if the duty is to take reasonable care not to expose Mr Benyatov to conviction, that itself is wide enough to embrace the steps which led to conviction including investigation, arrest, imprisonment and incarceration. In my judgment, it is artificial to frame the duty simply by reference to the loss of earnings subsequent to conviction. There is no reason to restrict it to these losses consequent upon conviction and not to include losses suffered in 2007-2008 referred to above. It is right to say that the losses were not total losses of earnings because Mr Benyatov was still remunerated throughout this period, and it is therefore a different extent of loss that ensued after conviction. However, it was nonetheless ‘real’ as distinct from ‘minimal’. His own evidence bears out both that the incarceration was the cause of his losses and that the losses amounted to hundreds of thousands of dollars.

300.

It therefore follows that if, contrary to the above, the Court had found that there was a duty of care or a breach of a duty of care, this is a case where Mr Benyatov cannot defeat the statute of limitations by claiming only in respect of damage occurring within the limitation period given that he has suffered damage from the same wrongful act outside that period. By parity of reasoning, a person cannot tailor the duty situation so as to capture losses claimed in time when in fact the duty, if it had existed, would capture losses outside the limitation period. For this reason also, the claim in tort fails.”

Ground 3 reads:

“The Judge erred in law in finding at [297]-[300] that the tort claim was time-barred. Mr Benyatov’s loss flowed from his inability to work as a finance professional, which did not take place until after his conviction, and was thus within time.”

Mr Ciumei’s first submission before us was that it was up to the Claimant how he formulated his claim. If he chose to rely on a duty of care to prevent losses caused specifically by his conviction, the Judge could not go behind that choice. I do not accept that. The proper formulation of a duty of care is a matter for the Court to determine, having regard to the nature of the harm against which, judged objectively, the defendant was under an obligation to protect the claimant; it cannot be artificially divided into a series of discrete duties, defined by reference to the particular form which the loss of the relevant kind took on this or that occasion. If contractors carrying out works near my house owe me a duty of care in tort to see that no damage is caused, that cannot be broken down into a series of discrete duties to prevent damage to different parts of the house – the walls, the roof, the windows – in order to circumvent limitation difficulties.

Mr Ciumei argued that, even if that were correct, the pleading of a duty of care to prevent loss of earnings specifically as a result of a conviction (rather than simply as a result of criminal proceedings) was not artificial. The loss of earnings, past and future, caused by his conviction was of a different character from the loss of bonus caused by the earlier stages of the proceedings because it was inherently career-long. I cannot accept that either. In both respects the Claimant has suffered a loss of earnings (or, as regards the future, will do so): the difference between the two is simply one of degree. I did not understand Mr Ciumei to be drawing a distinction between bonus and other kinds of earnings, but if he was it is unsustainable: bonus is simply a form of earnings (and in the financial services industry very often more substantial than basic salary). I appreciate that the Claimant might not have thought it worthwhile or prudent to start a claim against the Bank when he suffered the initial loss, at a time when he was still an employee. But it was entirely foreseeable that if he was eventually convicted his losses would become much greater; and sometimes the risk of a claim being statute-barred means that a claimant will have to start proceedings in order to preserve their position for the future (unless they can reach a tolling agreement with the potential defendant).

In his skeleton argument Mr Ciumei contended that the Judge should have treated the adverse events occurring prior to the Claimant’s conviction as contingent only, referring to Law Society v Sephton [2006] UKHL 22, [2006] 2 AC 543. That is, with respect, hopeless. The losses of earnings in 2007 and 2008 to which the Judge refers were actual losses suffered in those years.

It follows that I believe that the Judge was right to hold that the negligence claim was statute-barred, for essentially the reasons that he gave, and I would dismiss ground 3.

I would accordingly dismiss each of grounds 1-3 and uphold the Judge’s dismissal of the negligence claim.

The Bank filed a Respondent’s Notice advancing a number of additional reasons why the Judge should have dismissed the negligence claim. These included challenges to the Claimant’s case on causation, contending (among other things) that, even if the Bank was subject to the pleaded duty, loss of earnings of the kind for which he claims was not within the scope of that duty. But we heard only very limited submissions on those issues, and I need not lengthen this judgment further by considering them.

Para. 19 of the AMPOC pleads the following terms of the Claimant’s contract of employment (referred to as “the Contract”):

“The following terms were implied into the Contract as a matter of law, fact or business efficacy. At all material times, the Defendant owed to Mr Benyatov the following duties:

19.1

To indemnify Mr Benyatov in respect of all losses, costs, expenses and claims he has suffered arising from or in consequence of faithfully, diligently or properly performing his duties on its behalf; and

19.2

To indemnify Mr Benyatov in respect of all losses, costs, expenses and claims he has suffered arising out of any unlawful enterprise upon which he was required to embark without knowledge that it was unlawful.

19.3.

The duties pleaded at paragraph 24 below, owed in parallel in contract (but in respect of which no contractual cause of action is advanced).”

I will refer to the indemnity pleaded in para. 19.1 as “the general indemnity” and to that pleaded in para. 19.2 as “the unlawful enterprise indemnity”. Para. 20 pleads that the terms in para. 19 continue to apply after the termination of the contract.

Although for completeness I have set out para. 19.3, which pleads that the “duties pleaded at paragraph 24 below” (i.e. the duties owed in tort) are also terms of the contract, it will be seen that it expressly disavows any claim based on those terms; and no breach of them is indeed pleaded. At first sight that may seem odd, but the explanation is no doubt that any claim based on them would be statute-barred, since it would run from the date of breach and any breaches will necessarily have occurred no later than the date of the Claimant’s arrest in November 2006. It is, however, worth noting that the absence of any such claim means that, although we are in this case dealing with alternative claims in tort and in contract, this is not the familiar situation where the two claims are essentially the same: the contractual indemnities on which the Claimant relies are different in character from the duties relied on in tort.

The terms in para. 19 are said to be implied “as a matter of law, fact or business efficacy”. That reflects the well-recognised distinction between terms which the law implies generally in the case of certain kinds of contractual relationship, such as a contract of employment, and terms which fall to be implied in the circumstances of a particular case. The distinction is discussed in Chitty on Contracts, 34^th ed., at para. 16-015, and I need not elaborate on it here. It may be that it is not absolute (see per Lord Wilberforce in Liverpool City Council v Irwin [1997] AC 239, at p. 254); but since the distinction is applied in the pleading it is right to observe it in this case.

As regards implication in fact, para. 25 of the AMPOC pleads that the Claimant relies on the same factual matters as those which are said to give rise to the duties of care in tort. In this regard, therefore, the claimed indemnity and the duty of care considered above arise out of the same circumstances – essentially, the allegedly high-risk nature of the Claimant’s work (as the Judge put it at para. 107 of his judgment, they have “a common factual substratum”); but the latter requires the Bank only to take reasonable care whereas the claimed liability under the indemnity is strict. Para. 25.21 also pleads:

“The terms are necessary to give business efficacy to the Contract and/or are so obvious that they go without saying, are consistent with the Contract, and are reasonable and equitable.”

As regards implication in law, para. 26 pleads:

“Facts and matters giving rise to the implication of the terms referred to in paragraphs 19.1, 19.2 and 19.3 of the Amended Particulars of Claim as a matter of law include the following:

26.1.

The facts and matters set out in paragraph 25 (including its sub-paragraphs other than sub-paragraphs 25.22 and 25.23) of the Amended Particulars of Claim; and

26.2.

The terms are a necessary incident of the relationship between employer and employee and/or principal and agent, as a consequence of reasonableness, fairness and balancing the competing policy considerations.”

The reference in para. 26.2 to “reasonableness, fairness and balancing the competing policy considerations” is taken from para. 36 of the judgment of Dyson LJ in Crossley v Faithful & Gould Holdings Ltd [2004] EWCA Civ 293, [2004] ICR 1615, where he says:

“It seems to me that, rather than focus upon the elusive concept of necessity, it is better to recognise that, to some extent at least, the existence and scope of standardised implied terms raise questions of reasonableness, fairness and the balancing of competing policy considerations.”

In Société Générale v Geys [2012] UKSC, [2013] 1 AC 523, Lady Hale observes that “there is much to be said for that approach”: see para. 56 of her judgment. There has been some debate about whether Dyson LJ meant that the test of necessity applicable in cases of implication as a matter of fact does not apply in cases of implication as a matter of law: see the discussion in ch. 6.1 of Lewison The Interpretation of Contracts, 7^th ed. I do not, however, need to resolve that issue here, and I am content to proceed on the basis of the Claimant’s pleading.

As we have also seen, paras. 53-55 of the AMPOC plead the identical “Indemnity/Loss and Damage” as regards the claims in both contract and tort. As there appears, the claim is only for loss of earnings (past and future). Thus, although the terms as pleaded in para. 19 formulate the Bank’s obligation as being to indemnify the Claimant against “all losses, costs, expenses and claims”, the only relevant element for our purposes is “losses”: it is not alleged that he has not been indemnified against any costs or expenses incurred in the course of his duties or that he has been subjected to any claims on the Bank’s account. (The Bank has of course paid all his legal expenses, though Mr Goulding was cautious about accepting that it had been under a legal obligation to do so.)

At para. 13.1 of the Amended Defence the Bank accepts that it was an implied term of the contract that it would

“indemnify the Claimant in respect of expenses and liabilities reasonably incurred by him in carrying out his duties as the Defendant’s employee and within the scope of his authority”.

Although the pleading does not say so in terms, I understand the admission to be that that implication arises as a matter of law, and that was Mr Goulding’s position before us. Para. 13.3 amplifies the Bank’s position as follows (so far as material):

“More particularly, as to paragraphs 19.1 and 19.2:

13.3.1.

It is denied that the Defendant had any or any implied obligation to indemnify the Claimant in respect of losses, costs, expenses or claims incurred by him as a result of, or arising from the acts of third parties or other intervening acts, whether wrongful or otherwise, including, for the avoidance of doubt, the events in Romania complained of by the Claimant.

13.3.2.

If and insofar as the Defendant owed the Claimant a duty to indemnify him in respect of expenses and liabilities, or losses, costs and claims, for the avoidance of doubt, such duty is limited to sums paid or payable by the Claimant to third parties, and does not extend to any sums not received by the Claimant, in particular earnings from employment.

13.3.3.

The alleged implied terms do not satisfy the tests for the implication of terms in fact or in law.

13.3.4.

… .”

In summary, therefore, it is common ground that there was an implied term of the Claimant’s contract of employment that the Bank would indemnify him against some forms of harm suffered in doing his job. The issue raised by this appeal is whether the forms of harm covered by the indemnity extend to a loss of earnings caused, as here, by the acts of a third party and without the need to establish any fault on the part of the Bank.

The main focus of the argument before us was on the general indemnity pleaded in para. 19.1, and I will consider that first.

As we have seen, in para. 26 of the AMPOC the Claimant bases his claim that the general indemnity falls to be applied as a matter of law both on (broadly) the particular circumstances of the present case (para. 26.1) and on the averment that it is “a necessary incident of the relationship between employer and employee” (para. 26.2). In my view, however, the former way of putting it falls to be considered in the context of implication as a matter of fact. At this stage, therefore, we are concerned only with the proposition that a general indemnity against loss of earnings “arising from or in consequence of [the employee] faithfully, diligently or properly performing his duties” – in short, in consequence of their doing their job – is a necessary incident of the relationship between employer and employee (and of principal and agent).

Since the issue as regards this part of the case is one of pure law, I will, without disrespect to the Judge’s careful analysis, proceed directly to a consideration of the Claimant’s case without summarising his reasoning.

It was Mr Ciumei’s case before us, as before the Judge, that such a term was indeed a necessary incident of the employment relationship because it gave effect to a general principle that if a person acts on the instruction of another they are entitled to be indemnified against all losses, of any kind, suffered as a result of doing so (Footnote: ⁵). In support of that proposition he cited a number of statements in the textbooks. Para. 40 of the section on Employment in vol. 39 of Halsbury’s Laws reads:

“An employer is under an implied duty to indemnify or to reimburse the employee, as the case may be, against all liabilities and losses and in respect of all expenses incurred by the employee either in consequence of obedience to his orders, or incurred by him in the execution of his authority, or in the reasonable performance of the duties of his employment.”

He also referred us to similar passages in vol. 1 of Halsbury, in the section on Agency (para. 113); Chitty on Contracts, 34^th ed. (para. 42-117); and Bowstead and Reynolds on Agency, 21st ed., article 62 (para. 7-05). The feature on which he relied in all of them is that they refer without qualification to “losses” or “losses and liabilities” (Footnote: ⁶).

General statements of that kind, however, are of limited value. The term “loss” can be used in a variety of contexts and with a variety of shades of meaning. Whether the law does indeed recognise a principle of the width asserted by Mr Ciumei can only be established by a review of the authorities from which the statements in the textbooks derive. Recognising that, Mr Ciumei took us (as he did the Judge) through a large number of cases, which I review below in the order followed in his oral submissions.

Before I embark on that exercise, it is worth spelling out the nature of the loss suffered by the Claimant in this case. It does not consist in any expense or liability incurred to a third party: rather, it is an unliquidated loss requiring assessment by the Court. The loss was, when proceedings were commenced, partly in the past but mostly extending far into the future: as regards the future element the Claimant asserts a present right to compensation. That is the kind of loss for which recovery is typically sought by a claim in damages, as indeed it is in his negligence claim against the Bank: I note how the claims for damages and under the indemnity are assimilated in para. 53 of the AMPOC and in the prayer (see paras. 18-19 above). Those features make this an unusual kind of obligation to find described as an indemnity. I accept that “indemnity”, like “loss”, is an elastic term, and there are certainly some cases in which an indemnity takes the form of a promise to compensate for an unliquidated loss (Footnote: ⁷); and in any event we are concerned with the substance of the term to be implied and not how it is labelled. Nevertheless it is as well to bear in mind the peculiar character of the claimed indemnity as we go through the authorities.

Mr Ciumei’s first authority was The Fanti [1991] 2 AC 1. In that case the rules of a P&I club provided for a member to be “protected and indemnified against … claims and expenses which he shall have become liable to pay and shall in fact have paid” for loss and damage to cargo carried on their vessel. Cargo owners whose cargo had been lost and who had not been paid by the insolvent ship owners claimed under that rule (relying on the Third Parties (Rights against Insurers) Act 1930). The issue in the appeal concerned the effect of the “pay to be paid” requirement. Mr Ciumei relied on the following passage from the speech of Lord Goff (at pp. 35-36):

“I am unable to accept [the submission of counsel for the cargo owners] that a condition of prior payment is, at common law, implicit in a contract of indemnity. I accept that, at common law, a contract of indemnity gives rise to an action for unliquidated damages, arising from the failure of the indemnifier to prevent the indemnified person from suffering damage, for example, by having to pay a third party. I also accept that, at common law, the cause of action does not (unless the contract provides otherwise) arise until the indemnified person can show actual loss: see Collinge v. Heywood (1839) 9 Ad. & E. 633. This is, as I understand it, because a promise of indemnity is simply a promise to hold the indemnified person harmless against a specified loss or expense [emphasis supplied]. On this basis, no debt can arise before the loss is suffered or the expense incurred; however, once the loss is suffered or the expense incurred, the indemnifier is in breach of contract for having failed to hold the indemnified person harmless against the relevant loss or expense.”

The indemnity with which the House was concerned in that case arose under a contract of insurance, and it may be doubtful whether the passage quoted sheds much light on the kind of contract with which we are concerned. But what Mr Ciumei attached importance to was Lord Goff’s description of a contract of indemnity – which he said extended beyond the insurance context – as a promise to “hold harmless”: see the sentence which I have italicised. But with respect I do not see how that advances the argument. The obligation of an indemnifier is indeed traditionally described in that way, but it is still necessary to determine the kind of harm to which the indemnity extends – what Lord Goff refers to as “the specified loss or expense”. In the case of The Fanti that harm was the liability to the cargo owner. That is a very different kind of harm from the loss of earnings which is the subject of the present claim.

Mr Ciumei next took us to a series of authorities concerning claims for indemnities by agents (in the wide sense of anyone acting at the request of another) and which he said supported the general principle identified at para. 111 above. The authorities were Fletcher v Harcot (1622) Hutton 56 (123 ER 1097); Adamson v Jarvis (1827) 4 Bing 66 (130 ER 693); Frixione v Tagliaferro (1856) X Moore 175 (14 ER 459); The James Seddon (1866) LR 1 A&E 62; Birmingham and District Land Company v London and North Western Railway Company (1886) 34 Ch D 261; Lord Mayor of Sheffield v Barclay [1905] AC 392; and Re Famatina Development Corporation Ltd [1914] 2 Ch 271. I need not go through all of these authorities seriatim. They can be sufficiently addressed as follows.

Most of the cases afford no more than particular examples (sometimes on rather idiosyncratic facts) of principals being held to be obliged to reimburse their agents for amounts which they had had to pay out to a third party and/or costs which they had incurred, as a result of legal proceedings arising out of the performance of their duties. That sheds no light on the issues before us. Mr Ciumei advanced some submissions depending on the precise nature of the liabilities and costs in respect of which the agent recovered in the various cases. But even when those could be ascertained (which was not always the case) the point led nowhere because in none of the cases was there any issue about the kinds of harm covered by the indemnity, and in none of them was there a claim analogous to the claim for loss of earnings in this case.

In relation to one of the cases, Adamson v Jarvis, Mr Ciumei was able to make a rather more specific point. An auctioneer claimed an indemnity from a seller for damages that he had had to pay to the buyer when it transpired that the seller did not have title to the goods (together with his costs of defending the action and associated expenses). One of the defendant’s submissions involved the proposition that there was no contract between the seller and the auctioneer. Best CJ rejected that argument, saying (see p. 73):

“Here is a contract. The plaintiff is hired by defendant to sell, which implies a warranty to indemnify against all the consequences that follow the sale [emphasis supplied].”

Mr Ciumei submitted that the word “all” showed that the indemnity covered any kind of harm that the auctioneer had suffered. I do not believe that any significance can be attached to that, for essentially the same reason: there was no issue in the case as to the extent of the seller’s liability if he was subject to the indemnity, and Best CJ cannot be understood to have been saying anything that necessarily applied to the very different kind of loss in this case.

The Birmingham and District Land Company case requires slightly fuller consideration. It concerned the validity of a third party notice, which depended on whether the party serving the notice was claiming an “indemnity” within the meaning of the relevant rule. The decision of this Court is sufficiently summarised in the headnote as follows:

“In order to bring a case within [the rule] it is not enough that if the plaintiff succeeds the defendant will have a claim for damages against the third party, but the defendant must have against the third party a direct right to indemnity as such, which right must – generally, if not always – arise from contract express or implied, and that here there was no ground for implying such a contract.”

Mr Ciumei asked us to note that the judgments in that case had been approved by the Privy Council in Eastern Shipping Company Ltd v Quah Beng Kee [1924] AC 177: see at p. 183. He relied on it for two points.

First, he referred us to the judgment of Cotton LJ, at p. 272, where he said:

“Of course, if A requests B to do a thing for him, and B in consequence of his doing that act is subject to some liability or loss, then in consequence of the request to do the act the law implies a contract by A to indemnify B from the consequence of his doing it. In that case there is not an express but an implied contract to indemnify the party for doing what he does at the request of the other.”

He submitted that the reference to A’s obligation to indemnify B “from the consequence of his doing [the act in question]” must mean all the consequences of that act; he also noted that Cotton LJ, when dealing at p. 273 with a submission based on Adamson v Jarvis, had cited the passage from the judgment of Best CJ quoted above. I do not accept that submission. No more than in Adamson was the Court required to consider the nature of the losses to which an implied indemnity might extend, and it cannot be read as expressing any considered opinion on that question.

Second, he referred to a passage from the judgment of Fry LJ at p. 276, where he expounds the difference between an indemnity and a right to damages and concludes:

“Therefore the right to … damages is not a right to indemnity, although when you come to ascertain what the measure of damages is, it may be that indemnity will properly express that measure of damages.”

Mr Ciumei’s submission was that the observation that “indemnity” might represent the measure of damages for a breach of contract showed that damages and indemnity “can be the same kind of thing”, the significance of that being (as I understood him) that in principle a person indemnified might recover any amount that he could properly recover as damages. That is reading too much into what Fry LJ was saying. He was doing no more than making the (not very surprising) point that in some cases the remedy for a breach of contract might produce an order for payment of damages in the same amount as if the defendant had promised to indemnify the claimant against the loss suffered. It does not follow that there is any general equivalence between the two kinds of obligation, which would indeed be contrary to the whole thrust of the decision. The kind of loss covered by an implied contractual indemnity depends on the limits of the term which falls to be implied, whether in fact or in law.

Mr Ciumei also referred us to Burrows v Rhodes [1899] 1 QB 816, but that was not a case of a claim on an indemnity: the plaintiff was a volunteer who had been injured while participating in the Jameson Raid and was bringing a claim for fraudulent misrepresentation against its promoters (Cecil Rhodes and Jameson himself). The judgments do mention Fletcher v Harcot and Adamson v Jarvis but only in the context of whether the claim was barred because of the unlawfulness of the enterprise. Although the authority added some historical spice to the hearing I was unable to understand how it was relevant to the present case.

Finally, Mr Ciumei referred us to two other shipping cases – The Sagona [1984] 1 Ll Rep 194 and Petroleo Brasileiro SA v ENE Kos 1 Ltd [2012] UKSC 17, [2012] 2 AC 164 (“ENE Kos”). I take them in turn.

In The Sagona the master of a vessel which was the subject of a time charter had, at the direction of the charterers, delivered the cargo to a third party who was not entitled to receive it. As a result the vessel was arrested for two months. The owners claimed against the charterers for losses of two kinds, described in the judgment as “loss of earnings whilst the vessel was under arrest” and “various expenses”, the amounts of both of which were agreed (see p. 196, col. 2). It was common ground (p. 204, col. 1) that the owners were entitled to those amounts under an implied indemnity if, but only if, the charterers had directed the master to make the mis-delivery which led to the arrest. Staughton J held that they had done so, and the claim was accordingly allowed. Mr Ciumei submitted that the element for “loss of profits” in the amount awarded was of the same character as the claim for loss of earnings in the present case. Since the amount was agreed, and there was accordingly no discussion in the judgment about its basis, the decision is of limited value as an authority; but the underlying point is better considered in the context of considering ENE Kos.

In ENE Kos (which was not relied on before the Judge) owners had withdrawn a vessel from charter for non-payment of hire but it did not become immediately available because of the need to discharge the cargo. The owners claimed payment for the service of the vessel during that interval. They relied on clause 13 of the charterparty, by which the charterers promised to indemnify them “against all consequences or liabilities that may arise … from the master … complying with charterers’ orders”. One of the issues was whether the loss claimed fell within the terms of the indemnity. The Supreme Court, by a majority, held that it did. Lord Sumption, who delivered the leading judgment for the majority, characterised the loss as being the opportunity cost to the owners of the detention of their ship and treated “the measure of the indemnity” as being the market rate of hire at the time: see para. 17 (2). Mr Ciumei submitted that the loss recovered under the indemnity (as in The Sagona) was of the same character as the Claimants’ loss of earnings claimed in the present case, and that ENE Kos was accordingly binding authority that such a loss was recoverable.

One answer to that submission might be that in that case, unlike this, the Court was concerned with an express indemnity. But it is not quite as straightforward as that, since, as Lord Sumption notes in paras. 9 and 10 of his judgment, it is well-established that there is an implied term in a contract between owner and charterer in substantially equivalent terms to clause 13 (the claim in The Sagona was indeed under that implied indemnity).

However, I would not in any event accept Mr Ciumei’s submission. Lord Sumption made it clear at paras. 10-12 of his judgment that the scope of the indemnity, despite its apparent width, was not unlimited. He points out in para. 12 that the reference in clause 13 to the “consequences” of complying with the charterers’ orders applied a causative criterion, and continues:

“Like all questions of causation, this one is sensitive to the legal context in which it arises. It depends on the intended scope of the indemnity as a matter of construction, which is necessarily informed by its purpose … The real question is whether the charterers’ order was an effective cause of the owner having to bear a risk or cost of a kind which he had not contractually agreed to bear. I use the expression effective cause in contrast to a mere but for cause which does no more than provide the occasion for some other factor unrelated to the charterers’ order to operate.”

The references in that passage to the construction of the indemnity are literally inapt in a case like the present where it is not express; but Lord Sumption’s reference to its scope being determined by reference to all relevant contextual considerations applies equally (if not indeed more strongly) to an implied indemnity. The context of a charterparty is very different from that of a contract of employment, and that by itself precludes treating the decision in ENE Kos as authoritative in the present case. But a more specific difference is that in that case the loss was caused by the action of the charterers themselves in instructing the master to discharge the cargo, whereas in the present case the loss of earnings was caused by the act of the Romanian court in convicting the Claimant. The question for us is whether the indemnity extends to a loss of that character: ENE Kos has nothing to say about that. (I return to the question of causation at paras. 143-144 below.)

The result of that regrettably prolonged trawl through the case-law is that there is in my view no support in the English authorities – either in the form of a statement of principle or by way of decision on analogous facts – for Mr Ciumei’s submission that there is a general principle, reflected in the agency and employment cases, that if a person acts on the instruction of another they are entitled to be indemnified against all losses, of any kind, suffered as a result of doing so. The Judge reached the same conclusion on the basis of a similar analysis in the Bank’s submissions before him (Footnote: ⁸): see paras. 319 and 322 of his judgment. As he put it at para. 326:

“… there is not a single case in which an indemnity implied into an agency or employment contract has permitted the agent or employee to recover lost income. Rather, in every reported case where an indemnity was ordered it was for payments that the agent or employee had made, or was liable to make, to a third party.”

The question then is whether, notwithstanding the absence of authority, the general indemnity should nevertheless as a matter of principle extend to cover all losses of any kind suffered by an employee as a result of doing their job, and more particularly a loss of earnings of the kind suffered in this case, irrespective of any fault on the part of the employer. I do not believe that it should. Specifically, I do not believe that to imply such an indemnity would be reasonable or fair or that it would balance the competing policy considerations.

My principal reason for that conclusion is that a general indemnity of the kind contended for would wholly subvert the way in which both the common law and legislation have addressed the issue of the obligations of employers. The law has developed so as to create a wide range of circumstances in which they are obliged to compensate employees for harm suffered in the course of their employment, sometimes on the basis of strict liability and sometimes on the basis of breach of a duty of care. The most obvious example is where the employee suffers personal injury at work, but there are also circumstances in which other forms of harm may be compensated. One example is where the employee suffers stigma, and consequent financial loss, as a result of the employer conducting a dishonest business: see Malik v Bank of Credit and Commerce International SA [1997] UKHL 23, [1998] AC 20. That entire structure would be redundant if the Claimant’s submission were correct.

The same point was made by the Court of Appeal of New South Wales in National Roads and Motorists’ Association v Whitlam [2007] NSWCA 81. The claim in that case was for reimbursement of legal costs incurred by an employee in bringing defamation proceedings against a broadcaster arising out of statements made about him in connection with his employment. He based his claim in part on what his counsel claimed was “a general principle of law that a person acting on behalf of another is entitled to be indemnified for loss that they may suffer as a result of so acting”: see para. 83 of the judgment of Campbell JA. Reliance was placed on, among other things, article 64 in the then current edition of Bowstead and Reynolds on Agency, which read:

“Every agent has a right against his principal to be reimbursed all expenses and to be indemnified against all losses and liabilities incurred by him in the execution of his authority.”

Para. 94 of the judgment of Campbell JA reads:

“If the general principle stated by Bowstead & Reynolds were applied so that ‘losses’ included losses of types that can be compensable by action in tort at the suit of the person who suffered the loss, the civil law would be very different to what it in fact is. If, for instance, ‘losses’ included the type of damage that is remediable by an action seeking damages for personal injury, the mere fact that A had requested B to do a task, and B was injured in the course of performing it, would mean that B was entitled to be indemnified by A for the injury he had suffered.  Any such entitlement would sweep aside those aspects of the law of tort that require there to have been a recognised tort committed by A before B is entitled to be compensated by A for his injury. It would mean that, in the paradigm case in which worker’s compensation payments are made, where a worker in the course of carrying on his duties is injured, the worker would have had a right of indemnity under the general law from his employer just because the employer had requested him to do the task in the course of which he was injured, quite independently of any obligation created by the worker’s compensation legislation, and the indemnity would be to provide a full indemnity, not merely the limited scale of benefits conferred by workers’ compensation legislation.  I do not believe that a general principle of law that alters the civil law in such radical ways, exists but has hitherto gone unrecognised.”

(The reference to “worker’s compensation payments” is to the New South Wales statutory scheme of no-fault compensation for injuries at work; but Campbell JA’s point applies to tort liability generally.)

The Judge found that passage persuasive. At para. 327 of his judgment he said:

“The decision in Whitlam is to the effect that if an employee could recover for such losses pursuant to an implied indemnity, the law books in relation to the existence of a duty of care on the part of the employer would have to be rewritten. Very large parts of the law of employer’s liability could be re-written and simplified as the implied indemnity would eclipse the learning on the law of negligence and breach of statutory duty. The law reports are replete with cases where complicated arguments were run including in the appellate courts which would have been unnecessary in the event that the implied indemnity had existed in the form contended for in this case. The employee could say that there was a strict liability because of the scope of the indemnity, obviating the need for the other duty contended for.”

He went on to illustrate the point by referring, at paras. 328 and 329, to the decisions of this Court in Reid v Rush & Tompkins Group plc [1990] 1 WLR 212 and Greenway v Johnson Matthey plc [2016] EWCA Civ 408, [2016] 1 WLR 4487. I agree with him that those cases are good illustrations of the point made in Whitlam, and I should briefly address them.

In Reid v Rush & Tompkins the claimant was employed by the defendant in Ethiopia. He was injured in a road accident there while driving his employer’s vehicle. He was not at fault but there was no realistic possibility of recovery from the other driver. He had no insurance cover. He claimed that it was an implied term of his contract that the employer would arrange insurance cover for him in Ethiopia, or at least advise him to do so himself; and that, even if it was not, such a duty arose in tort. The claim failed. Ralph Gibson LJ, in a careful analysis, rejected the contractual claim on the basis that there was binding authority that it was not an implied term of a contract of employment that the employer would take care “to protect his servant from economic loss, caused through the wrongdoing of a person for whom the employer is not responsible” (see p. 231 B-C); and he went on to hold that it would be wrong to impose a higher duty in tort (p. 232 F-G). As the Judge observed:

“If it had been the case that the implied indemnity applied to all injuries and losses of earnings arising from an incident in the course of employment, then the complicated arguments in Reid could have been avoided and a simple route to recovery established through the implied indemnity.”

In Greenway the claimant employees had in consequence of their work developed a condition known as platinum sensitivity, as a result of which they were either redeployed or dismissed and suffered (at least potentially) a loss of earnings. It was held in this Court that platinum sensitivity did not constitute a physical injury and that there was no implied term of the contract (and thus also no duty in tort) to take care to protect the employees from loss of the kind claimed. Sales LJ, with whom the other members of the Court agreed, noted at para. 37 of his judgment that:

“The classic formulation of the duty owed by an employer to an employee is focused on protection of the employee from physical injury, not protection from economic harm (albeit if there is physical injury then damages may be recovered for consequential loss of earnings), and this is true both in contract and tort … .”

He held at para. 45 that in the absence of a physical injury “it could [not] properly be said that it is fair, just or reasonable to extend the obligation of the employer to cover this type of loss in this case”. Again, the Judge’s point is not that Sales LJ’s reasoning was directly applicable in this case, but rather that it was wholly unnecessary if the employees could have relied on a general indemnity in the terms contended for by the Claimant. (Footnote: ⁹)

The impossibility of reconciling an indemnity of the scope for which the Claimant contends with the structure of the existing law is a sufficient reason for rejecting his case on the general indemnity. But I should also say that I think that that structure strikes a fair balance between the parties’ interests, for the following reasons.

The law provides for fault-based liability on the part of the employer – typically, liability for breaches of duty of care – in a wide range of circumstances; and typically in such cases (subject to the rules about remoteness and causation) the employee can claim compensation for loss of earnings suffered as a result of the breach, including, where the facts support such a claim, on a career-long basis. It also, as we have seen, provides for a right of reimbursement for expenses and liabilities incurred by the employee in the course of their work.

By contrast, the effect of the indemnity for which the Claimant contends is that the employer should be liable to compensate employees for a loss of earnings even where that loss is caused by someone else and there has been no fault on the employer’s part, simply because the relevant harm was done “in consequence of” them doing their job. That means that the employer would be liable in an extraordinarily wide range of circumstances. An obvious example is where the loss of earnings is consequent on personal injury – the employee driving for work who is injured by the negligence of another driver; or who visits someone else’s office for a work meeting and trips on a defective stair; or whose work brings them into contact with a carrier of an infectious disease and they become seriously ill. However, as the present case illustrates, the indemnity would also operate where there was no physical injury. An employee’s reputation and career prospects might be injured by an association with a client of their employer who (unforeseeably) turned out to be dishonest; or they might be libelled in a work-related context; or they might lose their job, and suffer long-term loss of earnings, because the employer’s customer unreasonably takes against them and they cannot be moved to other work. In some, though not all, of these examples the employee might have a claim against the third party (who might or might not be solvent or insured), but that is immaterial: if the Claimant is right he or she has in all cases a straightforward right to be compensated by the employer irrespective of whether there is a claim against anyone else.

In all such cases the employee has suffered a serious misfortune – sometimes an injustice – at the hands of a third party, as the Claimant says that he has at the hands of the Romanian Court. But I do not believe that it follows that the employer should be liable simply because that occurred in consequence of their doing their job. In part that is because as a matter of principle the Court is wary of imposing liability in the absence of fault. As Sales LJ says at para. 56 of his judgment in Greenway:

“The law does not furnish a remedy for every harm suffered by an individual, and in particular does not do so where the infliction of the harm in question does not constitute a ‘wrong’ in the contemplation of the law: see D v East Berkshire Community Health NHS Trust [2005] 2 AC 373, at [100] per Lord Rodger of Earlsferry.”

But there are also practical considerations. As the Judge says at para. 336 of his judgment:

“An imposition of strict liability through the implied indemnity would impose a huge burden on employers including employers in a very different financial position from the Bank in the instant case. It would have to be seen how, if at all, and at what cost, they would be able to insure against such a liability.”

I do not lose sight of the particular nature of the relationship of employer and employee, where there is typically a fundamental disparity in the economic positions of the parties. But that is properly addressed by the imposition of appropriate duties of care and by employment protection legislation. It does not justify making the employer, in effect, the employee’s insurer in respect of all harm suffered in consequence of his work.

Mr Ciumei submitted that the scope of the indemnity was not as wide as it might at first appear because it would only apply in cases where the employment was the effective cause of the loss rather than merely the occasion for it. He said that the position as regards the implied indemnity was the same as that propounded by Lord Sumption in relation to the express indemnity in ENE Kos: see para. 127 above.

I do not believe that that avoids the problem. I note by way of preliminary that in a case where the loss to the employee is caused by the act of a third party, without any fault on the part of the employer, it is at least arguable that the employment is no more than the context for the third party’s act rather than an effective cause; if that is right, the Claimant’s case in these proceedings would fail for that reason. But that was not argued before us, and I am content to proceed on the basis that Mr Ciumei’s point would have a potential application in cases of this kind – that is, that it would have to be decided in any given case whether the connection between the third party’s act and the employee’s work was sufficiently close for the work to be regarded as an effective cause of the loss. I am bound to say that the need to perform that exercise in each case seems to me to make the application of the indemnity undesirably uncertain. But in any event in any case where the requisite causal connection was found the fundamental objection which I have identified above to the extension of no-fault liability still remains.

Mr Ciumei also submitted that a degree of overlap between the rights enjoyed under the indemnity and those recognised at common law (or under statute) was unobjectionable in principle: he referred to Spring v Guardian Assurance, where the House of Lords held that it was not an obstacle to the imposition of a duty of care (in tort or contract) on a person giving a reference for a former employee that it might give wider rights than an action in defamation because of the absence of a defence of qualified privilege. But the two cases are not parallel. The advantageous position of the employee derives from the employment relationship, which is a feature not present in cases of defamation generally.

The Claimant’s challenge to the Judge’s conclusion on the issue of implication constitutes his ground 4. I have addressed Mr Ciumei’s arguments in support of that ground, and for the reasons given I would dismiss it.

The Judge addressed the question whether the indemnity pleaded in para. 19.1 should be implied as a matter of fact – i.e. because of the particular circumstances of the case – in paras. 359-365 of his judgment. He referred at para. 361 to the Claimant’s response to a Request for Further Information setting out the facts and matters said to give rise to the implication. I need not quote the response in full but, consistently with para. 25 of the AMPOC, it focused on the facts that it was envisaged that the Claimant “would be working in a high-risk country, namely Romania and/or in high-risk transactions, namely the privatisation of significant state-owned companies” and that the risks in question included the risk of being “subjected to politically or commercially motivated judicial or criminal action including but not limited to arbitrary detention, criminal charges, prosecution, conviction and imprisonment or other corrupt action or interference with an individual's fundamental rights and freedoms”. The Judge continued, at para. 362:

“At the heart of these points is a case that the risk profile of Mr Benyatov in Romania as regards the country and/or the transactions in which he was involved was high-risk, such that there should be a broader indemnity in fact than the indemnity in law.”

He pointed out that his earlier findings in relation to the negligence claim meant that none of the risk factors relied on had been established and that it followed that “the substratum of the alleged indemnity is not made out.”

Ground 5 of the grounds of appeal reads:

“The Judge was wrong in law in failing to find that there was an indemnity implied in fact [359]-[365]. He applied the wrong analysis to the question and simply relied on his findings in relation to the duty of care without any regard to the different considerations that are necessary to the question whether an indemnity was to be implied as a matter of fact (as opposed to as a matter of law).”

As developed by Mr Ciumei in his oral submissions, the point is that by treating his findings on the duty of care issue as dispositive the Judge failed to answer the questions required in a case of implication as a matter of fact, most recently reviewed in Marks and Spencer plc v BNP Paribas Securities Services Trust Company (Jersey) Ltd [2015] UKSC 72, [2016] AC 742. I will refer to the test expounded by Lord Neuberger in that case as “the necessity test” (though that is of course an inadequate shorthand). More specifically, Mr Ciumei submitted that the existence and terms of the indemnity had to be judged at the date that the parties entered into the contract. At that date it was not known exactly where the Claimant might be working, but it was known that it could include unquestionably high-risk countries such as Russia. Accordingly the Judge had to consider what term needed to be implied at that point, applying the necessity test. It was the Claimant’s case, on the basis that he might have to work on transactions involving Russia, that it was necessary to imply a term of the kind pleaded in para. 19.1. By focusing only on his findings about the risks in Romania the Judge failed to address that case.

I do not accept that the Judge made any such error. He was clearly prepared to accept that it might be necessary to imply an indemnity against losses incurred by the Claimant as a result of working on high-risk transactions and/or with high-risk countries (for short, “high-risk work”), of which Russia may well have been an example. The question for him was whether the Claimant’s work in Romania was high-risk, as pleaded both in para. 25 of the AMPOC and in the Further Information. His findings on the negligence claim were that it was not. It necessarily followed that the Claimant was not indemnified against the losses in respect of which he claimed. In so far as it is being argued that, once it was contemplated that the Claimant might be asked to do some high-risk work (e.g. in Russia), it was necessary to imply an indemnity in the full terms of para. 19.1, i.e. covering all losses consequent on all other aspects of his work, that is obviously wrong: such an indemnity would not satisfy the necessity test.

At para. 56.1 of his skeleton argument Mr Ciumei made a further point. He submitted that the risk of conviction was one which the Claimant “did not and would not have undertaken as a normal part of his employment”. In that connection he referred to para. 54, where he had said:

“The question is a simple contractual one: Who bears the risk that Mr Benyatov’s work for the Defendant – done in accordance with international banking standards (Judgment [373]) – might result directly in his conviction and the destruction of his career and earning potential? Self-evidently, the Defendant. It stood to make a significant profit from his work. Clearly that was a risk that he did not undertake to be responsible for.”

I do not believe that asking whether the Claimant “undertook” the risk of conviction in Romania, and the consequent financial loss, advances the argument. Clearly neither party contemplated that risk. The question for the Court is whether it is necessary to imply a term that if such a risk nevertheless eventuated the Bank would compensate the Claimant for that loss. I do not think that that question can be answered simply on the basis that the Bank stood to make a profit from the transaction (as indeed did the Claimant, by way of bonus) or, which I take to be the underlying point, because it had the deeper pockets. The test is one of necessity. If an officious bystander had intervened during the negotiation of the Claimant’s contract of employment and suggested that an express term should be included that if (without any fault on the part of the Bank) the Claimant was wrongly convicted in Romania in consequence of his work and lost his regulated status the Bank would pay him a sum representing his career-long loss of earnings, he would certainly not have been suppressed by both sides with a testy “of course, that goes without saying”.

I would dismiss ground 5.

I have put para. 19.2 under a separate heading, but in truth it raises no further issues, and I deal with it here only for completeness. It was added by amendment. It is confined to the situation where the losses are incurred as a result of an “unlawful enterprise upon which [the Claimant] was required to embark without knowledge that it was unlawful”. That language derives from the speech of Lord Tucker in Lister v Romford Ice and Cold Storage Ltd [1956] AC 555, where he says, at p. 595:

“It has always been an implied term that the master will indemnify the servant from liability arising out of an unlawful enterprise upon which he has been required to embark without knowing that it was unlawful.”

As the Judge points out at paras. 345-348 of his judgment, Lord Tucker was not propounding a distinct kind of indemnity in addition to the uncontentious implied obligation of an employer to indemnify the employee against expenses and liabilities incurred in the course of the employment. Rather, he was simply making the point that if the expenses/liabilities were incurred in the course of an unlawful enterprise that did not deprive the employee of the benefit of the indemnity provided that he or she was unaware of the unlawfulness. It follows that the scope of the indemnity in an “unlawful enterprise” case is no wider than in the ordinary case, and if the claim under para. 19.1 fails, so also must any claim under para. 19.2. The Judge made the further point that it was hard to see how, on his own case, the Claimant could assert that he was engaged in an unlawful enterprise, but it is unnecessary for me to consider that.

I would dismiss this appeal.

I agree.

I also agree.