Markel International Insurance Company Ltd v Higgins

This appeal arises out of a finding at trial of dishonesty against an underwriting agent in his early seventies who was then suffering from Alzheimer’s disease and had probably been suffering from it to some degree, possibly at a pre-clinical stage, at the time, two or three years earlier, when the events which led to that finding were taking place. The essential issue raised on this appeal is whether the judge, Teare J, was wrong in his conclusion that Alzheimer’s was not an explanation, either of the agent’s actions for which he had to render an account or of difficulties in his evidence at trial, such as should have led the judge to reject what were otherwise powerful allegations of dishonesty.

The essence of that dishonesty, as found by the judge, was that the appellant had conspired with others to defraud two insurance companies from whom binders had been obtained, by entering into contracts on their behalf for sums in excess of authorised limits, so retaining excess premium for themselves. The period of this conspiracy was during the currency of these two binders, namely from approximately January 2005 to September 2006.

The sole appellant is Timothy Higgins, who was born on 16 December 1936 and turned 70 just after this litigation commenced. He had begun working in the insurance industry in 1958. In 1981 he started his own business, College Hill Underwriters Limited, an insurance underwriting agency specialising in the writing of surety bonds for the construction industry. Surety bonds are undertakings issued by the surety, usually an insurance company or a bank, to pay the beneficiary a sum of money (up to a stated limit) in certain events, typically the failure of the party requesting the issue of the bond to meet his contractual obligations to his contract partner, the bond’s beneficiary. The risk implicit in the bond needs to be rated by a skilled underwriter, and the party requesting the issue of the bond pays a premium to the surety as the price of the bond. Mr Higgins became a leading figure in a sector of the surety bond market, namely the writing of bonds for small to medium sized construction companies in the UK and Ireland. His company, College Hill, was successful and in 2003 he sold it to a major European group.

In the following year Mr Higgins formed a new company in collaboration with colleagues who in turn became co-defendants in the present litigation. The company was Surety Guarantee Consultants Limited (“SGC”), which was incorporated on 6 February 2004. Those colleagues included Barry Williams, Clifford Felstead and Ralph Brunswick. Mr Higgins and Mr Williams became directors and Mr Felstead became an employee of SGC. Mr Higgins was nominally and legally the owner of 100% of the share capital, but, although the precise details are not clear cut, it was not disputed on this appeal that under declarations of trust or other less formal arrangements Mr Higgins held his shares in part on behalf of Mr Brunswick and Mr Felstead (but not Mr Williams). Thus on 14 October 2004 Mr Higgins had entered into a declaration of trust in favour of Mr Brunswick to hold 66 of his 100 shares for him, albeit another document refers to Mr Brunswick’s interest as the 34 shares rather than the 66. There is also another document entitled “Declaration of Trust”, but this was unsigned, which set out an intention that Mr Felstead was to be given a beneficial shareholding in SGC. In his witness statement Mr Felstead said that he had been promised some shares by Mr Higgins, but that he had never received any.

Mr Brunswick was chairman, chief executive officer and a director of Templeton Insurance Limited (“Templeton”), a small insurance company incorporated in the Isle of Man. Templeton had acted in the past as one of the issuers of bonds which Mr Higgins had underwritten at College Hill. As a small company, however, its paper was not acceptable to all purchasers and beneficiaries of bonds. I will have more to say about Templeton below, but the essence of the matter is that SGC had represented to the claimant insurance companies who had issued surety bonds under binders to SGC that those bonds had been written within the binders’ stipulated limits, whereas they had in fact been written in excess of those limits and fictitious Templeton bonds had been referred to in bordereaux or put on file to account for the difference between those stipulated limits and the total amount underwritten. Mr Brunswick was a defendant at trial and found to have conspired with Mr Higgins and Mr Felstead to defraud the claimants. He did so in part by signing the bogus Templeton bonds just referred to. Having run out of funds to afford representation, he did not attend the trial until its fourth week, when he appeared to give evidence and be cross-examined and thereafter stayed to represent himself. He has not appealed. His participation in a conspiracy to defraud the two companies which had given binders to SGC is thus established.

Mr Higgins was vague about how he had come to collaborate with Mr Brunswick in relation to the business of SGC. In his first witness statement he said that he had always been suspicious about Mr Brunswick “needing to base his business in the Isle of Man”, and had warned his own son, who was also in the insurance world, against dealing with him. He was unable to say how or why he had disregarded his own advice and gone into business with him in 2004. Even so, back in December 2002, Mr Higgins had deposited €500,000 with Templeton, and in March 2004 he had asked Mr Brunswick to retransfer €300,000 as working capital for SGC (and another agency set up by Mr Higgins at about the same time, conducting general insurance broking work).

Mr Felstead had a conviction for insurance fraud, for which he had been sentenced to a term of three years imprisonment. He came out of prison in December 1997 and obtained employment back in the insurance world. Since 1993 he had worked in the field of surety and bond insurance. Mr Higgins was equally vague about how he came to have collaborated with Mr Felstead. He said that he had known about his conviction (and thought everyone else knew about it too). He said that he had also warned his son against doing business with Mr Felstead, and equally could not say how he had come to disregard his own advice about Mr Felstead. Mr Higgins’ case was that while he, Mr Higgins, had been used for his rating and underwriting skills, it had been Mr Felstead who had been the real engine of SGC, at any rate so far as paper-work and general administration were concerned: he had control over everything and “was essentially running it”, as well as participating in the rating and underwriting, and liaising with Mr Brunswick. The judge accepted this picture of Mr Felstead’s role, which was supported by other evidence (while not accepting the corollary that Mr Higgins was Mr Felstead’s dupe).

As early as February 2005 SGC’s compliance officer, a Mr Brady, took advice from Clyde & Co, solicitors, as to whether the FSA would approve of Mr Felstead, given his conviction, carrying out “controlled functions”, and was told that the FSA would not. That does not appear to have affected Mr Felstead’s role within SGC. Later, in October 2005, one of the two companies which had entered into binders with SGC heard a rumour about Mr Felstead’s conviction and asked Mr Felstead if he had ever been convicted of insurance fraud. He answered “no, never”. Nevertheless, shortly thereafter the company discovered the truth and raised the matter with Mr Higgins, who confirmed the true position. On the next day, 15 November 2005, Mr Higgins wrote to the company that Mr Felstead would “leave SGC with immediate effect” and would have “no further involvement with issuance of bonds and/or administration of our bond account”. However, Mr Felstead not only remained physically in SGC’s office but continued to be involved in SGC’s surety bond business, even if no longer as an employee. In August 2006 Mr Higgins wrote that Mr Felstead had remained as a consultant.

Mr Felstead was not represented at trial and did not appear: he explained in a letter to the judge that he was on benefits and could not afford the cost of the rail journey; and a doctor’s letter said that he had a heart condition and should avoid the daily attendance at court. The judge found that he had also participated in the conspiracy to defraud. There is an outstanding application by Mr Felstead for permission to appeal, which has been adjourned by this court to a time after the determination of Mr Higgins’ appeal. Mr Felstead has two grounds of appeal. His primary ground is that, as a litigant in person without representation in a complex case, he did not receive a fair trial, so that there has been a breach of article 6 of the ECHR. His secondary ground, in the light of the appeals that had been granted in the case of Mr Higgins and also Mr Williams, was that he should be entitled to take the benefit of anything helpful which might emerge from those appeals. Although Mr Felstead has that outstanding application for permission to appeal, Mr Higgins’ appeal has proceeded on the basis, as it has to, that the judge’s findings against Mr Felstead stand. In any event, it was part of Mr Higgins’ defence at trial and it remains so on appeal, that Mr Felstead was the real villain and deceiver in the business, and that he, Mr Higgins, had been among the deceived.

Messrs Higgins, Brunswick and Felstead, who in the case of the first two participated beneficially in the shares of SGC and in the case of the third had been promised a beneficial interest, had something further in common: in that all three were intended to share beneficially in a company incorporated in the British Virgin Islands and registered in Jersey, known as General Commercial Limited (“GCL”). The documents reveal that two payments were made by SGC to GCL, the first on 22 March 2006 for £200,000 and the second dated 23 May 2006 for £288,000. The first transfer was ordered by Mr Higgins himself. Although nominally all the shares in GCL were owned by Mr Brunswick, Mr Higgins accepted that he had a beneficial interest in GCL. Moreover, the back of the document which transferred the £200,000 payment had written on it by Mr Higgins:

“Trust document needs sorting out.

Ralph [Brunswick] 40%

Tim [Higgins] 40%

Cliff [Felstead] 20%”

Mr Higgins accepted in cross-examination that the monies received by GCL were to be split between himself, Mr Brunswick and Mr Felstead. The proportions of that split are indicated by Mr Higgins’ manuscript.

What were these transfers about? A service agreement existed between SGC and GCL pursuant to which GCL was to receive 30% of SGC’s net commission for the introduction of business by GCL to SGC. However, it is not disputed that this was a sham: as the judge said, there is no evidence that GCL introduced any business to SGC. Moreover, there is undisputed evidence that the two transfers did not represent some proportion of SGC’s commission income, but rather represented and derived from premium income received by SGC which, if SGC had been an honest business, ought to have been accounted for either to Templeton or to the two companies which had given SGC its binders. Those facts are not in dispute, but it is submitted that Mr Higgins, in his illness, may have thought that this was SGC’s legitimate commission which could be conveniently parked in GCL.

Mr Williams, whom I have briefly mentioned in para 4 above, was another defendant at trial. He was also a director of SGC. However, there was no evidence that he participated (or had been promised any participation) in either SGC or GCL. The judge acquitted him of being a party to the conspiracy to defraud. However, the judge found that he had dishonestly assisted in SGC’s breach of trust as the fiduciary of the two companies from which it held its binders: he had closed his eyes to the suspicion of wrong-doing at SGC and had been too easily fobbed off by the explanations of others, including those of Mr Higgins himself. Mr Williams obtained permission to appeal from the judge’s findings of liability against him, and his appeal was due to come on together with that of Mr Higgins. However, a few days before the hearing, that appeal was settled on terms unknown save that it involved the withdrawal and thus dismissal of Mr Williams’ appeal with no order for payment of any costs. The judge’s findings of dishonesty against Mr Williams therefore survive.

In this appeal, ie Mr Higgins’ appeal, there are, save for limited exceptions, no grounds which challenge the judge’s detailed findings of fact. Thus it is not disputed that a fraud was perpetrated on the two companies from which SGC had its binders, nor that Mr Brunswick and Mr Felstead played the roles which the judge found that they had played in the overall conspiracy. Nor is it contended that Mr Higgins did not do what the judge found that he did (with an exception relating to the creation of bordereaux to which I will come below). What rather is said is that the judge erred in his assessment of Mr Higgins’ essential honesty, and that he did so because he failed to make proper allowance for the disability that Mr Higgins was under, both at the time of the events in question and at the time of trial when preparing and giving his evidence, as someone suffering from the effects of Alzheimer’s. It was common ground on the medical evidence at trial that Mr Higgins had been and was so suffering. However, that same medical evidence left it to the judge to decide, on all the material in the case, whether Mr Higgins was honest or not, in the light of his disability. Did the judge err in his ultimate assessment that Mr Higgins was dishonest, and that his actions and evidence were not to be excused on the ground of Alzheimer’s? That is the issue on this appeal. As such, it is an issue almost exclusively on fact and assessment, albeit in a context in which the great majority of the judge’s findings are not disputed.

The two companies which gave to SGC their binders and became claimants in this litigation, and are now respondents, are respectively Markel International Insurance Company Ltd (“Markel”) and QBE Insurance (Europe) Limited (“QBE”).

Markel granted SGC a Binding Authority commencing as from 1 January 2005 “to bind surety bonds”. It was signed on 12/13 January 2005 by Mr Williams on behalf of SGC and by Mr Peter Smith on behalf of Markel. Mr Smith was the specialty underwriter at Markel. SGC had been incorporated on 6 February 2004. On 30 June 2004 Mr Higgins had met Mr Smith. A note of this meeting speaks to a suggestion that Markel might participate in the surety bond market to deal with the difficulty of some of SGC’s clients not wanting to accept the security of Templeton paper. On 20 October 2004 a “Proposed Structure” envisaged limits of £2m any one bond and £5m any one obligor (the client who requested the bond), but in the event Markel was only prepared to go ahead on the basis of £1m any one bond and £2.5m any one obligor. The binder’s schedule particularised these and other limits: thus the maximum period for any bond was to be 3 years, and the principals had to be domiciled in the UK and Ireland. Section 22 of the binder prohibited “joint surety bonds”, ie bonds under which both Markel and another insurer were both liable. This was contrary to a previous suggestion made during the negotiations. However, a “modus operandi” side-agreement made at the same time explained how fairness would be maintained between SGC’s existing relationship with Templeton and its new relationship with Markel. Thus –

“Any business introduced by Tim Higgins that accepts Templeton security will be split 50/50 between the Templeton and Markel facilities. Markel will only be able to accept the entire Tim Higgins order in instances where Templeton paper is not acceptable to the original client.”

The judge observed that the “modus operandi” agreement did not purport to be an amendment to the binder but to describe how it would be operated. It contemplated separate bonds, and several liability to be shared between the two insurers, in circumstances where Templeton paper was acceptable to the client. It was only where Templeton paper was not acceptable to the client that Markel could be preferred as the sole insurer. In the course of cross-examination Mr Higgins accepted that that was its meaning and effect. Thus in a note of a meeting dated 23 February 2005 Mr Higgins is recorded as confirming that bonds will be issued “on a several basis not joint and several”.

The modus operandi agreement had originated in a meeting dated 2 December 2004 at which the £1m/£2.5m limits had been agreed. The relationship with Templeton had been discussed, and it had been noted that Mr Smith would draft a letter to the effect that “Anything that qualifies for Templeton paper will be done 50/50”. The following is also recorded in the note of this meeting:

“[Mr Smith] asked about rumours that [Mr Higgins] was purely a marketing man. TH confirmed he is very much the underwriter and always has been. He has never been involved in pure marketing…[Mr Higgins] advised [Mr Smith] that he will be writing marketing material that will state he can look at business with a limit up to £10m. This is purely to attract interest.”

Despite the limits under the binder, SGC from the outset bound bonds beyond the prescribed limits. The judge listed at Annex 1 to his judgment a number of such bonds (selected out of a much greater number about which Markel made complaint) which were focused on at trial as having exceeded the Markel binder’s limits of £1m per bond or £2.5m per principal (or equivalent limits which were agreed for Euro and US Dollar currency bonds). There were 18 bonds in the first category and further bonds in the second category. Some bonds hugely exceeded these limits. For instance 3 bonds were bound for PC Harrington for £2m, £3m and £5m respectively. A further bond was bound for Interhealth for £10m. I will refer further to the Interhealth bond below. Nevertheless, the monthly bordereaux produced for Markel did not show any of these bonds to exceed £1m (save for two only marginally in excess of the limit).

On 10 August 2005 Mr Smith, who was the underwriting link at Markel with whom Mr Higgins dealt, resigned from Markel and joined Amalfi Underwriting Limited, a managing general agency established in that year to act for underwriters who wrote speciality lines of business (“Amalfi”). Amalfi was looking to involve QBE in its plans for Mr Smith. So it was that on 22 and 23 September 2005 a Management Agreement was entered into between Amalfi and SGC and a Binding Authority was entered into between QBE and Amalfi, each to commence on 1 October 2005. The QBE binder authorised Amalfi to bind surety bonds for QBE and the Management Agreement authorised SGC to submit surety bonds for approval to Amalfi. On 24 October 2005 an addendum to the Management Agreement was agreed which entitled QBE to exercise Amalfi’s rights under it. This addendum came to be known as the “cut through” agreement. It is therefore unnecessary to distinguish between Amalfi and QBE and I shall refer to the second binder as if it had been agreed directly between QBE and SGC. In the circumstances the Markel binder was terminated on 1 November 2005 on 30 days’ notice. Thus Mr Smith’s move from Markel to Amalfi was instrumental in the Markel binder giving way to the QBE binder.

The QBE binder contained the same limits as the Markel binder, namely £1m per bond and £2.5m per principal. One difference was, however, that initially the maximum period of a bond was expressed as 5 years (instead of Markel’s 3 year limit), but that 5 year limit was later reduced to 3 years by amendment dated 24 November 2005. There was again a provision against any joint surety bonds. On 16 November 2005 an addendum to the QBE binder was agreed between Amalfi and SGC (albeit not disclosed to QBE) which was to the same effect as the Markel “modus operandi” agreement so far as the relationship with Templeton was concerned.

In late October 2005 QBE received information that an employee of SGC had a conviction for fraud. Mr Smith was asked to investigate. He sent an email to both Mr Higgins and Mr Felstead, asking them, “have you ever been convicted of insurance fraud?” Each replied “No, never”, but Mr Higgins did not say that Mr Felstead had been convicted. QBE subsequently learned of Mr Felstead’s conviction. At a meeting on 14 November 2005 Mr Higgins confirmed that QBE’s information was true. On the following day Mr Higgins wrote to the chairman of Amalfi that Mr Felstead would “leave SGC with immediate effect” and would have “no further involvement with the issuance of bonds and/or the administration of our bond account”. Nevertheless Mr Felstead remained physically in the office (being employed in Mr Higgins’ other agency company which shared offices with SGC) and continued to be intimately involved in SGC’s surety bond business. Subsequently he was involved in the production of bogus Templeton bonds and their use in SGC’s files. In a letter dated 18 August 2006 Mr Higgins was subsequently to say that Mr Felstead had remained as a consultant to SGC.

On 24 November 2005 Mr Smith sent an email to Mr Williams to circulate to everyone at SGC to underline the limits to underwriting under the QBE binder. Mr Williams confirmed that he had passed the email on, among others to Mr Higgins. The email began (point 1) with the £1m limit for individual bonds. The email ended with the request to ensure “full compliance”.

On 30 November 2005 Mr Smith informed SGC that he wanted to introduce a number of compliance measures, to be contained in a “full risk review sheet for each and every file/bond”. This would state inter alia bond value, current principal aggregate before and after bond issue, and would have to be signed by Mr Higgins. The review sheets allowed for signature by Mr Smith where he approved a bond for a period in excess of 3 years.

Both before and despite these precautions, SGC continued to bind bonds for values in excess of the agreed limits. The judge again set out the bonds on which attention had been focused at trial, in Annex 2 to his judgment. There were 19 bonds in excess of £1m and others which exceeded the £2.5m limit per principal. The out-of-limit bonds started on 1 October 2005 with two bonds for Gala each for £4.24m, and continued on 25 April 2006 with 2 bonds for Canadian Shipping each for €5.17m. A bond for Ocon dated 21 April 2006 was for £4.8m, and one for PC Harrington dated 14 August 2006 was for £4.1m. The monthly bordereaux presented to Mr Smith, however, consistently showed such bonds to be within limit. Thus the Gala bonds were shown with a value of £1m, and the Canadian shipping bonds were shown with a value of only €1,035,294.

In December 2005 Markel requested an audit of the risks which it had underwritten. To anticipate the audit Mr Brunswick (of Templeton) signed bogus Templeton bonds in favour of the beneficiaries of many of the Markel bonds listed in the judge’s Annex 1 for a sum equal to the difference between the value of the Markel bonds and the agreed limit of Markel’s liability under its binder. At the same time copies of the out-of-limit Markel bonds were replaced in the bond files of SGC with new bonds (the so-called “Markel dummy bonds”) for sums which did not exceed Markel’s limits. Debit notes referring solely to Markel were also replaced by debit notes making reference to Templeton. Premium advice notes addressed to Templeton were also placed on the file. Neither the bogus Templeton bonds nor the dummy Markel bonds were delivered to the beneficiaries.

When the Markel audit took place, the auditor reported “no significant issues”.

At about the same time Mr Brunswick likewise signed bogus Templeton bonds in favour of certain of the beneficiaries of the QBE bonds for the difference between the value of those bonds and the QBE binder limits. These bonds were similarly not delivered to the beneficiaries.

In May 2006 a further Markel audit did not discover anything amiss. In June 2006 QBE commenced an audit. The fake bonds were not discovered, but the audit was regarded as unsatisfactory because of the poor state of SGC’s records. This led in turn to the discovery that the Canadian Shipping bonds had been written in breach of the binder limits and to the termination of the QBE binder by letter dated 21 August 2006 with effect from 20 September 2006.

On 16 November 2006 QBE, and on 8 December 2006 Markel commenced these proceedings, each by separate action, against SGC and GCL, and Messrs Higgins, Felstead, Brunswick and Williams. The claimants variously alleged conspiracy to defraud, breach of fiduciary duty, dishonest assistance in breach of trust, and procurement of breach of contract by SGC. The two actions were tried together by Teare J. He found that Messrs Higgins, Felstead and Brunswick had together conspired to defraud Markel and QBE. He found that Mr Williams had not been party to that conspiracy, but had nevertheless been dishonest in his reckless failure to distance himself from the consequences of that conspiracy and thus in dishonestly assisting in a breach of fiduciary duty. The trial on liability lasted 20 days. The judge directed quantum proceedings to take place, which have now been completed.

Mr Brunswick was the chairman, chief executive officer and a director of Templeton, but that did not stop him playing an important role in a conspiracy to defraud Markel and QBE, with the assistance of SGC, in which he held a beneficial shareholding interest. Nor did it stop him attempting to defraud Templeton itself. Although he signed the bogus Templeton bonds in December 2005 and/or January 2006, in a pretence that Templeton had shared in the insurance of those bonds, nevertheless hardly any premium was paid to Templeton by SGC in respect of those bonds. On the contrary, as will be seen below, the secret profits earned by SGC and/or diverted by SGC from their proper destination, whether that be Markel, QBE or Templeton, to GCL, remained with SGC and/or GCL.

On 17 May 2006, the same day that SGC requested Mr Brunswick to sign more bonds, he resigned from Templeton. Thereafter no further bonds were issued by Templeton.

The judge made the following findings about Mr Brunswick and Templeton. First that he had a beneficial interest both in SGC and GCL. Secondly, that his interest in SGC was a secret one, disguised because his shareholding was hidden behind the legal 100% shareholding of Mr Higgins himself. (There was no good reason why the chairman, chief executive officer and a director of an insurer, Templeton, should be a secret shareholder in an agency company, SGC, which had an authority from Templeton to bind it to the underwriting of surety bond business. Mr Brunswick denied this interest in SGC, but he had no explanation for the declaration of trust by Mr Higgins in his favour.) As for GCL, Mr Brunswick was personally involved in its incorporation and he accepted that he had a one-third beneficial interest in it (somewhat but not all that different from the 40% indicated by Mr Higgins in his manuscript note on the back of the transfer of £200,000 by SGC to GCL). Thirdly, the bonds which Mr Brunswick signed on behalf of Templeton in December 2005 and/or January 2006 were bogus bonds. He gave evidence that he believed these were genuine bonds which were to be delivered to their beneficiaries by SGC, having been agreed between him and Mr Higgins over the telephone at various times over the course of the Markel and QBE binders. Yet the failure to bring such bonds into existence at the time of their negotiation, and to deliver them at that time to the beneficiaries, counted against their genuineness. As the judge remarked: “The batch of bonds signed in December 2005 and the batch of bonds requested to be signed in May 2006 coincided with an audit to be conducted by Markel in December 2005 and by QBE/Amalfi in mid 2006” (at para 249). Mr Brunswick accepted that he did not give any active consideration to the underwriting of the risks beyond acting upon the advice of Mr Higgins. There was no satisfactory documentation to support the genuineness of the bonds in either SGC’s documentation or in Templeton’s.

Fourthly, as to the payment of premium by SGC to Templeton, the judge found that there was no proper accounting for such premium. Some premium was paid to Templeton during the Markel period but none during the QBE period. Mr Brunswick suggested that the premium was held on trust by SGC for Templeton, but there was no formal or detailed accounting between SGC and Templeton in respect of premium save for one e-mail (dealing with a specific bond).

Fifthly, Mr Brunswick was implicated in the transfer of premium from SGC to GCL. In this respect the judge made the following findings (at para 252):

“GCL was the recipient of £200,000 and £288,000 to which I have already referred. They derive from premium received by SGC and not accounted for to Markel or QBE/Amalfi. There is, moreover, a very clear connection between Mr Brunswick and the payment of the £288,000, namely two letters to SGC dated 12 August 2006 apparently signed by Mr Brunswick relating to this sum. The first describes the sum as commission and requests that it be paid to GCL. The second describes the sum as collateral and requests that it be paid to GCL. The sum was neither commission nor collateral and these letters are dated after the sum was transferred…Mr Brunswick had no plausible explanation for the letters…In his oral evidence, when asked to explain why these documents did not show him to be party to fraud, he could only suggest that he had been “massively negligent” in signing them…The letters appear to be an attempt to justify the payment of £288,000. The two justifications put forward (commission or collateral) were untrue.”

The judge therefore concluded as follows:

“253.

Although Mr Brunswick denied involvement in any fraud I consider that the inevitable inference to be drawn from the above facts and matters is that Mr Brunswick was party to the fraud committed by Mr Higgins and Mr Felstead. The fact that they each had a beneficial interest in GCL and that GCL received premium which had not been accounted for is the clearest evidence that they were in the fraud together. Mr Brunswick’s role was to provide bogus Templeton bonds for SGC’s files so that auditors of SGC might be persuaded that the exposure of Markel and QBE/Amalfi was within the agreed limits.”

It is now necessary to explain how Templeton and/or its bogus bonds were used to demonstrate that the out-of-limit bonds signed by SGC were within the Markel and QBE binders. (Although we are essentially concerned with out-of-limit bonds which were for sums in excess of the financial limits imposed by the binders, or so-called “overwritten” bonds, there were other bonds which were out-of-limit in that they failed to observe different limitations, such as geographical limitations, or a prohibition against “on demand” bonds.) There were two stages in this attempt. The first stage was represented by the issue in the ordinary way by SGC to the beneficiaries of (out-of-limit) bonds in the name of Markel and QBE.

In the Markel period these bonds were for the most part signed by Mr Williams or Mr Felstead, although a small number were signed by Mr Higgins. In the QBE period Mr Higgins signed many of these bonds. However, whoever signed the bonds in question, he was significantly involved in the writing of them. He was the bond underwriting expert at SGC, trading on his long experience in this field.

At this first stage, the bonds would be in the amounts for which they were requested and for which they were paid in terms of premium. Thus, for instance, a £4m bond would, according to its own circumstances, of course require the payment of an appropriate premium for a £4m bond. However, such bonds would not be entered in the monthly bordereaux submitted to Markel and QBE in terms of their true amount or true premium. They would be entered for amounts within the financial limits of the binder concerned and for a lesser premium which reflected that lesser amount. Some bonds were not reported at all. The Canadian bonds were reported to relate to the UK not Canada.

At the second stage, in order to deal with the Markel and QBE audits, the bond files were altered to bring the bonds into line with what the bordereaux showed. Thus the original overwritten bond would be replaced by a dummy (Markel or QBE) bond which was within limit, and the balance of the original bond amount would be reflected by a bogus Templeton bond.

So far as Mr Higgins is concerned, the judge made no express findings to implicate him personally, as distinct from Mr Felstead and Mr Brunswick, with that second stage. Thus the judge found that Mr Felstead was “directly involved in this activity” (at para 181). I have already set out the judge’s findings concerning Mr Brunswick and his signing of the bogus Templeton bonds. There is no equivalent finding concerning Mr Higgins at this second stage. Nevertheless, the judge’s findings in general are that what each of the three conspirators did they did as conspirators to defraud Markel and QBE and to share in the unaccounted for premium as secret profits. An example is the judge’s finding at para 253 of his judgment (cited at para 35 above).

However, the judge concentrated on Mr Higgins’ role at the first stage. How were the overwritten bonds to be explained? At various times Mr Higgins, or those acting on his behalf, have offered four explanations, each of them designed to deflect any possible inference of dishonesty or collusion. (1) One was that what he did, he did deliberately, with the express agreement, on a bond by bond basis, of Mr Smith. (2) A second was that there was a “silent co-surety arrangement” with Templeton, whereby Templeton agreed to participate in each bond to the extent that it was signed for an amount beyond the relevant binder’s limits. (3) A third was that he was already then suffering from Alzheimer’s, and that this affected both his memory (of limits) and his ability to function within the business (and latterly, the trial) environment. (4) Finally, it was said that he was the dupe of the dishonest Mr Felstead, who ran the office. To some extent, these four strands in Mr Higgins’ defence overlapped, for instance strands (1) and (2) overlapped, and strands (3) and (4) overlapped. It is not so easy to see how strands (1) and (2) overlap with strands (3) and (4). At any rate, at trial strands (1) and (2) were in the ascendant, but on appeal strands (3) and (4), and in particular strand (3), have come to the forefront.

At trial, it was Mr Higgins’ evidence that what he did he did knowingly and deliberately, with Mr Smith’s agreement, and/or as part of a silent co-surety arrangement. The essence of his evidence in this respect is set out in this passage of the judge’s judgment:

“113.

In his statement dated 27 September 2007, which formed part of his evidence in chief, Mr Higgins described how and why he wrote surety bonds in the name of Markel in excess of the financial limits…He employed an arrangement which he described as “silent co-surety”. Templeton was the silent co-surety and would issue a bond for the difference between the amount of the bond written in the name of Markel, say £2m, and the financial limit which Markel had agreed to accept, namely, £1m. Thus, on this example, Templeton would issue a bond for £1m. That bond would not, however, be known to either the client or to the beneficiary. Hence it was “silent”. Mr Higgins said that “clearly Markel and Templeton knew and agreed to this arrangement” and that “Markel was only really at risk for £1m, as Templeton effectively agreed to indemnify (or reinsure) Markel for the other £1m.” He said that “Mr Smith knew about this arrangement and agreed to the same.” When Mr Smith moved from Markel to Amalfi “this silent co-surety arrangement also moved seamlessly to the Amalfi/QBE period and continued as it always had. At all times Peter Smith was the key person who knew all about this and gave his approval for it to continue.” He said that Mr Smith “happily provided ‘special acceptances’ (where bonds to be written were over the Markel/QBE binder) as and when we requested them.” Mr Higgins explained that if he knew that he was writing a bond which would exceed the limits he would make sure that Peter Smith agreed to that and, if necessary, that some co-surety/reinsurance was involved.

114.

In his further statement dated 26 October 2007 he said that breaches of the financial limits and period limits would be orally agreed by Mr Smith “although in the case of the QBE/Amalfi binder, I am sure in several cases Peter Smith would sign off the Bond Risk Review Sheets to confirm such exceptions.”

115.

In his oral evidence he gave similar evidence:

“There is no way that I would have exceeded the limits unless I had obtained approval from Markel or whoever, I think it was Peter Smith or somebody. The practice was that mainly Peter Smith would come into the office and look at individual cases. When I wanted to go above the agreed limit of £1 million, I think it was £1 million under the contract, he would say yes, no or whatever and I would explain why. Then he would initial the facing slip and I would proceed on that because I thought he was acting – well, he was acting on behalf of Markel.”

However, the judge rejected this account, and it is accepted that there is no way back on the judge’s relevant findings. The documentation showed that Mr Smith only ever signed off on a number of cases where the 3 year limitation on the bond period was extended. This was something which the Markel binder specifically allowed for, the QBE binder did not, but the QBE bond risk review sheet later expressly permitted under a review signature. There is no documentation in which Mr Smith signed off on a bond written for more than £1m (or £2.5 in the aggregate for any one principal). Nor is there any documentation to support a silent co-surety arrangement. Nor indeed was Mr Higgins’ evidence entirely clear about exactly how such an arrangement was meant to work. He ultimately agreed that there was no support for it in the modus operandi document in the case of Markel or its equivalent in the case of QBE. Indeed, Mr Higgin’s explanation varied between the case where Markel or QBE insured the whole amount of an overwritten bond but there was nevertheless an understanding that Templeton would effectively reinsure the overwritten amount; the case where Templeton would provide its own bond for the overwritten amount, but it would be “silent” because such a bond would not be provided or known to the client or beneficiary; and the case where each of Markel/QBE and Templeton would provide a bond for their respective amounts.

An example of the last case was a bond risk review sheet issued on 4 May 2006 in favour of St George Central London Limited at the request of Ardmore Construction Limited. This concerned a bond for the total sum of £2,461,300 for a period of 5 years. The extension to 5 years had to be specifically “noted and agreed” by Mr Smith. The bond value, however, was specifically described in the review sheet as “£2,461,300 split with Templeton”, expressly giving Templeton 59.37% or £1,461,300 and QBE 40.63% or £1m. All these figures were set out in the sheet. Thus Mr Smith was being told that there were separate bonds in those amounts. That would reflect the binder limitations that any arrangement with Templeton would only be on a several and not on any joint basis. Mr Higgins also signed this review sheet. It reflected the way in which, at the second stage, with the help of dummy Markel and QBE bonds and bogus Templeton bonds, the file documentation was sought to be brought into line with what the insurers were entitled to expect. However, there was never any documentation either as between SGC and Markel or QBE, or as between SGC and Templeton, to reflect any overall arrangement whereby Templeton co-insured or reinsured Markel/QBE bonds. Indeed, Mr Higgins’ explanation of the silent co-surety arrangement was inconsistent with Mr Brunswick’s understanding of the situation: Mr Brunswick denied any such arrangement, and regarded bonds signed by Templeton as being bonds upon which the beneficiary could sue.

Despite this lack of documentation to support Mr Higgins’ case, it was submitted on his behalf that Mr Smith was giving false testimony in denying any knowledge of overwritten bonds or of such a co-surety arrangement and that Mr Smith “must have known” and agreed to what was going on. The “central pillars” of that case are set out in para 120 of the judge’s judgment, but the case failed. In reaching his conclusion, the judge had to take into account his own view that Mr Smith was in many respects an unsatisfactory witness and that he was bound to exercise considerable caution before accepting his evidence. Nevertheless he did so, and set out with care his reasons (at paras 126/137). Amongst those reasons was the consideration that if such a silent co-surety arrangement existed, or if Mr Higgins had even believed that Mr Smith had assented to any such arrangement, there would have been no need for the bogus Markel/QBE bonds to have been placed on file at the second stage (at para 132). It is clear from this passage that, although the judge did not found his conclusions about Mr Higgins’ dishonesty on his personal participation in that second stage, he did not acquit him of all responsibility for it.

In this connection the judge was right to observe that if, as he concluded, there was no silent co-surety arrangement between Markel or QBE and SGC, then the case that Mr Smith had assented to bonds being written in excess of the financial limits “must fall away” (at para 134). The judge also went on to find that not only was Mr Smith unaware of the overwriting of financial limits, but also he did nothing to lead Mr Higgins to believe that he was consenting to the writing of non-compliant bonds.

It is important to observe (and the importance of this will become more obvious when I go on to consider the reliance which is placed in this appeal on Mr Higgins’ condition as a sufferer from Alzheimer’s) that the defence based on the silent co-surety arrangement was no mere reconstruction of events when it came to trial. It had been the explanation given by Mr Higgins to Mr Williams during the period of the binders themselves, when Mr Williams questioned what was going on in terms of bonds being written in excess of the financial limits. Mr Williams was very conscious at the time of the events that bonds were being written in excess of the applicable limits, but, as the judge found, he allowed himself to be persuaded to overlook his concerns, which persisted or even increased, in favour of a “quiet life”.

Thus Mr Williams gave evidence that he had been totally ignorant about surety bonds, had questioned what was happening, and had been assured by Mr Higgins (and by Mr Felstead) that surety bonds were quite different from insurance generally and that the co-surety arrangement was the appropriate explanation for what was going on. The judge considered Mr Williams in general to have made every effort to answer questions truthfully and to the best of his recollection, and regarded him as a frank witness. He accepted Mr Williams’ evidence concerning the explanations given to him, acquitted him of participation in a conspiracy to defraud, but concluded that he had been nevertheless dishonest in lending assistance to what was going on. Thus:

“201.

As I have already said, I accept Mr Williams’ evidence that he asked Mr Higgins whether Mr Smith agreed to the signing of bonds which exceeded the limits...If, as I accept, he was told that it was in order to sign bonds in excess of the limits because Templeton would, as a co-surety, issue a bond for the difference between the value of the bond and the agreed limit of Markel’s liability it is again very likely, given his experience and intelligence, that he would have reacted to such information with scepticism. He would surely have wondered how such an arrangement could work…

204.

It seems clear that during the QBE/Amalfi period Mr Williams’ suspicion that SGC’s business was not being conducted in an honest and proper manner increased…The continued involvement of Mr Felstead in the surety bond business notwithstanding the discovery of his conviction and the assurance given to QBE/Amalfi must have fuelled his suspicions.

205.

Prior to the Markel audit in December 2005 he discovered that Mr Felstead was tampering with the bond files. Similarly, in mid-2006 Amalfi carried out an audit. Mr Williams accepted in cross-examination that in dealing with the audit queries he became aware that the QBE bond files had been tampered with and that the bonds in them were not the correct ones. These discoveries must have convinced him that the manner in which the surety bond business was being conducted was not proper. Indeed, he said in evidence that “at the time I thought there is something not right here”. He said he raised the matter with Mr Higgins but accepted that the answer he received was not satisfactory.”

The judge therefore concluded that Mr Williams did not honestly believe that his conduct in signing bonds in excess of the stated limits was justifiable.

Thus, at trial it was Mr Higgins’ evidence that he was entirely conscious of the financial limits imposed on him; that any talk by him of being entitled to write bonds for sums even up to £10m was merely “marketing” talk, to raise interest; that in going beyond the binder limits he was covered in every case by Mr Smith’s oral or written agreement; and that the arrangement which supported that agreement was that by which Templeton would share the liability (in some way) as a co-surety, albeit silently. In that evidence, he was supported by Mr Williams on the basis of his own conversations with Mr Higgins on the subject, albeit, as the judge found, Mr Williams ultimately had no real or honest belief in the truth of what he was being told.

This section of my judgment can be quickly stated. This is not an appeal on law, nor is there any dispute about the law. It is common ground that where a claimant seeks to prove a case of dishonesty, its inherent improbability means that, even on the civil burden of proof, the evidence needed to prove it must be all the stronger: see Hornal v. Neuberger Products Ltd [1957] 1 QB 247, Re H [1996] AC 593 at 586/7, Re Doherty [2008] UKHL 33 at paras 23/29, Re B (Children) (Care Proceedings: Standard of Proof) [2008] UKHL 35, [2008] 3 WLR 1. The judge did not cite these cases other than Re H, but that is sufficient. Moreover, it is clear to me reading his judgment that he proceeded on a basis, the right basis, which must have been common ground before him. He said right at the beginning (at para 4):

“The burden of proving this fraud lies upon the Claimants who must do so on the balance of probabilities. But cogent evidence, commensurate with the gravity of the allegations made against the defendants, is required to prove the allegation on the balance of probabilities: see Ikarian Reefer [1995] 1 Lloyd’s Rep 455 and Re H [1996] AC 563.”

There is no sign in his judgment that the judge forgot or failed to apply the requisite standard of proof. He was particularly sensitive in his evaluation of Mr Higgins’ evidence, seeing that it was common ground on the medical evidence that he was, and had been, suffering from Alzheimer’s (see below). It is clear that at the crucial point of his judgment, he expressly reminded himself of the necessary requirement of cogent proof. He said –

“177.

I have reminded myself of the standard of proof required to prove the serious allegations made against Mr Higgins. But in my judgment it is an irresistible inference from the facts that I have found, coupled with the evidence concerning GCL, that Mr Higgins was involved in the failure to account fully to Markel and QBE/Amalfi in respect of premium, that he conspired with others to injure Markel and QBE/Amalfi by obtaining a secret profit and that he could not honestly have believed that his actions were justified. I am therefore compelled to reject Mr Higgins’ evidence that he was not involved in any dishonest or fraudulent activity.”

There is similarly no dispute about the law which obtains on an appeal against a judge’s findings of fact in such a case: see for instance Piglowska v. Piglowska [1999] 1 WLR 1360 at 1372 per Lord Hoffmann, and Assicurazioni Generali SpA v. Arab Insurance Group [2002] EWCA Civ 1642, [2003] 1 WLR 577 at paras 12/23. The importance of the opportunity for assessing the truth which a judge of trial has by reason of seeing and hearing the witnesses give their evidence is well attested in authorities such as Watt v. Thomas [1947] AC 484 at 487/8 per Lord Thankerton and Benmax v. Austin Motor Co Ltd [1955] AC 370 at 375 per Lord Reid. Moreover, this is not one of those cases where the documentary material, and the inherent probabilities and possibilities of the case, detract from the judge’s evaluation of the witnesses he has heard: see Armagas Ltd v. Mundogas SA (The Ocean Frost) [1985] 1 Lloyd’s Rep 1 at 57 per Robert Goff LJ. Thus it is not submitted that the documents and the probabilities are in conflict with the judge’s finding of fraud. On the contrary, Mr Gruder accepts that there was a conspiracy involving Mr Felstead and Mr Brunswick. The question was rather whether Mr Higgins, who was the principal behind SGC and its underwriting expert, was a party to that conspiracy; and in that connection whether his dementia was a significant factor which the judge failed to evaluate properly.

I therefore turn directly to the medical evidence about Mr Higgins and to his important role as the underwriter of the bonds.

On this appeal, Mr Jeffrey Gruder QC on behalf of Mr Higgins has to accept the bulk of the judge’s findings, but, he submits, the judge erred by failing to appreciate that Mr Higgins’ condition as a sufferer from Alzheimer’s exposed him to forgetfulness and to exploitation at the time of the events in question and to a situation at trial where he could not do himself justice.

Mr Gruder builds his appeal on the following broad foundations. First and foremost, he submits that the judge failed to make sufficient allowance for Mr Higgins’ condition. Dishonesty should not have been found where an alternative explanation can be said to exist in disability. Secondly, and in reliance on his first point, he seeks to ameliorate or if necessary set aside the judge’s conclusions on certain critical aspects of the case against Mr Higgins: that he knowingly and dishonestly wrote bonds above the binder limits; that he was involved in the dishonest preparation of bordereaux; and that he participated in any dishonest attempt to salt away illicit proceeds to GCL. Thirdly, he submits that the judge acquitted him of any participation in the use of bogus bonds at the second stage of events.

Under this heading, I will set out the medical evidence and the way in which the judge treated it, especially for the purpose of considering Mr Higgins’ role as the underwriter of the bonds.

Mr Higgins is now 72. In 2005/6, at the time of the binders, he was 68/69. At the time of the trial he was 71. It is not in dispute and was not in dispute at trial that he is and was then suffering from Alzheimer’s. He was diagnosed as so suffering in the run-up to trial in 2007. He was seen by three consultant physicians. The first to see him was Dr Robin Powell, a consultant psychiatrist and honorary senior lecturer at Imperial College School of Medicine, whose report dated 20 January 2007 recommended further investigations, but concluded that he had a history of at least two years and possibly longer gradual decline in his memory and intellectual functioning, typical of a dementing illness in its earlier stages. Although the deficits were at that time relatively mild, the illness had been developing over a prolonged period. Mr Higgins had scored 24/30 on a mini mental state examination. (A score of 26 and above is normal).

The trial expert consulted by the claimants was Dr Angus Kennedy, a consultant neurologist at the West London Neuroscience Centre at Charing Cross Hospital and Chelsea and Westminster Hospital. He examined Mr Higgins on 5 June 2007, who scored 25/30 on the mini mental state examination administered to him on that occasion. Dr Kennedy concluded that he was suffering from a “mild” form of Alzheimer’s and had probably been suffering from some level of preclinical intellectual under-functioning, likely to have been significantly less than the then current situation, in 2005/6.

The trial expert consulted by Mr Higgins was Professor Sube Banerjee, inter alia professor of mental health and aging at the Institute of Psychiatry, King’s College. His two reports are dated 25 July 2007 and 16 January 2008. He also diagnosed Alzheimer’s, then of “moderate severity” and increasing in severity for at least the last two years. Mr Higgins scored 24/30 on his mini mental state examination. Professor Banerjee doubted Mr Higgins’ capacity to give instructions for the purpose of trial. It was therefore decided to use his son, Timothy, as his litigation friend for the purpose of instructing lawyers.

Dr Kennedy and Professor Banerjee signed a joint medical report dated 11 February 2008. They agreed on the diagnosis of Alzheimer’s, subject only to a disagreement as to whether it was “mild” (Dr Kennedy) or “moderate” (Professor Banerjee). They agreed that Mr Higgins’ memory was currently “moderately” impaired; that his disease could have been diagnosed for a period of some years before 2007; that earlier symptoms would almost certainly have included impairment of memory; but that variability between patients made it difficult to predict impairment either prospectively or retrospectively. In the absence of direct measurement, however, some information on levels of functioning could be obtained from collateral reports, ie from the observations of those who knew Mr Higgins at the relevant time. (At trial Professor Banerjee spoke of collateral histories as “absolutely the core of our ability to be able to work out what is happening with the person with dementia alongside the history taken from the person with dementia and the examination and investigations”: Day 14.51/52.) They were agreed that it was very unlikely that Mr Higgins was either feigning or exaggerating the deficits of the disease.

How then should the judge regard Mr Higgins’ evidence? The judge was clear about three important matters. First, that Mr Higgins’ ability to recall events in 2005/6 was “significantly” impaired (a generous interpretation of the experts’ agreed evidence) and that it would therefore be unfair to conclude that there was any evasion in his professed failure to remember (at para 105). Secondly, that it would not be safe or fair to conclude that Mr Higgins was playing upon his condition (at para 111). Thirdly, he was guided by the expert evidence to observe that memory deficits in such a patient were likely to be intermittent rather than constant; and that therefore the court had to look at Mr Higgins’ evidence as a whole to see if he demonstrated consistency and understanding (at para 109). Fourthly, it may be said that it would be valuable as well to consider the experience of other observers of Mr Higgins’ condition. Fifthly, he had evidence from the experts that the most recent quantitative medical assessment of the current degree of Mr Higgins’ dementia (the mini mental state examination scores) was 24 or 25 out of 30, where 26 and above was within the normal range and where, on Dr Kennedy’s evidence, a falling off of 1 or 2 points a year was to be expected. On that basis, Mr Higgins would have been scoring normally at the time of the events in question.

The judge expressed his conclusions about Mr Higgins in these paragraphs from his judgment:

“154.

I had the advantage of listening to Mr Higgins give evidence over 3 days. Whilst there were occasions when he did not follow the question or appeared to contradict himself there were many periods when he carefully explained his understanding of the issue under discussion and answered many questions on a particular issue. He did not appear to me to be someone whose memory and executive function (that is, his ability to manipulate information, form concepts and solve problems) were so impaired that he would, in the course of writing surety bond business in 2005 and 2006, regularly confuse or forget the limits on his underwriting authority. The importance of adhering to such limits would have been something which he would have been aware of throughout his professional life and so less likely to forget. Whether the scale of his present dementia be mild or moderate the manner in which he gave evidence strongly suggests that in 2005-2006 his dementia was not such as to cause him to confuse the limits on the Markel and QBE/Amalfi agreements with those of other and older binders or to forget the relevant binder limits.

155.

Members of his family did notice incidents of forgetfulness, though there was uncertainty as to how many years back these went. I do not consider these instances to be inconsistent with the view I have formed having listened to him give evidence. Mr Justice Cooke described Mr Higgins as not well and to have very little recollection when giving evidence before him in Oxus Gold Plc v Templeton Insurance Ltd [2006] EWHC (Comm). Again, I do not consider that this assessment of Mr Higgins’ reliability as a witness in 2006 to be inconsistent with the view I have formed that in 2005 and 2006 Mr Higgins is unlikely to have confused or to have forgotten the limits on his underwriting authority on a regular basis.

156.

Mr Higgins’ son-in-law Mr Ray, who had worked with him until the end of 2003, said that he regarded Mr Higgins as a “sick man” since 1998 and doubted his ability to conduct surety bond business from 2003. Mr Ray’s evidence does not fit with the other family evidence because it puts the onset of Mr Higgins’ symptoms much earlier than other family members put it. It seems to me that Mr Ray’s evidence is probably best explained by serious differences between Mr Ray and Mr Higgins as to how to conduct the surety bond business rather than as a genuine sign of Alzheimer’s disease significantly affecting his cognitive abilities.

157.

I am therefore unable to accept the submission made on behalf of Mr Higgins to the effect that it is more likely than not that from about 2004 his defects in memory and executive functions led him either to confuse or forget the relevant binder limits. On the contrary I find that it is more likely than not that, although he was suffering from the early stages of Alzheimer’s disease in 2005 and 2006, Mr Higgins was aware of the importance of adhering to the limits of his underwriting authority and that he remembered those set out in the agreements with Markel and QBE/Amalfi. The Markel and QBE/Amalfi limits were of course part of his recently acquired information and therefore more likely to be forgotten but his own case requires that where he exceeded those limits he was permitted to do so by ensuring that Templeton was a secret co-surety. There is in truth no evidence that he forgot the limits. His own case requires that he appreciated when the limits were to be exceeded and made provision for it. The submission that it is more likely than not that from about 2004 his defects in memory and executive functions led him either to confuse or forget the relevant binder limits ignores the reliance placed by Mr Higgins on the supposed role of Templeton as silent co-surety and the reasons for such reliance. It is improbable that he simply wrote surety business without reference to limits and no such submission was made. He himself accepted in cross-examination that he was entirely conscious of the limits when underwriting.” [emphasis added]

Mr Gruder nevertheless submits that the judge, for all his apparently careful assessment, got it wrong. Thus it is said that the judge erred in observing, in the sentence highlighted above, that “There is in truth no evidence that he forgot the limits.” Mr Gruder says that there was such evidence, principally coming from Mr Williams himself. Mr Gruder referred to passages in Mr Williams’ evidence where he spoke of Mr Higgins saying that he could write £5m or even £10m business; of occasions when it was not possible to tell whether Mr Higgins was joking in such protestations; of how Mr Liam Curran, one of SGC’s employees, resorted to sticking a note of the limits over the screen of Mr Higgins’ desktop computer (possibly as sharing in the joke); but also of one occasion when he and Mr Higgins had a blazing row about the binder limits, with Mr Higgins insisting that the limit was £5m. Mr Williams also referred to Mr Higgins as on occasions “very, very erratic at best” (at Day 12.148), or “totally forgetful, hopelessly forgetful” (at Day 12.149), or being willing to “sign every bond placed in front of him” (Mr Williams’ witness statement). Mr Gruder also relied on the impressions of others, such as Cooke J in 2006 (“plainly not well…very little actual recollection”, referred to by the judge himself, see above at para 155 of his judgment), or of QBE’s Ms Lawton reporting that Mr Higgins’ decisions were causing concern in the market and that his family were concerned as to whether he should be working at all (see her email dated 23 March 2006). He also referred to the evidence of members of his family who gave examples of his forgetfulness on occasions: such as getting lost on a walk in the woods, or forgetting in which car park he had left his car.

In this connection Mr Gruder also emphasised certain answers given by Mr Higgins himself in his evidence: such as that he did not know what the allegations were against him or why he was really in court (evidence which the judge took into account as being genuine, see paras 106/111); or other passages in his evidence where he appeared to be genuinely confused, or to be mixing up the Markel/QBE binders with previous underwriting episodes.

On the other hand, Mr Michael Swainston QC on behalf of Markel and Mr Derrick Dale on behalf of QBE and Amalfi responded by emphasising other aspects of the evidence before the judge. They particularly emphasised the evidence of Mr Higgins himself, who insisted on his absolute knowledge of the limits involved, and his positive case of the silent co-surety arrangement and of Mr Smith’s agreement to every instance of an otherwise out-of-limit bond; and that of Mr Williams who spoke repeatedly of the tight control which Mr Higgins together with Mr Felstead kept on the business, and of how Mr Higgins had the expertise to rate bonds and was an excellent and acute reader of accounts. They also relied on other evidence, such as that of Mr Hodgins, a solicitor on secondment to Markel, who said that in May 2006 he had found Mr Higgins clear and articulate; of Mr Godwin, a business partner of Mr Higgins, who said that Mr Higgins was “very astute” in early 2005; of Mr Coyne, another business partner of Mr Higgins, who said that he had “a razor sharp mind”; and of Mr Smith himself, who said that Mr Higgins was a highly intelligent man who had no trouble in addressing any issue which arose.

In my judgment, it is quite impossible to say that the judge erred in his assessment of the medical evidence or in his conclusion that in 2005/2006 Mr Higgins knew what he was doing and was not simply the victim of his developing illness. There can hardly be a question of fact which lies more in the province of the trial judge than the assessment that Teare J had to make, retrospectively, of Mr Higgins’ state of mind and intellectual functioning at the time, several years earlier, when he was operating the Markel and QBE binders. The judge had to take into account the medical evidence, assisted as he was by the fact that for the most part it was agreed; Mr Higgins’ own evidence at trial, which the judge had both in writing, in two lengthy witness statements, but also most importantly in three days of oral evidence; the evidence of those most intimately concerned with Mr Higgins’ business dealings, such as Mr Williams and Mr Smith, which the judge also had given before him at length and which he was able to probe in depth; and the evidence of other witnesses, all of whom could in their own way provide the sort of collateral accounts which the medical experts had said were useful touchstones of the ultimate conclusions to which the judge had to find his way. The experts essentially said that the assessment was for the judge to make, and they indicated how he could look to make it: by testing for consistency, and taking into account both Mr Higgins’ own history of himself and the collateral histories of others who could speak of their experiences of him.

In these circumstances, unless the judge proceeded on the basis of some critical error, or else plainly erred in his conclusion, it would be wholly unprincipled for this court to interfere. Mr Gruder submits that the judge did proceed on an error, when he said that there was in truth no evidence that Mr Higgins forgot the binder limits. However, it seems to me to be plain that the judge was using that expression to mean no material evidence. Intermittent forgetfulness of the kind spoken of by the experts as symptomatic of Alzheimer’s in its early stages, or exemplified by some of the witnesses, could not begin to explain the consistent overwriting of so many bonds over such a length of time. That would be an extraordinary coincidence. Moreover, Mr Higgins himself was adamant that he knew the limits, and always obtained Mr Smith’s consent to any bond which went beyond the binder limits; and that Mr Smith’s consent was facilitated by the silent co-surety arrangement. While there was admittedly some evidence of intermittent forgetfulness of one kind or another, Mr Higgins’ own dominant and positive case, supported by Mr Williams’ evidence, that he relied at the time on a silent co-surety arrangement in his explanations of how and why he might appear to be breaking limits but was not in fact doing so, was wholly antithetical to a finding that Mr Higgins had erred through illness.

In effect, Mr Gruder’s submission asks this court to say that all over-writing was done out of forgetfulness; that no one in the office, with the exception of Mr Williams’ somewhat half-hearted effort, was willing to deal with what ex hypothesi was not a case of deliberate over-writing but mere forgetfulness, by reminding him of the binder limitations and taking further steps if necessary to remedy the situation; that it was essentially coincidence that he had both forgotten how to conduct himself in accordance with the binders and had also fallen among thieves in the partners he had himself chosen to associate with at SGC and GCL; or else that despite his being reminded of his obligations his disease somehow led him not merely to intermittent or even pervasive forgetfulness but also to believe in a silent co-surety arrangement of which there was no documentary evidence but of which again his partners in SGC and GCL were prepared to take advantage, while he retained his innocence. The truth must be that if Mr Higgins had suffered from a dementia induced intermittent forgetfulness, the position would have been soon and easily remedied, not least by the inevitable disclosure, whether inadvertent or deliberate, of his mistakes to Mr Smith as being those of an honest albeit forgetful (or forgetful albeit honest) man.

However, in my judgment, the scenarios which have to be canvassed by Mr Gruder are unlikely scenarios, and will be seen to be still the more unlikely when other aspects of the judge’s findings are considered below. For the present I would comment as follows. The over-writing of so many bonds, persevered with in the face of not one but two negotiations of the binder terms with Markel and QBE successively wherein the limits and the requirement of several bonds (where more than one insurer might be concerned) were emphasised, in the face of so many meetings or conversations with Mr Smith in which the truth must have emerged if the truth was honest, accompanied by the drawing up and presentation to Mr Smith of bordereaux which did not honestly present any co-surety situation, leads in my judgment to the compelling and even inevitable inference that Mr Higgins’ explanation that he was honestly employing a silent co-surety arrangement was the untruthful explanation of a dishonest man. Whatever Mr Higgins’ involvement in the drawing up of the bordereaux (see below), if the silent co-surety arrangement had really been the honest basis of the overwritten bonds, he must have given instructions for the proper disclosure of the situation to those whose responsibility it was (if not his) to draw up the bordereaux.

Moreover, each element of that silent co-surety defence had to be rejected on the evidence. There was an attempt at one time to find the co-surety arrangement, at any rate as between Markel/QBE and SGC, in the modus operandi document or its QBE equivalent: but ultimately even Mr Higgins came to accept that that did not represent or justify it. There was no documentary support therefore of any silent co-surety arrangement as between SGC and Markel or QBE. Nor was there any documentary support for such an arrangement between SGC and Templeton. The closest one gets to any documentary support for such an arrangement between SGC and Templeton lies only in the bogus Templeton bonds. Whatever might be said about Mr Higgins’ complicity in the drawing up of those bonds and their insertion into SGC’s files late in the day, the fact that Mr Higgins’ alleged silent co-surety arrangement is supported in documentary terms only by such dishonest documents cannot do other than undermine his case.

Nor is it possible to conclude, in accordance with a new argument advanced on appeal, that Mr Higgins’ own explanation and evidence at trial, that everything was agreed and that the silent co-surety arrangement was the bedrock of that agreement, was simply a manifestation of the disease, that it was in itself simply his disease speaking, leading him to do himself, so to speak, less than justice. That is because the experts were also agreed that the disease did not lead its victim to confabulate, to make up a story out of confusion or deceit; nor did it lead its victim to lose his moral or ethical sense of right and wrong. In any event, such an explanation is again totally undermined by the fact that, as Mr Williams’ evidence demonstrated, the silent co-surety arrangement was relied on at the time, and not merely by way of subsequent (and perhaps faulty) reconstruction, to justify what had been done.

In the circumstances I am unable to accept Mr Gruder’s primary argument, that Mr Higgins’ disease, and not dishonesty, is a plausible possible explanation for the overwriting of so many out-of-limit bonds, and that the judge ought therefore to have concluded that it would not be safe to find that Mr Higgins was a party to the conspiracy which surrounded him in the persons of Mr Felstead and Mr Brunswick, with whom he shared an interest in both SGC and GCL.

Nevertheless Mr Gruder had a further argument, that the judge’s finding of dishonesty was itself mistaken for more general reasons: to the effect that the judge ought not in any event to have found that Mr Higgins knew about the dishonest and misleading bordereaux; nor to have found that he was implicated in dealing with the proceeds of unaccounted for premium through the off-shore company GCL. Mr Gruder also submitted that the judge acquitted him of complicity in the provision of bogus bonds at the second stage, in the face of the Markel and QBE audits.

Mr Gruder put it in this way. He said that the judge found there to have been dishonesty in four different aspects or components of the overall conspiracy: (i) in the over-writing of the out-of-limit bonds; (ii) in the preparation and use of the bordereaux; (iii) in the extraction from SGC of unaccounted for premium into GCL; and (iv) in the creation of bogus bonds and their insertion into SGC’s files. He submitted that whereas the judge had found Mr Higgins to be complicit in components (i), (ii) and (iii), he had acquitted him of complicity in component (iv). As for component (i), that was not evidence of dishonesty, because it was due to Mr Higgins’ illness and forgetfulness. As for component (ii), the judge was simply mistaken to find that Mr Higgins had been involved with the bordereaux. And as for component (iii), the judge should not have inferred against Mr Higgins any dishonest complicity. Component (iv) simply did not concern Mr Higgins. Therefore, there was no support at all for a finding of dishonesty.

The judge dealt with the bordereaux as follows. At issue 10 of his judgment, he asked the question, “Were Mr Higgins, Mr Williams and/or Mr Felstead involved in the creation of bordereaux which misstated the true position?” He acquitted Mr Williams of any such involvement, but found Mr Higgins and Mr Felstead to have been involved. Mr Higgins denied any involvement, and this was supported by Mr Williams. Both of them referred to Mr Felstead as being involved. The bordereaux were created on computer by Liam Curran. They were only occasionally printed off. When asked whether Mr Higgins had a role in producing the bordereaux, Mr Williams’ response was that “I don’t think he would have a clue how to do it, my Lord”.

The judge concluded with respect to Mr Higgins’ role as follows:

“167.

Counsel for Markel relied on the note of the meeting on 26 January 2005 which records that “Liam [Curran] sits with Tim [Mr Higgins] every morning and goes through the account”. Whilst this does not specifically refer to the bordereaux, Mr Higgins accepted in cross-examination that he worked closely with others in relation to the bordereaux. Those with whom he worked were Mr Felstead, Liam Curran (whom Mr Higgins was training up) and Mr Ward (the SGC accountant and Company Secretary). It was also apparent from his cross-examination that (as was to be expected) Mr Higgins was well aware of the important role of bordereaux, namely, to provide information to Markel and QBE/Amalfi as to the business which was being written in their name. Thus, although (i) Mr Higgins was only in the office for three days a week, and when in the office often took a leisurely lunch, (ii) he was not a user of the computer on which the bordereaux were prepared and (iii) he was suffering from the early stages of Alzheimer’s Disease, I consider that it is more likely than not he must have been involved in the production of the bordereaux and understood the information provided in them to Markel and QBE/Amalfi. I consider that Mr Williams’ answer which I have quoted above probably reflected Mr Higgins’ inability to create the bordereaux on a computer file.”

Mr Gruder submits that the judge was wrong to prefer the evidence of Mr Higgins (that he did discuss the bordereaux with Mr Curran and Mr Felstead) to that of Mr Williams that Mr Higgins was not involved. It was wrong to do so because of Mr Higgins’ condition, and the judge did not in this instance apply the requirement of cogent evidence. The judge’s reliance on the meeting note of 26 January 2005 was a flimsy basis for his conclusion.

In my judgment, however, the judge’s conclusion cannot be faulted. Although the judge asked and answered issue 10 in terms of the creation of the bordereaux, the evidence went well beyond their creation and extended, as was natural, to their review with Mr Smith. The meeting note of 26 January 2005 covers both aspects, for it not only explains how “Liam sits with Tim every morning and goes through the account”, but also shows the parties to that meeting (Mr Smith, Mr Higgins, Mr Felstead, Mr Curran) reviewing business. (For good measure another note item refers to the bond limit of £1m.) The judge could also well have referred to another meeting note of 23 February 2005 (the attendance included Mr Smith, Mr Higgins, Mr Felstead, and Mr Brunswick). Item 4 on that note, headed “Bordereau”, states that Mr Felstead “showed bordereaux of bound and quote[d] risks”. By that time two overwritten bonds had recently been bound, namely a PC Harrington bond (dated 11 February 2005, for £2m) and a Byrne Bros bond (dated 21 February 2005, for £1.623m). The meeting note also refers to the 50/50 modus operandi arrangement with Templeton, but emphasises that Mr Higgins “confirmed bonds will be issued on a Several basis not joint and several” (original emphasis).

Moreover, Mr Higgins’ evidence about both the creation and the review of bordereaux was quite clear; and he had no explanation at all for the bordereaux’s false content. For instance (at Day 8.33/35):

“Q…We know that you went through the account every morning with Liam Curran. You are nodding.

A.

That is right, sorry.

Q. We know that you supervised him and trained him up.

A. Um um.

Q. We know as well that you discussed the bordereaux with Mr Ward and Mr Felstead, did you not?

A. That is right.

Q. Given that we do not have a single bordereaux which gives an accurate account of the bonds and given that you were the underwriter exercising underwriting judgment, how do you explain how all the bordereaux are false?

A. How do I explain? I would have thought any bordereaux which I would have been responsible for would have been correct. I can only assume that they have been forged.

Q. But it is correct, is it not, that you would have provided the information on your bonds?

A. The information which was in our records would have been complete and the bordereaux should have correctly reflected every single bond that we issued.

Q…is it right that you went through them as well with Mr Smith?

A.

In the early days he came in frequently, I think a couple of times a

week, two or three times a week and went through it. So it was an ongoing process.

Q. So the answer is yes?

A. That is correct…

Q. But the question I am asking is, can you tell his Lordship why the bordereaux were false, why you did not react to the inaccuracy of the bordereaux when you went through them with Mr Felstead, Mr Curran and Mr Ward?

A. I cannot remember, I really cannot.

Mr Justice Teare: It is an astonishing thing, is it not, that the bordereaux should not record accurately the sums involved?

A. Well, I find it unbelievable…I am astonished.”

In his first witness statement, his review of the bordereaux was again emphasised. Thus

“47.

Once the Markel binder was up and running, I would meet with Peter Smith in SGC’s office fairly regularly. Liam Curran would usually be present during the meeting and we would run through the latest bonds we had written or were considering according to the bordereaux…”

There was another passage to similar effect in his second witness statement (at para 60).

A particular oddity of the bordereaux was the way in which the Interhealth bond was dealt with. This was the largest single bond written by Mr Higgins, at £10m. Markel’s original bond for £10m is dated 10 January 2005. The gross premium charged was £855,000. £7.5m was reinsured with EDC (Export Development Canada) and the reinsurance certificate is dated 7 January 2005. Nevertheless, the Interhealth bond does not appear in the bordereaux until May, when it appears as a £1m bond dated 13 April 2005 with a gross premium disclosed of £22,500. (£1.5m was “co-insured” with Templeton.)

Mr Gruder nevertheless relied on a further answer given by Mr Higgins in his oral evidence, but this time in re-examination (at Day 10.78/79):

“Q. Would you have been involved in preparing the bordereaux?

A. Would I have been involved in preparing the bordereaux? (Pause) Can I answer that a bit elongated?

Q. If it helps.

A.

At one time I used to attend to all the bordereaux myself and I would do it so I knew everything that was happening. Every item on the bordereaux would compare with the bond underwriting book and everything else. As time moved on, I no longer had an involvement in that. There was a young lad who would do the donkey work. I think most of that was then checked with Cliff Felstead. So the answer to the question really is that I was not involved in that to any real degree.”

In my judgment, however, it is not possible to say that the judge was not fully entitled to come to his findings in respect of Mr Higgins’ involvement in the bordereaux. Those findings reflect the documents, the probabilities, and Mr Higgins’ own evidence in cross-examination. To the extent that Mr Higgins’ answer in re-examination differs, the judge was fully entitled to ignore it. Nevertheless, it could be accepted that in due course Mr Higgins, having set the pattern in training Mr Curran, was prepared to leave such matters to become more a thing of routine, under the supervision of Mr Felstead. Even if that were to be the case, Mr Higgins’ answer in re-examination still retains what in all probability must have been the truth, namely “At one time I used to attend to all the bordereaux myself and I would do it so I knew everything that was happening”.

In other words, Mr Higgins, as one would expect, was not only involved in the underwriting of the bonds, but also in translating that business into the bordereaux and in satisfying Markel and QBE by means of those bordereaux that their binder business was being conducted properly and successfully. Since those bordereaux were false, plainly and astonishingly false, there is here another inevitable inference to be drawn that Mr Higgins was complicit in the conspiracy which also embraced his partners in the business.

The judge found that there had been no proper accounting to either Markel or QBE, or to Templeton, of the premium charged to clients for the overwritten bonds. Markel and QBE of course were only told on the bordereaux about the lower premium charged on bonds there represented as being within limit. The judge found as an inevitable inference that the three men interested in SGC, namely Messrs Higgins, Felstead and Brunswick, were seeking to defraud Markel and QBE of the secret profits that SGC was rolling up in this way. Moreover, expert accountancy evidence showed that some of these profits deriving from the excess premium charged on overwritten and undisclosed bonds was transferred out of SGC to GCL in March and May 2006. The judge explained:

“175.

In March 2006 a payment of £200,000 was made to GCL and in May 2006 a payment of £288,000 was made to GCL. A Service Agreement existed between GCL and SGC pursuant to which GCL was to receive a 30% commission for the introduction of business. However, this appears to be a charade because there is no evidence that GCL introduced business to SGC. Moreover, there is evidence that the two sums paid to GCL represented premium income in respect of which there had been no accounting to Markel or QBE/Amalfi. Thus an analysis of SGC’s records by an expert accountant suggests that the sum of £200,000 represents premium income received by SGC [which ought to have been accounted for to QBE]. The same expert suggests that the sum of £288,000 represents premium on the Interhealth bond and the PC Harrington bonds which had not been accounted for to Markel. There was no evidence to the contrary.”

I have already referred above to the judge’s further findings with respect to these payments to GCL in the context of Mr Brunswick’s role in the conspiracy.

Mr Gruder submits, nevertheless, that the evidence did not justify the judge’s findings that implicated Mr Higgins with GCL or the transfer to GCL of unaccounted for premium. He submits that there was overwhelming contemporaneous evidence that Mr Brunswick was the owner of GCL and exercised total control over it. He had given instructions for it to be set up, he was the nominal owner of all its capital, and certain letters which purported to be signed by Mr Higgins, one instructing Mr Brunswick to set up a company in Jersey on his behalf and another confirming that Mr Brunswick’s shareholding was wholly for his own benefit, were forgeries. Moreover, Mr Gruder submits that Mr Higgins’ evidence about GCL was incoherent (since he appeared to be thinking about another company entirely) or extremely vague.

It may be accepted that the signatures in question were not those of Mr Higgins: but the papers are full of such signatures where Mr Higgins authorised others to sign in his name. The judge was therefore entitled to regard that as of no moment. The fact remains that Mr Higgins was personally involved in transferring to GCL £200,000 which represented unaccounted for premium deriving from QBE bonds and that he indicated in his own manuscript on the back of that transfer that the moneys in GCL (or possibly the shareholdings in GCL) should be subject to a trust in favour of the three individuals who were interested in SGC, namely himself, Mr Felstead and Mr Brunswick. The fact remains that the explanation of that transfer (and of the second transfer) cannot be ascribed to either commission or collateral. Mr Higgins suggested a third explanation, which was that Jersey (he said the Isle of Man, but that was his error) was a good tax haven for such funds. What he was unable to explain was why he and his two colleagues, Mr Felstead and Mr Brunswick, were seeking a tax advantage in respect of premium paid by clients for insurance from Markel and/or QBE and long unaccounted for.

It may be true that the judge does not ascribe personal involvement by Mr Higgins in the signing of sham Templeton bonds by Mr Brunswick, or in the replacement of the overwritten bonds in SGC’s files by new, dummy, Markel or QBE bonds. However, it is not true that the judge acquitted him of such involvement. On the contrary, the judge referred to evidence given by Mr Williams that, when he discovered that Mr Felstead had been tampering with the bond files prior to the Markel and QBE audits and that the bonds in them were not the correct ones, “he raised the matter with Mr Higgins but accepted that the answer he received was not satisfactory”. Moreover, the judge’s finding in relation to Mr Brunswick and his signing of the sham Templeton bonds was that in so acting he was part of a conspiracy with Mr Higgins and Mr Felstead to defraud Markel and QBE by obtaining secret profits from them. See also the judge’s reasoning at para 132 of his judgment (referred to at para 45 above).

Markel and QBE were interested in promoting the court’s interest in what they alleged were earlier, ie pre SGC, instances of dishonest conduct by Mr Higgins in his insurance dealings. In answer to the forensic point that it was unlikely that Mr Higgins should descend to fraud after a long and successful career in the insurance world, and that the correct explanation of apparently inculpatory events was to be found in his illness and the opportunity which that gave for less scrupulous men to take advantage of him, Mr Swainston and Mr Dale submitted, in effect, that Mr Higgins was a leopard who had not changed his spots. The judge felt no need to go into these earlier matters. Nor do I.

In my judgment, the evidence against Mr Higgins of his involvement with Mr Felstead and Mr Brunswick in a conspiracy to defraud Markel and QBE, or in dishonest assistance in a breach of fiduciary duty, is overwhelming. Mr Higgins was at the time suffering from the early, possibly pre-clinical, stages of Alzheimer’s disease, which had progressed by the time of trial and its preparation to the mild or moderate state referred to in the evidence of the medical experts. Nevertheless, the judge made full allowances for Mr Higgins’ condition, and had the inestimable advantage in a case such as this of seeing and hearing Mr Higgins give his evidence over three days. We in this court do not share that advantage.

It is plain that SGC was formed by Mr Higgins together with Mr Felstead and Mr Brunswick. Mr Felstead had no business being involved, and Mr Higgins knew it, but persevered in maintaining his connection with Mr Felstead even after his conviction for insurance fraud had to be disclosed. Mr Brunswick had no business being involved if an important part of the SGC strategy was to use Templeton as a co-surety, and a “silent” one at that. In these circumstances, and from the very beginning, and despite the clearest of limitations on SGC’s authority to bind, Mr Higgins proceeded to underwrite far in excess of those limits, with the £10m Interhealth bond being an early and egregious example. His excuse for doing so, both at the time when he was questioned by Mr Williams, and at trial, was to the effect that Templeton stood in as a co-surety or effective re-insurer: but the bonds were not written in that way, and there was no documentary support for that excuse – at any rate until the bogus Templeton and Markel/QBE bonds were produced at the second stage in the teeth of Markel and QBE audits. In any event, the silent co-surety scheme would have been a plain breach of the binders and of Mr Higgins’ personal assurance that any bonds shared with Templeton would be on a several and not a joint and several basis.

Mr Higgins’ undoubted involvement in the underwriting and issuance of the bonds was compounded by his involvement in the creation and review of the bordereaux, which were false. It was his own evidence that he was so involved. In any event it stood to reason. It was he who had underwritten the bonds, and, on his own evidence, had done so on a basis which was not transparent, in that a bond had been issued in an amount which was out of limit, and whose possible legitimacy had to be explained, if that could be done at all, by reference to a sharing of the underwriting with a “silent” partner in the shape of Templeton: of which there was no sign in any documentation. Only Mr Higgins could explain all that to Mr Curran, who it seems was the employee tasked with the drawing up of bordereaux. At a later stage, these false bordereaux, which of course also had to misstate the premium earned by the bond in question, were reviewed by Mr Higgins together with Mr Smith. Mr Higgins’ case at trial was that Mr Smith knew everything, and approved of everything: but that case was not accepted by the judge. On the contrary, for all the difficulties of Mr Smith’s position, and the unsatisfactory aspects of his evidence, the judge concluded that he was telling the truth about this matter. The judge proceeded cautiously in that conclusion.

At the root of all this was premium. Bonds were issued for out-of-limit amounts and the premium earned on these large bonds was due to Markel and QBE. That was undoubtedly so, because Markel and QBE were liable on such bonds, subject to any successful repudiation of them. The fact that such liability might have been (secretly) shared with some other insurer was a collateral matter. SGC ought to have accounted to Markel and QBE for such premium, but did not do so. Alternatively, in accordance with the “silent co-surety” arrangement, SGC should have accounted to Templeton for a large part of this premium flow, but did not do so. Mr Higgins must have had some regard for the financial affairs of his company. Be that as it may, in due course he is found to participate with Mr Felstead and Mr Brunswick in transferring to GCL, in which Mr Higgins accepts that each of the three had an interest, unaccounted for premium which has been shown to derive from out-of-limit bonds issued under the binders.

In my judgment, each of these components of the case against Mr Higgins by itself would set up a powerful inference justifying the conclusion that Mr Higgins was acting dishonestly with regard to Markel and QBE. Mr Gruder himself conceded that unless he could undo the judge’s findings with respect to Mr Higgins’ involvement in the creation of the false bordereaux, his appeal was in difficulties. Cumulatively, however, these components become overwhelming, despite everything that can be said on behalf of Mr Higgins as, unfortunately, a sufferer from Alzheimer’s. It is not in dispute on this appeal, if it ever was, that there was a conspiracy, and that at least Mr Felstead and Mr Brunswick were conspirators; it can no longer be in dispute that Mr Smith did not sanction the over-writing of the out-of-limit bonds or a silent co-surety arrangement with Templeton; there was no silent co-surety arrangement with Templeton, or none that went beyond the dishonest involvement of Mr Brunswick; and the three men who were behind and interested financially in SGC were the same three men who were interested in GCL, to which unaccounted for premium was transferred by a combination of Mr Higgins himself and Mr Brunswick.

In sum, the misfortune of Alzeimer’s may have provided Mr Higgins with a legitimate defence of forgetfulness (to some degree, but it must not be forgotten that the medical evidence was that such forgetfulness would only have been intermittent). However, it is striking that at trial Mr Higgins did not pray in aid his Alzheimer’s as explaining his underwriting, only as enabling Mr Felstead to take advantage of him. However, Mr Felstead was his colleague and the man whom he was willing to associate with himself both in SGC and in GCL; also the man whom he was willing to keep on within the business even when he, Mr Felstead, was supposed to have left. Mr Higgins’ ethical judgment, may, regretfully, have failed with increasing age and illness, but it is impossible to say that the judge was wrong to find that he was part of a conspiracy to defraud Markel and QBE.

On that basis it is conceded by Mr Gruder that Mr Higgins is liable at least on the ground of conspiracy to defraud and/or dishonest assistance in breach of fiduciary duty, and thus responsible for such damages as may have been assessed against him in the quantum proceedings.

I would therefore dismiss this appeal.

I agree.

I also agree.