On 22nd October 2007 at the Crown Court at Southwark, before His Honour Judge Rivlin QC and a jury, after a trial lasting many weeks, Michael John Bright was convicted on two counts of conspiracy to defraud (count one and count five). This is his appeal against the sentence of 7 years’ imprisonment on each count to run concurrently. He was also disqualified under section 2 of the Company Directors Disqualification Act 1986 for 12 years, but no appeal is brought against the disqualification order.

Philip Condon and Dennis Lomas, were also convicted, Condon on one count of conspiracy to defraud (count one) and Lomas on the same counts as Bright. Condon was sentenced to 3 years’ imprisonment, and Lomas to 4 years’ imprisonment on each count to run concurrently. They were both disqualified under section 2 for a period of 10 years.

These convictions arose from the notorious collapse of the Independent Insurance Group. Independent Insurance PLC was a public company quoted on the London Stock Exchange. Independent Insurance Limited was its principal trading subsidiary. The appellant started the business in 1988. It developed into the 9th largest general insurance concern in the United Kingdom. In June 2001, both companies were put into liquidation.

The collapse was one of the most serious commercial disasters to have occurred in recent years. Although the liquidator’s final accounts are unavailable, it is likely that the deficiency will be in the region of £1 billion. Approximately 1000 employees lost their jobs. Shares in the company became valueless. Not far short of £400 million has already been paid to policy holders by the Financial Services Compensation Scheme funded by the insurance industry, but not all claims will qualify for compensation.

It was never suggested that the company had been set up as a vehicle for fraud. The dishonesty began as a response to trading difficulties. What was unknown at first, but gradually became apparent to the conspirators, was that the company had been unprofitable over a period of years. It was then that the dishonesty began. The appellant was the Managing Director and Chief Executive Officer of the Independent Insurance Group of Companies. At the time when he was committing these offences, he was President of the Institute of Insurers. Condon was the Deputy Managing Director, and Lomas the Finance Director. The moving force behind the initial success of the business, and then the driver of the conspiracy was, as the judge found, this appellant.

The essential facts can be briefly summarised. This was a public company. Huge sums of money flowed in and out of it daily. The defendants were in a position of trust. Over a period covering at least two year end published accounts for 1999 and 2000, they dishonestly manipulated the company’s reserves. These were described by both Condon and Lomas as “woefully inadequate” and the defendants arranged to hide the need to improve the reserves by keeping claims data off the computer system and the company’s books of account. As Judge Rivlin summarised it, in the very simplest terms, this was achieved by instructions issued to senior employees so to manage the business that very many millions of pounds worth of claims were put on lists which were kept separate from the accounting systems. In this way they were concealed from the company’s independent actuaries. Accurate information was of course required to enable the actuaries to make appropriate projections and advise both the company, the shareholders, and the public at large accurately about the state of the reserves. Effectively the actuaries were deliberately and dishonestly tricked, and basing their certification upon the representations made by the defendants in their capacity as directors, they certified that the reserves were sufficient, when the defendants knew perfectly well that the certification on which so many relied was the product of deliberate and systematic fraud.

The second aspect of dishonesty was reflected in count five. Once the company’s considerable financial difficulties were appreciated, the appellant and Lomas negotiated three very substantial reinsurance contracts which had a corresponding beneficial effect on the accounts. These contracts were underpinned by several further, but unfavourable contracts. Assurances were given that the only contracts were the favourable ones. This was deliberately false. The effect was to alleviate concerns which would otherwise have troubled the company’s actuaries and accountants. In the result, whereas the board of directors, and the public at large, were given detailed information about the beneficial contracts, the unfavourable contracts were entirely concealed.

In his report for the year end 2000, the appellant was able to assert that the business had a “unique level of comfort”. This was nonsense, and the appellant knew it. The 2000 year end accounts were manipulated into showing a profit before tax just in excess of £22 million. In reality if the accounting material relating to case estimates had been included in the accounts, this profit turned into a loss of fractionally short of £30 million. As Judge Rivlin explained in his sentencing remarks

The judge was prepared to accept that these dishonest activities were not directly responsible for the collapse of Independent Insurance. However he was amply justified in reaching the conclusion that, as he put it, the effect of the conspiracies was to “put paid to any chance of mounting a legitimate rescue operation by the raising of fresh capital and, as the fraudulent conduct was prolonged, it must have played a significant part in the scale of the financial disaster that ensued”.

Numerous different categories of people were deceived. They included insiders, such as employees, other directors of the company, and the company’s accountants and auditors, and those outside the company, such as shareholders, creditors and policy holders. The scale was vast. The fraudulent activity continued despite express warnings from other directors. As Judge Rivlin pointed out the dishonesty was set in motion by the appellant as the “architect and driving force” behind the fraud, who determined on a course of dishonesty which “steamrollered out of control under his clear instruction and with his full knowledge”. Judge Rivlin went on to observe that “the truth is that you corrupted a lot of people along the way including, I believe, your co-defendants”, and he described the “breezy manner” in which the appellant sought to blame able decent and hardworking employees for dishonest practices which he had himself introduced and put into operation. The appellant was the individual who led the dishonesty to a “new and exceedingly serious level” when he played the “hands on” control and leading role in the reinsurance aspect of the fraud covered by the second conspiracy count.

With that very brief summary of a complex fraud, the investigation into which required analysis of a million documents, we must come to the heart of the appeal. It is submitted by Mr Ian Winter QC on behalf of the appellant, that the sentence imposed on him was the maximum available sentence. For a variety of reasons, such a sentence was wrong in principle, and produced an overall sentence which was manifestly excessive. Pausing for a moment, however, before examining his argument, we should record that commendable efforts were made by all counsel and solicitors in this case to cooperate with the judge to enable him to exercise his case management responsibilities and to produce a trial which, notwithstanding the vast amount of material, took no longer than necessary.

At first the defendants’ criminality was embodied in a single count of conspiracy to defraud. The Crown decided to add a number of substantive counts of fraudulent trading contrary to section 458 of the Companies Act 1985 to cater for the possibility that the jury might conclude that although no conspiracy was proved between the defendants, one or other of them was guilty of fraudulent trading. It also became apparent that the original conspiracy to defraud count covered dishonest conduct which had two distinct and separate features. It is an essential feature of the issues to be examined in the appeal that, at the outset of the trial, the chief concern of the Crown was the effective management of the case. Accordingly appropriate amendments to the indictment were allowed.

At the end of the trial the jury convicted the defendants of conspiracy to defraud. With the agreement of the prosecution, Judge Rivlin accepted that it would be inappropriate for him to pass consecutive sentences in relation to the two counts of conspiracy to defraud. That said, however, on ordinary sentencing principles, it was an aggravating feature of the case against the appellant and Lomas that were involved in fraud which had two distinct manifestations.

The problem which arises in this appeal is that Judge Rivlin accepted a submission on behalf of the defendants that as there was no effective difference between these particular conspiracies to defraud and what would have been conspiracies to trade fraudulently. For such offences the maximum available sentence was not 10 years’ imprisonment for conspiracy to defraud, but 7 years, for what would have been conspiracies to trade fraudulently. When passing sentence, Judge Rivlin carefully explained that he had,

The first question which arises is whether Judge Rivlin’s analysis of the maximum sentence available to him was correct, and if not, whether, in the context of an appeal against sentence, we are nevertheless bound by it. It is no disrespect to Mr Winter’s forceful argument on behalf of the defendant that in reality his submission that the maximum sentence was indeed 7 years’ imprisonment represented a reworking of the submissions made to Judge Rivlin, before the trial proper began, that the jury should not be invited to consider the counts of conspiracy to defraud just because these allegations did not in fact extend beyond the ingredients of the statutory offence.

At the preparatory hearing in March 2007, during the discussions about the form of indictment, Judge Rivlin expressed the view that “the charges of conspiracy to defraud were “really tailor made” for the allegations that the Crown is making in this case”. He immediately recognised that, as he put it “simply and bluntly”, one matter in Mr Winter’s mind would be the differences in the maximum penalty available for the different offences.

At the hearing on 1st May 2007 consideration was given to the then new guidance issued by the Attorney General on the use of the common law offence of conspiracy to defraud. Mr Winter submitted that the guidelines should be reflected in counts relating to the substantive offence, rather than conspiracy to defraud. When the judge canvassed the possibility of the indictment alleging substantive offences of fraudulent trading coupled with conspiracies to the same effect, Mr Andrew Baillie QC made clear that the Crown was not advancing submissions about the correct form of the indictment on the basis of any sentencing eventualities if the defendant were convicted. He explained the reason for the “conspiracy to defraud route” was appropriate because there was no “criminal allegation against the other directors of the company, and where the company is a public company, and what is said is (an) allegation of two very specific allegations, we prefer to put that as an agreement to put other people’s interests at risk, conspiracy to defraud, rather than to describe it as a conspiracy to trade fraudulently”.

The Crown therefore maintained the position that the counts of conspiracy to defraud should be maintained. No one had suggested that to do so would create some kind of unfairness or prejudice to the defendants in the forthcoming trial. There was no such risk. In any event, the conspiracy count would be the natural way to describe the conduct alleged against the defendants, “where the company and its shareholders and others concerned with the company are the alleged victims”, and “fraudulent trading” whether as a conspiracy or a substantive offence, would not be natural. “It is of course the law that one can carry on a business of a limited company for a fraudulent purpose, which are the key words, if one is engaged in defrauding the company and its shareholders and creditors, and so on. So it does in fact as a matter of law fit these facts. But it is not such a natural description of what happened here.”

Mr Winter submitted that he did not understand Mr Baillie’s argument. Giving a short ruling, Judge Rivlin believed that conspiracy to defraud charges were “tailor made” for the case. He reached that conclusion having considered R v Rimmington:R v Goldstein[2006] 1 AC 459 and the Attorney General’s guidance. He was “entirely satisfied” that charges of conspiracy to defraud expressed “most succinctly, clearly and naturally what it is that the Crown is seeking to prove.” The judge offered to give a fuller ruling if it was required.

Following a request that he should do so, on 30th May in a much more detailed judgment, after setting out a detailed narrative of the way in which the indictment had developed. Judge Rivlin recorded that it had never been suggested that conspiracy counts were either unnecessary or inappropriate. It had always been accepted, in Judge Rivlin’s view “rightly”, that the conspiracy charges were properly made, and that it was also appropriate for alternative substantive charges against each defendant alone to be included. He recorded that it had been and remained his “strong instinct” that “the crime of conspiracy to defraud was really “tailor made” for what the Crown was alleging - both in relation to the alleged conspiracy to withhold claims data from the actuary and also in relation to that known as “the reinsurance” conspiracy.” That, effectively, was his ruling. In his judgment he made an observation about the maximum penalties, and the difference between the maximum penalty for conspiracy to defraud which was greater than the penalty for fraudulent trading, but noted that the penalty for conspiracy to defraud was not unlimited and, as a maximum, was 10 years imprisonment.

The trial proceeded. The jury convicted the defendants of conspiracy to defraud. None of them has appealed against conviction. The rulings about the indictment were made by a pre-eminent judge in this specialist field, and they were and remain unassailable.

Before sentence submissions were advanced on behalf of the appellant, in effect repeating the submission in relation to the indictment, in the context of sentence. In short, there was no principled distinction between the counts of conspiracy to defraud of which the appellant was convicted, and what would have been conspiracies to trade fraudulently. “The court should approach sentence on the basis that the maximum sentence for counts one and five should properly be 7 years’ imprisonment. All parties accept that the facts supporting counts one and five fall within the statutory offence of conspiracy to trade fraudulently, which would carry a maximum of 7 years’ imprisonment (see R v Bourgass [2006] EWCA Crim 3397….) In all the circumstances, it would not be proper to pass consecutive sentences. Originally, the indictment contained a single count. The prosecution’s approach, therefore, was that the criminality amounted to a single transaction. That single count was divided in two (counts one and five) for trial management reasons only”.

Judge Rivlin raised with counsel for the Crown whether the case should be treated as a single transaction, and Mr Baillie made clear that in broad terms the original single counts were separated into two counts for “case management reasons”. During his submissions he said that the decision was not “taken…deliberately for the purpose of extending the sentencing judge’s powers to more than one count”. Judge Rivlin did not apparently seek assistance, and nothing said at this stage by Mr Baillie amounted to a concession or agreement by the Crown that the maximum sentence was limited to 7 years: he simply and rightly accepted that consecutive sentences would be inappropriate. When, as recorded in paragraph 13, the judge referred to concessions made by the prosecution in this context, he was under a misapprehension. In our judgment the conclusion that the sentences following these convictions were limited by the maximum sentence which would be imposed for an offence or offences of fraudulent trading was wrong.

We recognise that it may sometimes be appropriate for an offender who has been convicted of an offence which falls within the definition of a different offence with a lower maximum sentence to be sentenced as if that lower maximum applied. Examples are R v Quayle [1993] 14 CAR (S)726 and R v Blair [1996] 1 CAR (S) 336, where the idiosyncrasies of the legislation governing sexual offences meant that a defendant could be prosecuted for indecent assault when he was in fact guilty of unlawful sexual intercourse, but not prosecuted for it because the prosecution was brought outside the 12 month time limit which applied to sexual intercourse but not indecent assault. Accordingly, Keene J giving the judgment of the court in Blair, concluded that there was “some force in that argument that where unlawful sexual intercourse itself can attract a penalty of no more than 2 years’ imprisonment, indecent assaults on a defective, which falls short of sexual intercourse, need to pay some regard to that maximum of 2 years’ imprisonment despite the fact that 10 years is technically available”.

In R v Bourgass [2006] EWCA Crim 3397 the defendant was convicted of conspiracy to cause a public nuisance. He was sentenced to 17 years’ imprisonment. It was submitted that the maximum available sentence was 14 years’ imprisonment under section 113 of the Anti-Terrorism Crime and Security Act 2001. This was the appropriate statutory offence committed by those taking action which involved the use of noxious substances which endangered human life and was designed to, in effect, influence the government or intimidate the public. As sentencing for conspiracy to commit a public nuisance was “at large”, Lord Justice Latham, the Vice President, reflected on Rimmington and Goldstein. He acknowledged the correctness of the submission that “the court should be guided in determining what is the right sentence by any statutory offences which could sensibly be said to cover, or, at least, parallel, the situation in any given case”. That, we emphasise, is the language of guidance, not prescription.

For present purposes it is noteworthy that the maximum sentence for conspiracy to defraud was set by section 12 (3) of the Criminal Justice Act 1987 after the enactment of section 485 of the 1985 Act. Nothing in the legislative structure suggests that conspiracy to defraud may not be prosecuted whenever fraudulent trading is integral to the conspiracy. Section 458 of the Companies Act 1985 is a very broad general offence, covering many types of fraudulent misconduct in connection with the trading business of a company and this conspiracy certainly included fraudulent trading on an exceptional scale. The counts of conspiracy to defraud were appropriate charges, not only for case management reasons, as well as those given by Judge Rivlin at the various preparatory hearings, but also because of the sheer magnitude of the defendants’ criminality. This appellant was convicted of two counts of conspiracy to defraud, properly brought and prosecuted, and properly tried. It was indeed a classic offence of this kind. Accordingly, both as a matter of law and practice, the maximum available sentence was 10 years’ imprisonment, not 7 years’ imprisonment

The next question which arises is whether on this appeal against sentence, it is open to us to go behind the maximum sentence publicly announced by Judge Rivlin. This is not a case in which the Crown made representations to the court, or to the appellant, about the maximum sentence on which he was entitled to rely, and even if the Crown had made any such concession, the responsibility of the court is to consider whether the sentence under appeal was wrong in principle or manifestly excessive on the basis of a correct, not a mistaken view of the maximum sentence. If it were otherwise, the court would be permitting a judicial error to curb its sentencing jurisdiction, and indeed to preclude a reference by the Attorney General on the ground that the judge made a mistake about the maximum sentence. The safeguard for the defendant is that, whatever error the judge may have made, save in the context of such a reference, the sentence may not be increased. In such references, of course, the court would exercise its discretion in the light of representations made by the Crown, if any, whether to the defendant, or to the court, which may have caused the appellant to act to his detriment or may otherwise have given rise to a legitimate expectation that the Crown would not resile from its position. Nothing like that arose here. We must therefore examine this sentence in the light of the statutory maximum as it is, not as it was thought to be.

In deference to Mr Winter’s submissions we shall, in any event, address the argument based on the proposition that the maximum sentence was indeed 7 years’ imprisonment, and that the maximum sentence should not have been imposed.

The maximum sentence permitted by statute is, of course, very rarely imposed, and nowadays when there has been a guilty plea, effectively never. Such sentences should be reserved for those cases which, at the end of the trial and within the statutory context, can fairly be regarded as crimes of the utmost gravity. It is sometimes loosely said that the maximum sentence should be reserved for the worst case of its kind, and from this, imaginative counsel for the defendant will urge examples of cases of greater criminality than the offence established against his client. The argument however is founded on the misapprehension that if a realistically more serious case can be imagined, the imposition of the maximum sentence is precluded. That is why we repeat, the maximum sentence permitted by statute is reserved not for the worst possible case which can realistically be conceived, but for cases which in the statutory context are truly identified as cases of the utmost gravity.

The principle is long established. In R v Amber and Hargreaves, unreported, November 24, 1975, but referred to in Current Sentencing Practice at A1-4CO1 a case involving offences of corruption in the context of bribery of prison officers, Lawton LJ observed:

Mr Winter submitted 5 reasons why this particular case did not merit the imposition of the maximum sentence. The criminal activity of the appellant and his co-conspirators did not cause the company to collapse. None of them, to use colloquial language, raided the till. None of them intended to cause loss, but rather to trade the company through a difficult period to better times. The fraud began almost imperceptibly, and grew under pressures on vast claims coming from the London Market, for which the appellant was not responsible, and of which he was ignorant. That submission was true up to a point, but could not be said to apply after November 1998. The appellant was also said to have continuously purchased shares in the company, and he was the biggest loser at its collapse. His own loss was £55 million. That again may be true, but the appellant’s determination to see the company through by dishonest behaviour was at least in part motivated by his own obsessive determination to save the company which he had founded, by whatever means were available, and of course, as with his co-conspirators, to continue to enjoy the fruits of what had been, but was no longer, a very successful business.

Mr Winter drew our attention to the observations of this court in R v Paulk and Smith where the court was considering sentences of 3 years imprisonment imposed on two defendants for offences of fraudulent trading. Potter LJ (as he then was) pointed out that there was no doubt that

Judge Rivlin was fully entitled to express himself in the trenchant terms he did. This was a conspiracy to defraud, for a prolonged period, with dire consequences for many individuals. It was indeed, even within the limited statutory context of section 485 of the 1985 Act, a crime of the utmost gravity.

Mr Winter submitted that the sentence ignored the appellant’s personal mitigation. He drew attention to the appellant’s age (63 years) his good character, his poor health, the medical condition of his wife, for whom he was caring before he started his sentence. He also submitted that the appellant lost literally everything for which he had worked, not least the destruction of the business which had soared to the heights from nowhere, not least because of his commitment and contribution to it. In summary, his life is in ruins. A maximum sentence, by definition, must have ignored all these elements of personal mitigation, and some allowance should have been made for them.

Our conclusion is that in this particular case not much, if any, allowance would be appropriate. It is not unknown for major white collar fraudsters - and that is what the appellant is – to be individuals of apparent impeccable rectitude and good character. Indeed this very reputation helps them to establish this standing in business which, for a variety of different reasons, later becomes the vehicle for fraud. At the time when he was committing these offences, he was a President of the Institute of Insurers, apparently a pillar of respectability. Suspicion would not fall on him, and if it did, could be brushed aside as unworthy. It is always sad to see a successful individual fall from heights which he has achieved through his own hard work. But the appellant’s criminal behaviour means that all his wounds were and are self inflicted.

A disparity argument was faintly mounted. We can see none. More important, the judge had absolutely no hesitation in deciding which of the three defendants before him carried the greatest burden of culpability, and led this dishonesty. There was no disparity. The differences in sentence reflected the different criminality of the defendants.

It was suggested that some allowance should have been made for the delay between the first investigations with the appellant and the date of sentence. He was first interviewed in January 2002 by the liquidators of the company, and by the Financial Services Authority in November 2004. The prosecution, applying that term broadly, was not responsible for any delay. The investigation involved study of literally a million documents. During his interviews the appellant was concerned to conceal the truth rather than reveal it. The case therefore had to be prepared and it eventually proceeded as a trial. The practice of allowing a discount against sentence to a defendant who pleads guilty is pragmatic. In this case an honest response to the interview process, and an indication of a guilty plea, would have avoided a huge investigation and a mammoth trial. It would be absurd for the defendant whose criminal activity requires massive investigation and whose lies contribute to the delay in bringing him to justice, then to seek to take advantage of the delay as a feature in mitigation.

We have reflected on Mr Winter’s submissions as they relate to what we shall describe as the maximum sentence issue. If we had been satisfied that the maximum sentence was indeed 7 years’ imprisonment, we very much doubt whether it would have been appropriate to order any reduction. In the light of his age, and ill health, and his wife’s ill health, some small reduction from the notional maximum might have been allowed as an act of mercy, but the judge was entitled to conclude that in view of the appellant’s conduct, he had forfeited any mercy which the court might otherwise have extended to him. However, and in any event, the actual sentence has effectively made every possible allowance for these matters, and represents a relatively significant discount from the available maximum sentence.

The last ground of appeal is based on the submission that the judge intended to achieve a custodial term of 3½ years imprisonment, and that the eventual sentence failed to achieve his objective. The judge sentenced the appellant to 7 years imprisonment on each count. The appellant was told he would serve half that period in custody, and half on licence. The appellant was in fact subject to the sentencing regime created by section 33 of the Criminal Justice Act 1991 not, as the judge thought, the Criminal Justice Act 2003 regime. Therefore it did not follow that the appellant would automatically be entitled to be released at the half way point of the sentence. The error was corrected on the same day. The judge said that he had no intention of altering the sentences imposed on any of the defendants. He was “completely satisfied” that he would not have imposed any different sentence if he had had the provisions of the 1991 Act rather than the 2003 Act in mind. He therefore amended what he had said, to make it clear that the appellant would have to serve “at least” one half of the sentence in prison. Thereafter release would be at the discretion of the Secretary of State in the recommendation of the Parole Board. He added that he himself would expect that unless circumstances arose of which he had no knowledge, there “is a very strong likelihood indeed that the defendant would be released at that stage”.

Mr Winter submitted that the judge should have applied the terms of the Practice Statement (Crime: sentencing) [1992] 1WLR which provided that after 1 October 1992 it was necessary for the Crown Court when passing a custodial sentence “to have regard to the actual period likely to be served”. This Practice Statement does not bear the weight Mr Winter sought to put on it. It was directed to sentencers in order to highlight the then new statutory arrangements relating to remission and parole, and their impact on the times actually to be served when compared with sentences imposed before the changes came into force. The context was the coming into force of sections 32-40 of the 1991 Act which altogether abolished remission and removed parole from consideration in sentences of less than 4 years imprisonment. In these cases, the prisoner would be released at the half way point. Sentences longer than 4 years involved release on licence after two thirds of the sentence had been served, but not remission, and eligibility for parole half way through the sentence. In short, whereas if he had been sentenced under the provisions of the 2003 Act, the appellant would automatically be entitled to release when he had served 3½ years, to be followed by a period of 3½ years licence, a 7 year sentence to which the 1991 Act applied means that he will be eligible for, but not automatically entitled to be released at the end of 3½ years

From this Mr Winter sought to argue that as the judge intended a 3½ year sentence actually to be served, the sentence should in any event be reduced to 5¼ years. The submission is based on a fallacy. The actual sentence was 7 years imprisonment. The release provisions did not and should not have affected the judge’s sentencing decision. What he was required to do was to explain the effect of the sentence in the context of the applicable statutory provisions relating to release. He did not “intend” that the appellant should be released after 3½ years: that would simply have been the consequence if the 2003 Act had applied to the sentence, and he was required to state that consequence in open court. Precisely the same applies to his observations when reconsidering the sentence. He knew that the appellant would be eligible for release after 3½ years, but not automatically entitled to it. He was entitled to say, as a judge of considerable experience, that it would be very likely indeed that the defendants would in fact be released at the end of 3½ years. Again, that did not mean that he so “intended”. The sentence of 7 years’ imprisonment is not open to question on the basis that the order achieved a different result to that which the judge intended. Indeed he himself made it absolutely clear that he had no intention whatever of changing the sentence which he decided to impose.

This appeal against sentence is dismissed.