Court of Appeal, Civil Division

Society of Lloyds v Jaffray and others

[2007] EWCA Civ 586, [2007] All ER (D) 213 (Jun)

Philip Jenkins (instructed by Grower Freeman) for the applicants not in person.
Two applicants appeared in person
David Foxton (instructed by Freshfields Bruckhaus Deringer) for the respondent.
James Wilson Barrister (NZ).

JUDGES: Buxton and Moore-Bick LJJ

DATE: 20 June 2007

Court of Appeal — Jurisdiction — Application to reopen appeal — Application based on alleged new evidence showing fraud practised on trial — Whether Court of Appeal having jurisdiction to reopen appeal on basis of new evidence concerning fraud below — Whether on facts appeal should be reopened — Civil Procedure Rules 1998, SI 1998/3132, CPR 52.17.

Part 52.17 of the Civil Procedure Rules 1998, SI 1998/3132, provides: ֥The Court of Appeal or the High Court will not reopen a final determination of any appeal unless—(a) it is necessary to do so in order to avoid real injustice; (b) the circumstances are exceptional and make it appropriate to reopen the appeal; and (c) there is no alternative remedy.’

The action formed part of the ‘Lloyd’s litigation’ between the Society of Lloyd’s (Lloyd’s) and some of its ‘Names’. The Names were underwriting members of Lloyd’s who subscribed to policies of insurance through membership of one or more syndicates for particular years of account (each syndicate operating with a given membership for one calendar year only) and whose personal liability in respect thereof was unlimited. The instant case was concerned with the ‘threshold fraud point’. The issue was whether Lloyd’s had made misrepresentations which it knew to be untrue and/or as to which it was reckless as to whether they were true or false, and whether such misrepresentations were communicated to the Names and, if so, when. The Names’ case was in tort for fraudulent misrepresentation. They claimed that certain alleged representations derived from the brochures issued by Lloyd’s in 1981 and subsequently were false and fraudulent to the knowledge of Lloyd’s. Lloyd’s denied that they made the alleged representations in the brochures and the accounts, and, further, argued that no elements of the tort of deceit were made out. The judge held that no relevant representations had been made (see [2000] All ER (D) 1674). He further held that, even if the alleged representations had been made, the names had not proved that Lloyds did not believe them to be true, nor had the names proved that Lloyds had been reckless as to whether the representations were true or false. The Court of Appeal ruled that the representations had in fact been made, and that they had been untrue. It upheld the judge, however, in relation to the latter issues and therefore dismissed the appeal (see [2002] All ER (D) 399 (Jul)). The names applied to reopen the decision of the Court of Appeal under CPR 52.17, relying on Taylor v Lawrence [2002] 2 All ER 353. They contended that they had acquired evidence that showed that the judge below had been misled by the evidence of Lloyds at the trial as to the state of its knowledge and belief, and by necessarily basing its ruling on the judge’s findings, the Court of Appeal had been similarly misled.

The names argued that since the application was made under CPR 52.17 the court should start from, and apply, the plain wording of that rule, and in particular the reference in CPR 52.17(1)(a) to appeals being reopened ‘in order to avoid real injustice’. That general rule prevailed over any previous jurisprudence that might be argued to limit the jurisdiction to any particular cases, such as where the earlier decision had been obtained by fraud.

The application would be refused.

It was trite law that the CPR could not extend the jurisdiction of the court from that which the law provided, but could only give directions as to how the existing jurisdiction should be exercised. In fact CPR 52.17 , which referred to the jurisdiction ‘not’ being exercised ‘unless’ various conditions were met, was passed to limit rather than extend the supposed jurisdiction. Previous authority precluded the reopening of a determined appeal on the ground of fraud (the correct procedure being an action to set aside the original judgment), although some cases had assumed that it was possible. If an appeal was to be opened on that basis, it would be required (i) that the new evidence was admissible under Ladd v Marshall principles; (ii) that, on the basis of that evidence, it was probable that corruption had in fact occurred; and (iii) that there was a strong probability that that corruption had affected the result of the first trial.

On the facts of the instant case, the supposed new evidence fell far short of making out even an arguable case.

The applications would be dismissed, though the restriction on citation would be lifted.

Taylor v Lawrence [2002] 2 All ER 353; Flower v Lloyd (1877) 6 Ch D 297 and Ladd v Marshall [1954] 3 All ER 745 considered; Couwenbergh v Valkova [2005] All ER (D) 98 (Feb) distinguished.

[2007] EWCA Civ 586


20 JUNE 2007





1. These are unusual applications in an unusual case. The applicants seek, under the jurisprudence of Taylor v Lawrence [2003] QB 528, to reopen the decision of the court in this case [2002] EWCA Civ 1101. This court dismissed an appeal from Cresswell J [2000] All ER (D) 1674 in proceedings in which these and other applicants sought remedies in deceit against Lloyds [sic]. The terms of the representations said to found that deceit are in issue in one respect in this application, and we return to the point in §§ 54 and following below. For the moment, however, it suffices to say that the claim in this case, in the respects relevant to these applications, was that in the “brochure” published by Lloyds in 1981 and later years representations were made that Lloyds operated a rigorous system of auditing that enabled reasonable estimates to be made of the outstanding liabilities of syndicates, including unreported and unknown losses, and thus proper reserves to be made against those potential losses. On the strength of such representations the applicants, including the lead claimant in the action Sir W Jaffray, became underwriting members ("names") of Lloyds. They suffered serious financial losses by becoming liable to compensate insureds in respect of the latter’s liabilities for injury caused by earlier exposure to asbestos, the existence and extent of such pending liabilities not having been identified by the Lloyds accounting system, and thus not properly reserved against.

2. Cresswell J held that no relevant representations had been made. He was reversed on that issue by this court, which held that representations in the terms set out above had been made. This court further held (with, as the court said at §375 of its judgment, the benefit of hindsight) that those representations had been untrue. Cresswell J also held that, even if the alleged representations had been made, the names had not proved that Lloyds did not believe the representations to be true; nor had the names proved that Lloyds had been reckless as to whether the representations were true or false. This court upheld Cresswell J on those latter issues, and therefore in the event upheld his dismissal of the claims.

3. The names now assert that further evidence has come to light that shows that Cresswell J was misled by Lloyds evidence at the trial as to the state of its knowledge and belief, and thus this court, which necessarily based itself on that same evidence, was similarly misled. It will be convenient to mention here that the further evidence on which reliance is placed in this application is virtually the same as the evidence that is sought to be introduced into another appeal arising out of many of the same events and involving many of the same parties, Lloyds v Henderson [2005] EWHC 850 (Comm). In those proceedings Lloyds claims against various names in respect of liabilities as a result of policies written by them, which claims the names seek to resist because of misbehaviour, including fraud, on the part of Lloyds. In that action the names have an outstanding appeal to this court against a decision of Andrew Smith J refusing them permission to amend their pleadings to allege Misfeasance in Public Office against Lloyds (the alleged acts of misfeasance closely mirroring the acts of deceit alleged in this case), which also involves an application by the Names for permission to rely on the fresh evidence. That appeal is to be heard in the last week of July. It should have been heard last November, but was adjourned because on the day before the hearing the names parted company with counsel then instructed to argue the appeal. The present, Taylor v Lawrence, application had been filed at or about the same time as the appeal in Lloyds v Henderson, but the Taylor v Lawrence application was regrettably overlooked in the court’s filing system, a matter for which the court has apologised to the names and to their representatives. That is why it only comes on now. The major number of the names concerned, conveniently referred to as the UNO names, are represented by the same solicitors and counsel both in Lloyds v Henderson and in this case.

4. This Taylor v Lawrence application is being heard in open court, and not disposed of simply on paper as is the usual practice, because it arises in a long-standing and highly contentious dispute, and also because of its substantial links with the appeal in Lloyds v Henderson. It must by no means be thought that this process, and the elaborate exchanges to which it has given rise, creates any sort of precedent for other Taylor v Lawrence applications, which will continue to be discharged by the summary process mandated by CPR 52.17.4.

5. Before us the UNO names made common cause, represented by Mr Jenkins of counsel. Two applicants addressed us on their own behalf, Mrs Adams, for whom her husband spoke, and Mr Butler. In response to a direction of the court that sought to impose some order on the proceedings the UNO names had submitted a 27 page statement of their case, and also, in addition to making reference to the further evidence that has been produced in Lloyds v Henderson, submitted a further witness statement, running to 34 pages of single-spaced typing, from one of their principal witnesses, Mr Stephen Merrett. Further, on the day before the application was heard the names submitted a further substantial statement, from which it became clear that the focus of the argument had shifted somewhat since its original formulation. In reply to the names’ written arguments, Lloyds submitted a 56 page document, which additionally referred to the extensive written arguments in the forthcoming appeal in Lloyds v Henderson.

6. In order properly to understand the issues in this application it is necessary first to say something in general terms about the jurisprudence of Taylor v Lawrence.

Taylor v Lawrence and CPR 5.17

7. Mr Jenkins argued that since the application was made under CPR 5.17 the court should start from, and apply, the plain wording of that rule, and in particular the reference in CPR 52.17(1)(a) to appeals being re-opened "in order to avoid real injustice". That general rule prevailed over any previous jurisprudence that might be argued to limit the jurisdiction to any particular category of cases, for instance where the earlier decision had been obtained by fraud. Accordingly, the court should not take time with analysis of Taylor v Lawrence itself, or of the cases underlying it, but should ask itself whether this appeal should be reopened in order to avoid real injustice in a broadly discretionary, essentially palm-tree, frame of mind.

8. That approach is quite misconceived. The CPR, being rules of court, cannot extend the jurisdiction of the court from that which the law provides, but can only give directions as to how the existing jurisdiction should be exercised. That is very trite law, but if authority is needed for the proposition it can be found in the speech of Lord Herschell LC in British South Africa Co v Companhia de Mocambique [1893] AC 602 at p 628. And quite apart from that general rule, it is apparent from the wording of CPR 52.17(1) (which speaks of the jurisdiction not being exercised unless various conditions, including avoidance of real injustice, are fulfilled) that, as the helpful commentary in vol 1 of Civil Procedure explains, it was passed to limit, and not to extend, the operation of the supposed jurisdiction under Taylor v Lawrence.

9. The applicants cannot therefore avoid, any more than can this court avoid, detailed analysis of the extent of this court’s jurisdiction to re-open determined appeals.

The jurisdiction to re-open determined appeals: a summary

10. Taylor v Lawrence concerned (alleged) misconduct by a court, in that the judge was said to have been biased. There is no authority in that case for extending the recognition of jurisdiction to re-open an appeal on grounds of bias to a case such as the present, where the allegation is not that the court misbehaved, but that the court was misled by one of the parties; and there is authority directly denying the existence of jurisdiction in the latter case both in this court (Flower v Lloyd (1877) 6 ChD 297) and in the House of Lords (Jonesco v Beard [1930] AC 298), it being held in both of those cases that the proper remedy is to bring a collateral action to set aside the judgment allegedly obtained by fraud.

11. We have not felt able to act on that conclusion in the present application, even though what is in issue is a matter of jurisdiction, first because the point was not argued; and second because in a series of cases since Taylor v Lawrence this court has assumed, without the matter being argued out, that the jurisdiction of Taylor v Lawrence does extend to cases of fraud as well as to cases of bias. However, even on that assumption, there is no doubt that the very stringent limits stated in Taylor v Lawrence, that the jurisdiction should only be exercised in exceptional cases where there is no alternative remedy, apply just as much in cases of fraud as they do in cases of bias.

12. We now explain those propositions in some greater detail.

Taylor v Lawrence in a case of bias

13. In Taylor v Lawrence an appeal against the order of the court below was founded on an allegation that the trial judge had been relevantly biased. After that appeal had been dismissed, further evidence emerged, which was said to demonstrate the judge’s partiality to one of the parties more clearly than the evidence relied on in the original appeal. The first, and indeed main, issue that confronted this court was whether, despite the general principle of finality of judgments, it had jurisdiction to re-open a final determination of the Court of Appeal in a case where bias was alleged on the part of the lower tribunal. Although that issue was debated at some length, the answer to purely jurisdictional questions in a case of bias was to be found in the decision of the House of Lords in Pinochet (No2) [2000] 1 AC 119, in the speech of Lord Browne-Wilkinson at p132. The House of Lords relied on its jurisdiction to control its proceedings, unfettered by any statutory limitation. This court recognised in Taylor v Lawrence that, unlike the House of Lords, the Court of Appeal’s jurisdiction is wholly created by statute, but held that nonetheless, like any other court, it had inherent power to take steps to maintain the court’s character as an (unbiased) court of justice: see per Lord Diplock in Bremer Vulcan [1981] AC 909 at p977, cited to this effect by Lord Woolf in Taylor v Lawrence at §52. Accordingly, the Court of Appeal could take the same approach to the corruption of the judicial process by bias as had the House of Lords in Pinochet.

14. That, however, was as to the existence of the jurisdiction. As Lord Woolf pointed out, it is a very different question to ask how that jurisdiction should be exercised. The court gave guidance on the exercise of the jurisdiction in §§ 54-55 of Taylor v Lawrence:

There … needs to be a procedure which will ensure that proceedings will only be reopened where there is a real requirement for this to happen. One situation where this can occur is a situation where it is alleged, as here, that a decision is invalid because the court which made it was biased. If bias is established, there has been a breach of natural justice. The need to maintain confidence in the administration of justice makes it imperative that there should be a remedy. The need for an effective remedy in such a case may justify this court in taking the exceptional course of reopening proceedings which it has already heard and determined. What will be of the greatest importance is that it should be clearly established that a significant injustice has probably occurred and that there is no alternative effective remedy. … Where the alternative remedy would be an appeal to the House of Lords this court will only give permission to reopen an appeal which it has already determined if it is satisfied that an appeal from this court is one for which the House of Lords would not give leave.

Taylor v Lawrence in a case of fraud

15. So far so good. But the present case is not one in which the court is said to have misbehaved, by bias; but one in which the court is said to have been misled by misbehaviour, in the form of perjury, by one of the parties. That issue was also addressed in Taylor v Lawrence, and was seen by this court as raising distinctly different issues from those that arise in a case of bias. Before the creation of an appellate court with general statutory authority by the Judicature Act 1873 decisions of lower courts said to have been obtained by fraud could only be dislodged by collateral means, by the bringing of a fresh bill in chancery. In Flower v Lloyd (1877) 6 ChD 297 this court had to decide whether the new Court of Appeal could, in effect, take over that jurisdiction, by permitting the reopening of determined appeals that were alleged to have involved a fraud on the court. This court held that it had no such jurisdiction. The complaint as understood by the Court of Appeal, 6 Ch D at p299, was that there had been "subsequent discovery of facts which show or tend to show that the order of the Court of Appeal was obtained by a fraud practised on the court below". That, it will be appreciated, is exactly the same complaint as is made in our case. The Court held that such a complaint must continue to be made, as before the Judicature Act, by collateral action. As Jessel MR put it, at pp 300-301, if this court has once determined an appeal, it has no further jurisdiction.

16. In §25 of his judgment in Taylor v Lawrence Lord Woolf pointed to dicta in other cases shortly after Flower v Lloyd in which a different view had been suggested. He also mentioned two cases of more recent vintage, including Wood v Gahlings The Times, 29 November 1996, a decision of Lord Woolf MR, that assumed, without deciding, that the court had jurisdiction to reopen an appeal in a case of fraud as well as of bias. But Lord Woolf did not suggest, nor with deference could he have done, that Flower v Lloyd had ceased to be good law: indeed, at §25 of Taylor v Lawrence Lord Woolf accepted that there is no subsequent case which is in direct conflict with Flower v Lloyd.

17. That this court in Taylor v Lawrence accepted the continuing authority of Flower v Lloyd in cases of fraud is further illustrated by its appreciation that in order to reach the conclusion that it had jurisdiction in a bias case the court had to distinguish Flower v Lloyd. That it did by saying, at its §49, that

It is the reality of the situation which means that we cannot, as Sir George Jessel MR did in Flower v Lloyd, take refuge in the fact that there is an alternative remedy. If there is no effective right of appeal to the House of Lords and this court is the only court which can provide a remedy then in our judgment there can arise the “exceptional circumstances” to which Russell LJ referred in In re Barrell Enterprises [1973] 1 WLR 19.

And it was in those circumstances that there arose the power of the court, like any other court, in the words of Lord Diplock in Bremer Vulcan at p 977, already quoted, to do what was necessary to maintain its character as a court of justice. That, as we have seen, was what permitted the Court of Appeal to deal with cases of bias on the basis adopted by the House of Lords in Pinochet (No 2).

18. That distinction turns on the perceived absence of an alternative remedy in a bias case, at least where the court is satisfied that there is indeed no alternative because the case is one in which the House of Lords will not give leave. But in a fraud case there is always an alternative remedy, as Jessel MR and Lord Woolf said, in the shape of a fresh action to set aside the original decision.

19. In terms of precedent, therefore, Taylor v Lawrence itself is only authority for the proposition that the Court of Appeal has jurisdiction to reopen an appeal in which judicial bias is alleged, and then only if satisfied that the House of Lords would not give leave to appeal the original decision of the Court of Appeal. That the court has no such jurisdiction in a case of fraud is decided by the authority of the Court of Appeal in Flower v Lloyd. But that proposition is supported not just by authority in this court but also by a binding decision of the House of Lords.

20. [1930] AC 298 is important not only because the House approved Flower v Lloyd, but also because it held that complaints of fraud practised on a lower court must be made by way of collateral action and not by way of appeal: that is, not even by an initial appeal, let alone by seeking to reopen an appeal that was already concluded. In that case the Court of Appeal had set aside a judgment that had been obtained by fraud, and ordered a retrial. The House of Lords reversed that decision. Lord Buckmaster, giving the only substantive speech, said at pp 300-301:

It has long been the settled practice of the Court that the proper method of impeaching a completed judgment on the ground of fraud is by action which, as in any other action based on fraud, the particulars of the fraud must be exactly given and the allegation established by the strict proof such a charge requires. In Flower v Lloyd (1877) 6 Ch D 297, 302 the Court of Appeal, consisting of Jessel MR, James and Baggallay LJ, held that there was not jurisdiction in the Court of Appeal to entertain a similar application with regard to one of their own judgments. James LJ states that "you cannot go to your adversary and say ‘You obtained the judgment by fraud and I will have a rehearing of the whole case’ until that fraud is established." Flower v Lloyd, Cole v Langford [1898] 2 QB 36 and Baker v Wadsworth (1898) 67 LJ (QB) 301 show that the right procedure for that purpose is by action. That, however, there is jurisdiction in special cases to set aside a judgment for fraud on a motion for a new trial may be accepted. Hip Foong Hong v Neotia [1918] AC 888 is such a case; but it should be remembered that this case had come up to the Privy Council on this procedure and the Board would naturally be unwilling to defeat a case at its last stage on such a ground.

21. It should be noted that Hip Foong Hong was not concerned with the reopening of an appeal, the subject-matter of the rule in Flower v Lloyd and of the application before us, but with a motion to the original court for a retrial. However, even in that context, and to touch on a matter that we will have to consider in more detail below, the Board at p 894 made some general observations about procedure:

… where a new trial is sought upon the ground of fraud, procedure by motion and affidavit is not the most satisfactory and convenient method of determining the dispute. The fraud must be both alleged and proved; and the better course in such a case is to take independent proceedings to set aside the judgment upon the ground of fraud, when the whole issue can be properly defined, fought out, and determined, though a motion for a new trial is also an available weapon and in some cases may be more convenient

Taylor v Lawrence as seen in subsequent cases

22. We are accordingly of the view that Flower v Lloyd and Jonesco v Beard preclude the reopening of an appeal on grounds of fraud perpetrated on the lower court, the proper recourse in such a case being an action to set aside the original judgment. That, however, is not how the issue has been approached in decisions subsequent to Taylor v Lawrence. We were shown four cases in this court subsequent to Taylor v Lawrence in which jurisdiction to reopen appeals on grounds of fraud has been assumed. Those cases are Couwenbergh v Valkova [2004] EWCA Civ 676; Re Uddin [2005] 1 WLR 2398; Pell v Express Newspapers [2005] EWCA Civ 46; and First Discount v Guinness [2007] EWCA Civ 378. The issue, based on Flower v Lloyd and the treatment of that case in Taylor v Lawrence, of whether the Court of Appeal has any jurisdiction to reopen an appeal in a case of fraud, was not addressed in any of these cases: though it right to say that in Uddin, at §20, this court viewed the Taylor v Lawrence jurisdiction as appealing to the correction of injustice, stated in very broad terms, and not limited to particular cases such as bias or fraud. How that view sat with Flower v Lloyd and Ionesco [sic] v Beard was not explored: Flower v Lloyd was not put to the court in these terms, and the court was not even shown Ionesco v Beard.

23. We therefore remain of the view that the Flower v Lloyd issue remains open, and in another case it will be possible to explore whether this court indeed has any jurisdiction to reopen an appeal on grounds of fraud perpetrated on the court. We accept, with some hesitation, that that will indeed have to be in another case, despite the issue being one of jurisdiction, because, no doubt because other divisions of this court have accepted jurisdiction in fraud cases, this preliminary point was not relied on by Lloyds in this application. We will therefore proceed on the assumption, but by no means deciding, that this court does have jurisdiction to reopen a determined appeal on grounds of fraud.

24. That, however, leads us to the question of when, if at all, that jurisdiction should be exercised. In only one of the four cases mentioned above, Couwenbergh, was relief actually given, and then in very singular circumstances as we will later demonstrate. In the other four cases, careful warnings were given about the strict limits that have to be placed on the exercise of the jurisdiction in a fraud case.

25. It will be convenient to start an analysis of the proper exercise of the jurisdiction in a fraud case with the observations of Butler-Sloss P in Uddin: and that not least because in First Discount v Guinness this court held, at its §10, that to the extent that Uddin differs from Couwenbergh the former must prevail. The guidance given in Uddin has, we emphasise, to be read bearing in mind the doubts mentioned in §22 above as to whether the court should have been addressing the issue at all. That guidance can be summarised as follows:

a) Fresh evidence admitted under the rule in Ladd v Marshall [1954] 1 WLR 1489, sufficient to be taken into account in a first appeal, will not in itself justify a second appeal under Taylor v Lawrence

b) For the Taylor v Lawrence test to be passed, there has usually, but not necessarily always [§20], to be shown corruption of the judicial process [§18]

c) Absent "corruption", merely to demonstrate that the wrong result was reached below may justify reopening the earlier appeal; the earlier appeal will not be reopened simply on claims that the first decision may have been wrong [§22].

26. That analysis was developed by this court in First Discount v Guinness by saying that it may be enough to invoke Taylor v Lawrence to demonstrate that corruption of the earlier process, eg by the giving of perjured evidence, has probably occurred; but that the court must then be satisfied that there is a strong probability that the corruption affected the result of the first case. How is it to be decided that the fresh evidence sufficiently demonstrates corruption of the earlier process? That question was addressed by Rix LJ in Pell. He referred to what had been said in Jonesco v Beard about the proper method of impeaching a judgment being by way of action, and then said at his §48:

Even if that practice were to be regarded as less settled, so that it is a matter of discretion on the particular facts of each case whether such an allegation of fraud should proceed by way of appeal or by way of fresh action, nevertheless it is in my judgment clear that in this case Mr Pell should be required to proceed by way of fresh action and not by way of appeal. First, the alleged fraud is hotly disputed. Second, in a complex case the allegation has yet to be clearly and properly set out in the way in which only pleadings can clarify. Thirdly, the appeal process is an unsatisfactory basis for the analysis of an untried allegation of fraud.

27. Those observations importantly remind us, amongst other things, of the limits of this court’s procedures and powers. Take the present case. The central allegation is that evidence has emerged since the first appeal that arguably demonstrates that evidence given at the trial was perjured. If the appeal is to be reopened on that basis the following steps have to be gone through. First, the court must be satisfied that the new evidence is admissible under Ladd v Marshall principles. That was the basis of the court’s approach in Uddin. Second, it must be shown on the basis of that evidence to be probable that corruption of the earlier process in fact occurred (First Discount). Third, there must be a strong probability that that corruption affected the result of the first trial. The nature of the second and third enquiries means, as Rix LJ pointed out, that unless the case is so clear as to be in effect incontestable this court is neither equipped to undertake the enquiry nor able to give any immediate remedy if it were to find that the allegations are established. That is why Sir Martin Nourse said, summarising the jurisprudence in §25 of the judgment of this court in Sohal v Sohal [2002] EWCA Civ 1297, and echoing Jonesco v Beard, that even a first appeal should be pursued on grounds of perjury only at the trial, as opposed to leaving the matter for a fresh action, if the allegation of fraud "could be clearly established" (per Lord Woolf in Wood v Gahlings) or (which was thought to come to the same thing) the fresh evidence or its effect is not "hotly contested" (per Lord Phillips in Hamilton v Al Fayed, 21 December 2000, unreported).

28. These concerns also engage the remedies available in this court, as well as its powers of investigation. We again refer to the present case. Even if we were to be persuaded that there was a strong possibility that on the reopening of this appeal the court hearing that renewed appeal would conclude that it was probable that corruption of the earlier process had occurred, it is difficult or impossible to see how that court could grant any relief other than a retrial: because it would be beyond the power, or good sense, of the court to conclude that even if perjury had probably been committed in the respects on which the applicants rely it was so plain that that had affected the outcome of the case that the court should there and then substitute a different decision. Indeed, when Mr Jenkins was asked what ultimate outcome he envisaged as following from this application he agreed, after some discussion with the court, that at the reopened appeal his evidence would have to be tested, probably by extensive cross-examination, together with evidence on the other side, not least from those now said to have committed perjury; and still that the only feasible outcome of all of that process would be an order for retrial. The suggestion that that retrial could be limited to the matters recently raised was quite unreal. Those matters cannot be properly determined except in the context of the whole complaint.

29. To proceed by way of reopening the appeal would, therefore, simply add another step on the road to the sort of enquiry into the substance of the perjury and its effect on the original proceedings that would be provided by a collateral action. It is therefore not merely a matter of jurisdiction but also a matter of practicality that dictates that in all but the most exceptional case complaints such as the present should be pursued by way of a fresh action to set aside the original decision, rather than by reopening the appeal.

30. Such an exceptional case was identified by this court in Couwenbergh. The facts were in quite different compass from our case. The dispute concerned the validity of a will. At first instance, and in the original appeal, it was accepted that the will had been properly attested, the issue turning on the capacity of the testator. The party challenging the will lost at both levels, and incurred a very substantial liability for costs. Evidence then emerged, held by this court to be admissible under the Ladd v Marshall test, that suggested that the alleged attestors had not in fact signed the will, and that the claim to the contrary effect in the original trial had been fraudulent. The issue, although no doubt controversial, was thus limited to a short and discrete question, of whether or not the witnesses had attested the will: it being undisputable that, if they had not done so, the will was invalid.

31. This court, whilst setting out the limits of the Taylor v Lawrence jurisprudence, identified two particular features of Couwenbergh that enabled it to give permission for the reopening of the first appeal, as opposed to leaving the complainant to bring a fresh action. First, a retrial (which the court recognised would be the only relief that it could order) would be a short and simple matter, with no fresh pleadings because the issue was already clearly joined, and only a small part of the case needed to be retried. Second, the Court of Appeal in §28 of its judgment identified as the potential substantial injustice of the first trial that the order for costs there made might turn out to be unfair; and in §39 said that the most important consideration pointing towards reopening the appeal was that at a retrial the judge could make a flexible and comprehensive order as to the costs of the whole proceedings, in a way that, it was thought, would not be open to the judge hearing a fresh action. It is not by any means clear that the latter assumption was correct, since there is no obvious reason why the relief claimed in a fresh action should not include the discharge of the costs order made on an erroneous basis in the first action. But, however that might be, the perceived centrality of the costs issue in Couwenberg was a very powerful consideration driving the court towards ordering the reopening of the original appeal.

32. The present applicants originally sought to argue that this case is on the same level as Couwenberg, but that contention was impossible, and Mr Jenkins properly did not seek to press it in oral argument. A new trial in our case could not simply start from the pleadings and assumptions of the old trial. That is because the new evidence does not, as in Couwenberg, contest a simple issue of physical fact, whether the will was attested or not; but is said to cast light on the state of mind, knowledge and intentions not just of the men who are said to have committed perjury but also of the body of Lloyds members of which they were part. That will require sophisticated comparison of what is now said to be proved against the perjurers with what was demonstrated at the trial. We do not go so far as to say of Couwenberg what Rix LJ said in §49 of his judgment in Pell, that the idiosyncrasies of that case in no way guide the solution of this application, but it is quite clear that Couwenberg cannot be relied on for any solution outside the ambit of its own particular facts.

The allegations in this case: preliminary

33. The applicants are faced with a significant preliminary difficulty. There are serious issues as to whether the evidence on which they seek to rely is admissible at all at appellate level, bearing in mind the preconditions for admission of evidence laid down in Ladd v Marshall. Lloyds raise two objections. First, it is strongly argued by Lloyds that most or all of the witnesses were available to the applicants at the time of the trial before Cresswell J, or could have been obtained by the exercise of reasonable diligence. Mr Merrett falls into a somewhat different category, but his unavailability was largely the result of the fact that at the time of the original trial he was one of the persons accused by the names of fraud in Lloyds v Jaffray; a claim that as we understand has still not been formally withdrawn, though no doubt no longer pursued. Second, much of the new evidence, and in particular that of Mr Merrett, consists of commentary on evidence given at the trial, rather than first-hand reporting of new facts.

34. We intend to deal with these complaints as follows. The Ladd v Marshall objection is one of substance, which strictly speaking should be determined before we embark on the specific Taylor v Lawrence enquiry: see the approach of Butler-Sloss P in Uddin. But that objection, or an objection as close to it as to be indistinguishable in practical terms, forms an important part of the forthcoming appeal in Lloyds v Henderson. It would not be right to pre-empt that decision by what in the present application is only a threshold decision. We will therefore assume in the present application that the Ladd v Marshall test is satisfied. That is an assumption distinctly favourable to the applicants, which would not have been made, or at least would not have been made without substantial argument, had it not been for the order in which the various forms taken by the dispute between the names and Lloyds have come before the court.

35. Second, however, that is no excuse for giving weight to those parts of the evidence that are not evidence at all. Indeed, so far as Mr Merrett’s comments are concerned, counsel was reduced to saying that they should be regarded as, or like, the evidence of an expert: see for instance the names’ statement, § 10(iii). But there are plenty of experts in the world, not least experts on the insurance market, who if it had been thought of at the time could have assisted at the trial. However, to exclude Mr Merrett’s commentary does not in the event significantly reduce his value to the applicants. The substantial part of their case in the application, as opposed to some of the forensic matter that accompanied it, is largely built not upon what Mr Merrett says, but upon certain documents that he produces. Those documents can be looked at for what they do or do not say, irrespective of their provenance.

36. The names’ allegations being of perjury, like any other allegation of fraud they have to be distinctly made and distinctly proved. Authority for that proposition scarcely needs to be cited, but it will have been noted from §20 above that the point was made specifically in the context of the present jurisdiction by Lord Buckmaster in Jonesco v Beard. In an attempt to meet the first limb of that requirement the names served with their application a document "Extended Grounds for Application", which they expanded on in the further statement of their case that was required by the court. It has only been possible to extract from those documents the three allegations that are addressed below, which Mr Jenkins accepted were the basis of his case. We deal in turn with those allegations, separately identified by letter. We then deal with two further claims that are not easy to extract from the written submissions, but which were pursued in oral argument.

Allegation A: Sir David Rowland committed perjury before Cresswell J

37. Sir David Rowland was a prominent member of Lloyds, and at one time its Chairman. The allegation is that he committed perjury when he told Cresswell J that with the benefit of hindsight, but only with the benefit of hindsight, it can be seen that in 1982 reserves against asbestos liability were inadequate. The new evidence is said to show that Sir David well knew in 1982 that reserving against asbestos liability was impossible, and he positively concealed that knowledge in the answer that he gave to Cresswell J.

38. This allegation was based on comments by Mr Merrett, but more substantially on documents that he produced in relation to the affairs in the crucial period of 1981-1982 of Pulbrook Underwriting Management Ltd. Sir David was in 1982 the Chief Executive of Stewart Wrightson Holdings, the ultimate parent company of Pulbrook. The evidence now said to inculpate him in perjury at the trial consisted of a series of Pulbrook minutes and papers alleged to show very serious exposure to asbestos liability, which it was to be assumed Sir David would have been shown. That was the smoking gun that showed him to have known and realised much more than he had admitted to the judge.

39. The fatal flaw in this claim was that Lloyds were able to demonstrate that almost all of this material had in fact been before the court in the original trial, and Sir David had been cross-examined on a substantial part of it. Faced with that demonstration, Mr Jenkins made no attempt to support the original case. Rather, he concentrated on the one document produced by Mr Merrett in this connexion that had not been before the court. This was a letter of 18 March 1982 from American attorneys, Mendes & Mount, reporting to Wrightsons about a meeting with London underwriters to discuss the state of litigation in the USA with a particular insured. The claim is that the letter showed an extremely alarming prospect for the underwriters; and Sir David Rowland must have known about it.

40. To deal first with the claim that the letter had been concealed from the court in the trial in this case, the letter, being held by individual underwriters, was not in the power, possession or control of Lloyds; and indeed we were shown a statement by leading counsel then representing the names before Cresswell J that he did not seek to pursue attorneys’ reports relating to particular insureds. The letter may or may not have been in the power possession or control of Sir David Rowland. We would in fact think it very unlikely indeed that that was so in respect of Sir David personally, as opposed to his company; but that is irrelevant to the present issue, because Sir David was not a party to the action.

41. Accordingly, to the extent that it was now alleged that there had been fraud or concealment in relation to the letter itself that charge was not made out. The point however remains of the effect of the facts set out in the letter. We were told from the bar that Mr Merrett’s opinion was that the position set out in the letter was so serious that, if he had known of the letter at the time he “would have been unable to plead innocence in Jaffray”. Mr Merrett had not in fact said that in any of the 299 pages of evidence that he had filed in this application, but we were told that that was his view of a letter now seen as crucial to the names’ case. Having read the letter, and having seen the comparatively modest role assigned to the letter amongst all the material originally relied on, we will be permitted to express some scepticism. But however that may be, Mr Foxton was able to show us material that had been before Cresswell J in which the problem reported on in the 18 March 1982 letter had been fully exposed and discussed.

42. We are therefore unpersuaded that the 18 March 1982 letter significantly altered the landscape as it was understood to be at the trial. But even if we took a different view, this would quintessentially be material that drives towards a fresh action rather than to the reopening of the appeal. First, it would have to be established whether Sir David Rowland ever saw the letter. Next, the court would need to know what the market’s perception was of the matters set out in the letter, as well as similar matters affecting other insureds, an enquiry that would demand the calling of a great deal of evidence; and then whether that perception was shared by Sir David; and then, but only then, whether what he had said at the trial had been untrue. These are, as Mr Jenkins indeed said, sophisticated and detailed enquiries, wholly unsuitable for pursuit in the course of even a first, let alone a second, appeal. We remind ourselves of what was said by Sir Martin Nourse in Sohal v Sohal, by Lord Woolf in Wood v Gahlings, and by Lord Phillips in Hamilton v Al Fayed.

43. One other thing should be said about this part of the case. Mr Merrett’s evidence contained a good deal of comment about the implications that could be drawn from the evidence at the trial, the understanding of the market that the court should have obtained, and the light thrown on the process generally by the documents. We did not permit any of this to be developed, because on no view is any of it the stuff of Taylor v Lawrence. To take up the analysis in this court in Uddin, none of this suggested corruption of the earlier trial, nor was it any more than a claim that the decision at the trial might have been wrong, rather than that the wrong result was clearly reached.

Allegation B: Mr Murray Lawrence committed perjury before Cresswell J

44. Mr Murray Lawrence, who in 1982 was Deputy Chairman of Lloyds, was like Sir David Rowland the subject of the allegation in §12 of the UNO applicants’ "Extended Grounds for Applications" that

the evidence given to the court in Jaffray was misleading as to the knowledge about asbestos liabilities faced by the market on the part of at least two key witnesses (Murray Lawrence and Sir David Rowland) and that the Council/Committee of Lloyds knew in the 1980’s that the accounting systems in place at Lloyds were unable to produce reasonable estimates of the outstanding liabilities for a number of reasons, some continuing throughout the period and some of particular significance in relation to particular years

45. The basis of this bare allegation is set out in §11 of the applicants’ statement. The treatment is much more diffuse than that in the case of Sir David Rowland, and mentions a range of matters without specifying with any clarity how they demonstrate Mr Lawrence to have perjured himself. We very much doubt whether this presentation fulfils the initial requirement of a fraud case of the allegation being clearly stated and clearly proved. The most that could be said was that there were some matters that could be relied on to suggest that Mr Lawrence was less sanguine about the position in 1982 than he had suggested to the judge.

46. Lloyds went through this material in great detail in pp 28-38 of its skeleton, and was able to show that much or all of it had been in issue at the original trial: as had already been demonstrated in respect of the material relating to a dispute between Mr Lawrence’s syndicate and The Travelers Insurance Co in §67 of Lloyds’ skeleton in the appeal in Lloyds v Henderson. Mr Jenkins said that he nonetheless stood by two of the criticisms of Mr Lawrence. We do not pursue those further: first, because we cannot see that they materially alter the position as understood at the trial; and second because even if we took a different view the probity of Mr Lawrence would again plainly not be a matter suitable for pursuit by way of appeal.

47. Even more clearly than in the case of the allegations made against Sir David Rowland, the complaints against Mr Lawrence come nowhere near to establishing even an arguable case for further investigation of possible perjury on his part.

Allegation C: various people committed perjury in relation to the Murray Lawrence letter

48. This allegation is not made in the Extended Grounds, but it can be spelled out of §14 of the applicants’ skeleton. It is very doubtful whether the claim adds anything substantial to the allegations of perjury against Mr Lawrence himself, and indeed Mr Jenkins went no further than to say that there were features of the letter that added to that picture.

49. The Murray Lawrence letter [MLL] was a letter written by Mr Murray Lawrence on 18 March 1982 on behalf of the committee of Lloyds, in the context of concerns raised by some auditors (the "Neville Russell" letter) about reserves for asbestos liability. The claim made at trial, in very considerable detail, was (to put it extremely briefly) that the committee, or at least its most influential members, knew in 1982 that proper reserving for asbestos claims was impossible. In the MLL they dishonestly concealed that fact, by purporting to give advice and guidance as to how asbestos liabilities should continue to be reserved for. As an adjunct of that claim it was suggested that the MLL had been deliberately given a limited circulation, its main purpose being to serve to protect the backs of its authors should they thereafter be criticised. Cresswell J heard substantial evidence about the latter allegation, and rejected it.

50. Cresswell J also rejected the main part of the case, and held that the MLL had been an honest document. The Court of Appeal agreed with that conclusion, though expressing more doubt than had Cresswell J. But an important part of its reasoning was that the Court found to have been very powerful submissions by Lloyds that pointed to the subsequent behaviour and commitment to the market of those who originated the MLL: see §414 of the Court’s judgment. The conclusions of the Court of Appeal, including its explanations of its concerns about this issue, are extensively set out in §§ 428-432 of its judgment. We do not extend this judgment by setting them out again. The matter put before us in this application went nowhere near to dislodging that central part of this court’s reasoning.

51. The further evidence now relied on by the applicants substantially takes the case no further, and certainly does not demonstrate that Mr Lawrence, or anyone else, perjured themselves at the trial in relation to this issue. Mr Jenkins really was not able to establish this as a separate part of the case in Taylor v Lawrence terms.

52. We would also, in view of the weight placed by the applicants upon the evidence of Mr Merrett, mention another, albeit to some extent forensic, point on which Lloyds relied. In his evidence in Henderson v Merrett [1997] LRLR 265, when Mr Merrett as underwriter was sued by various members of his syndicate, he said amongst other things that "throughout the period from 1980 to 1988, while asbestos claims were naturally perceived at each year end to be a serious problem, at no time did we believe that it was not possible to make a reasonable estimate of the reserves needed to meet these claims". In his evidence in this application, §3.22 of the statement of 27 April 2007, Mr Merrett confirmed that that correctly represented his view at the time of the Henderson v Merrett trial. That view, he now says, was wrong, as is admitted to have been the same view held, amongst others, by Sir David Rowland. But the issue is not whether the view was wrong, but whether it was honestly held at the relevant time. The evidence of Mr Merrett, who was both a prominent underwriter and in 1982 a member of the Lloyds Committee and the Audit Committee to the effect that in 1982, and subsequently, he did not believe that it was not possible to make a reasonable estimate of reserves needed to meet asbestos claims puts obvious difficulties in the way of establishing a case of fraud and perjury against other members of those committees. And the need to offset or explain such evidence further demonstrates the extent and complexity of the enquiry sought by the names, and the impossibility of conducting that enquiry through the medium of an appeal.

53. That is the limit of the properly pleaded issues that even potentially raise questions under Taylor v Lawrence. Two other claims were however made before us, which we will briefly consider. It has to be said about both of them that they could not even arguably get off the ground unless Mr Jenkins were correct in the extensive view of this jurisdiction discussed in §§ 7-8 above. The first of them, even on its face, has nothing to do with fraud or corruption of the court process, and the second of them on proper analysis is in the same position.

The terms of the misrepresentation found against Lloyds

54. The allegation pursued before Cresswell J was that Lloyds had made misrepresentations about the adequacy of reserving for asbestos liability. However, this court, apparently of its own motion, stated the representation, that it then found to have been untrue, in terms that there was in place a rigorous system of auditing which involved the making of a reasonable estimate of outstanding liabilities including unknown and unnoted losses. The applicants placed great weight, for the source of this finding, on some observations that fell from the court during argument on Day 9 of the appeal; the formulation just set out is however taken from the court’s considered statement in § 587(i) of its judgment. There is a dispute between the parties as to whether this was a widening of the claim as pleaded by Lloyds. For present purposes we will assume that the representation found was different from that pleaded.

55. But in any event it is far from immediately apparent how this significantly altered the nature of the case, because a representation that it was possible to reserve for asbestos liability would seem to have pregnant within it a representation that the accounting systems necessary for proper reservation were in place. The applicants however argue that, having so defined the representation, it was not open to the court then to go on and find that Lloyds did not know that the representation was untrue; because the court had had no evidence as to Lloyds’ state of mind with regard to that representation. Justice, very broadly stated, required the reopening of the appeal, so that there could be heard evidence on that subject that the names now had available.

56. We do not go into the merits of all this, because the argument is misconceived. If it was the case that the court reached conclusions on an issue on which it did not have proper evidence, then that was a matter for original appeal: which the names, armed with skilled advice, including that of leading counsel, did not pursue. It is not a complaint that can be made many years later under the guise of Taylor v Lawrence.

Reinsurance to close [RITC]

57. The second ancillary ground on which the application is made is that the court was misled about the nature of RITC, and that as a result its conclusion that in 1982 Lloyd’s did not know that its accounting systems were incapable of producing a reliable estimate of outstanding liabilities is fatally flawed.

58. In order to close the accounts of a Lloyds syndicate it is necessary to make adequate provision for any outstanding liabilities, which must cover claims already made but not fully settled and liabilities incurred but not reported (“IBNRs”). The traditional way of making provision for outstanding liabilities is by RITC, which on the face of it involves an agreement by the names on the syndicate for the following year to reinsure the names on the closing year against all their outstanding liabilities in consideration of the transfer to them of the syndicate’s existing reserves and the payment by way of premium of a properly calculated reserve in respect of the outstanding liabilities of the closing year. It follows that in order to calculate the closing premium the syndicate managers must be able to make a fair assessment of outstanding claims, including IBNRs. If they cannot do so, the accounts must remain open.

59. It appears to have been generally assumed by most, if not all, of the claimants in this case that the effect of a completed RITC was to extinguish any further liability on the part of the members of the closing syndicate; in other words, they assumed that their outstanding liabilities were transferred, not merely in a business sense but in a legal sense, to the members of the syndicate for the following year. The legal effect of RITC was not in issue in the trial. The issue has emerged on this application only because Mr. Merrett says that Lloyd’s have recently asserted for the first time that RITC involves reinsurance, and not a complete discharge in the sense of a novation: with the consequence that names who underwrote policies in the 1950s (and subsequently) remain liable on them to this day. He says that is not the correct view of the matter, but that if it is, it shows only too clearly that, contrary to this court’s conclusion, Lloyd’s must have been aware that its accounting system was incapable of producing a reliable assessment of any syndicate’s outstanding liabilities, its solvency or Lloyd’s solvency as a whole, since no attempt was ever made to ascertain the solvency of names on closed years.

60. This part of the names’ case depends on the following propositions: (a) that when hearing the appeal the court understood that the effect of RITC was to bring about a legal transfer of outstanding liabilities from the Names on the closing year to the Names on the next open year; (b) that that was an important factor in its finding that Lloyd’s thought that the audit system was capable of producing reliable information about outstanding liabilities and solvency; (c) that if Lloyd’s is right in saying that RITC operates merely as a reinsurance, the court was misled; (d) if the court had realised that RITC operates merely as reinsurance, its finding about Lloyd’s perception of its audit system would have been different and fraud would have been established.

61. This argument fails at the first stage. In § 373 of its judgment the court adopted the description given by Mr. Outhwaite in his statement in Stockwell v RHM Outhwaite (Underwriting Agencies) Ltd of the way in which reserves were set. In the light of that evidence there is no basis for saying that the court was under the impression that RITC operated as a novation of names’ liabilities to the original policyholders. If that is right, it follows that when reaching the conclusion that Lloyd’s thought that its audit procedures were capable of producing reliable information for the purposes of reserving and solvency the court was not influenced by an understanding that RITC involved a transfer of legal liabilities, whether that be the correct view or not

62. The names are simply seeking to re-open the appeal in order to put forward an argument that was always available to them but which no one thought, or wished, to pursue. It does not depend on new evidence. At best all that has happened is that Mr. Merrett, and through him the names, has become alive to a view of RITC which had not previously occurred to him. And it is very doubtful whether the view of RITC which Mr. Merrett says is now being put forward for the first time is of any significance in relation to Lloyds’ understanding in the early 1980s of the effectiveness of the audit procedures. If it were, we are confident that the argument would have been pursued vigorously. These disputes and the way in which they arise are miles away from the proper ambit of Taylor v Lawrence.

Mrs Adams and Mr Butler

63. We have recorded that we were addressed also by Mr Adams and Mr Butler. We are grateful to them for the courtesy and restraint that they displayed, in a matter on which they feel extremely strongly, and which has burdened them for very many years. It is no criticism of them to say that they were understandably not alive to the constraints affecting this jurisdiction. Their submissions did not relevantly add to those that we received from the UNO names.


64. We have gone into these applications in some detail, both because of the seriousness of the allegations made in them and because of the strong feelings entertained by the applicants. In truth, however, the applications were completely hopeless. None of the points raised even start to make out an arguable case under Taylor v Lawrence. That is their condition in any event; but even if the case were a great deal more arguable than it is, the only possible form of procedure to accommodate it would be by a fresh action to set aside the judgment. We dismiss the applications.

65. Since this judgment deals with issues of general importance in relation to Taylor v Lawrence, it is released from the restriction on citation contained in Practice Direction (Citation of Authorities) [2001] 1 WLR 1001.