[Home] [Databases] [World Law] [Multidatabase Search] [Help] [Feedback]
	England and Wales Court of Appeal (Civil Division) Decisions
You are here: BAILII >> Databases >> England and Wales Court of Appeal (Civil Division) Decisions >> Sun Life Assurance Company of Canada & Ors v The Lincoln National Life Insurance Co [2004] EWCA Civ 1660 (10 December 2004) URL: http://www.bailii.org/ew/cases/EWCA/Civ/2004/1660.html Cite as: [2004] EWCA Civ 1660, [2005] 1 Lloyd's Rep 606, [2006] 1 All ER (Comm) 675, [2005] 2 CLC 664

[New search] [Printable RTF version] [Help]

IN THE SUPREME COURT OF JUDICATURE
COURT OF APPEAL (CIVIL DIVISION)
ON APPEAL FROM
Hon Mr Justice Toulson
Lower Court Reference 2003 Folio 575.

		Royal Courts of Justice Strand, London, WC2A 2LL
		10 December 2004

Lord Justice Mance:

INTRODUCTION

1. This appeal concerns the correct analysis of an arbitration award relating to a dispute between X and Y, and its relevance in a subsequent, separate arbitration between X and Z in which the legal position between X and Y is an issue. X represents the three appellants to this court, Sun Life Assurance Company of Canada, American Phoenix Life and Reassurance Company and Phoenix Home Life Mutual Insurance Company ("Sun/Phoenix"), Y represents Cigna Reinsurance Company of Europe SANV ("Cigna"), and Z represents the respondents in this court, The Lincoln National Life Insurance Company ("Lincoln"). Because arbitration awards are generally confidential, it is worth mentioning that the award between Sun/Phoenix and Cigna was made available to Lincoln with, it seems, Cigna's consent (and certainly to Cigna's knowledge and without Cigna applying to prevent it occurring) following an order made by the arbitrators in the arbitration between Sun/Phoenix and Lincoln.
2. Sun/Phoenix commenced an arbitration against Cigna, claiming that Cigna had reinsured them under Occupational Accident ("Occ/Acc") reinsurances (covering 50% of losses of (a) $32,500 excess of $17,500 and (b) $200,000 excess of $50,000). These reinsurances were allegedly concluded in August/September 1998 in respect of incoming business underwritten on Sun/Phoenix's behalf by Centaur Underwriting Management Limited of Bermuda ("Centaur"). Sun/Phoenix alleged that the Occ/Acc business so reinsured included seven whole account reinsurances accepted by Centaur to run for three years from 1^st December 1997 to protect a personal accident pool underwritten and managed by Unicover Managers Inc. of New Jersey ("Unicover"). The arbitrators were Mr Nicholas Legh-Jones QC, Mr Gordon Cornish and Mr Brian Kellett. Sun/Phoenix lost this arbitration, since the tribunal by its award dated 14^th March 2002 declared that the whole Cigna reinsurance had been validly avoided for misrepresentation and non-disclosure by the brokers, Stirling Cooke Brown Reinsurance Brokers Limited ("SCB"), by whom it had been placed.
3. A second arbitration was commenced by Lincoln against Sun/Phoenix. In it, after losing the Cigna arbitration, Sun/Phoenix contended by amendment to their counterclaim that the Occ/Acc reinsurances had never covered their exposure on the Unicover book of insurances, and that such exposure was recoverable accordingly from Lincoln under the terms of a reinsurance covering loss of US$90,000 excess of $10,000 any one loss in respect of risks attaching during the period 1^st October 1996 to 31^st December 1997. It is common ground that whether such exposure was recoverable from Lincoln depended upon whether the Occ/Acc reinsurances with Cigna would, apart from their avoidance, have covered it. This is because the Lincoln reinsurance incorporated a Net Retained Lines ("NRL") clause, signifying the following terms:

"This Agreement applies only to that part of the Original Policies which the Reinsured retain net for their own account and in computing the Ultimate Net Loss, only loss or losses in respect of such Net Retained part of the Original Policies shall be included.

The amount of Reinsurers' liability in respect of loss or losses shall not be increased by reason of the inability of the Reinsured to collect from any other reinsurers whether specific or general, any amounts which may have become due from them whether such inability arises from the insolvency of such other reinsurers or due to the cancellation of such other reinsurance by the Reinsured or by such other reinsurers for whatever reason or for any reason whatsoever."

It is not, and was not below, in issue that this clause applies to risks covered under a reinsurance which is later successfully avoided ab initio – even a reinsurance (like the Cigna reinsurance in fact) effected after the Lincoln reinsurance. Nothing in this judgment should therefore be taken as commenting on any argument that might have been raised to the contrary in a case like the present where the Occ/Acc reinsurances were not placed until after the Lincoln reinsurance.

4. The issue before us arises because Lincoln contend that, in addition to or in the course of holding that the Cigna reinsurance had been validly avoided, the Cigna arbitrators also determined that, apart from such avoidance, the Unicover book would have been covered by the Cigna reinsurances. That, Lincoln submits, was and is a binding determination as to the position as between Sun/Phoenix and Cigna, on which Sun/Phoenix cannot go back as against Lincoln in the present arbitration.
5. The Lincoln tribunal (composed of Mr John Rowland QC, Mr Brian Wood and Mr Anthony Robertson) considered by its award dated 27^th May 2003 that it was "bound by the decision of the Cigna tribunal", but it went on to conclude that the Cigna tribunal were "not focusing on the issues that [had] arisen in [the Lincoln] arbitration" and "were not analysing the evidence before them with a view to providing any insight into the issues that [had] to be decided" in the Lincoln arbitration. Accordingly the Lincoln tribunal concluded that it was necessary for it to "reach its own view as to whether or not Cigna could ever have been bound to indemnify [Sun/Phoenix] in respect of losses arising from the Unicover whole account reinsurances". Its own factual conclusion - based on the witness statements, transcripts of evidence and submissions from the Cigna arbitration, albeit subject to redaction of parts identified by Sun/Phoenix as immaterial - was that "Cigna never agreed to cover the Unicover reinsurances and could never have been required to pay losses arising from the Unicover reinsurances".
6. Lincoln obtained permission to appeal from Tomlinson J who certified two questions of law as being of general public importance. Only one remains relevant:

"Whether the [Lincoln] Tribunal was free to depart from a binding obligation by an earlier tribunal of rights and obligations arising between Sun/Phoenix and Cigna set out in the Award of [the Cigna Tribunal] dated 14 March 2002."

7. The appeal came before Toulson J who treated this question as requiring "analysis (a) of the findings of the Cigna arbitrators and (b) of their relevance in the Lincoln arbitration". He concluded in his judgment dated 26^th February 2004 (a) that, contrary to the conclusion of the Lincoln tribunal, the Cigna tribunal had addressed the issue whether losses on the Unicover book would have been recoverable under the Cigna reinsurances but for Cigna's avoidance thereof, and had "reached an unambiguous decision" to the effect that "the Unicover whole account reinsurances were reinsured by Cigna, but that the Cigna reinsurance[s] had been validly avoided"; and (b) that, that point having been raised before and decided after hearing full evidence by the Cigna tribunal, the Cigna tribunal's "award determined the position between Sun/Phoenix and Cigna, on which the liability of Lincoln depended" and could be relied upon by Lincoln against Sun/Phoenix. This was so, he said, even though (if the Cigna award had been to opposite effect) Sun/Phoenix could "not necessarily" have relied on it against Lincoln.
8. Toulson J gave permission to appeal to this court on both (a) "the [Lincoln] Tribunal's analysis of the findings in the Cigna Award; and (b) the relevance of the findings in the Cigna Award in the Lincoln Arbitration". Mr Jonathan Sumption QC representing Sun/Phoenix has presented submissions, and Mr Ian Hunter QC representing Lincoln, has responded under four main heads, which I recapitulate as follows:

(1) The correctness of the Lincoln tribunal's analysis of what the Cigna tribunal did or did not decide.

(2) Whether the Lincoln tribunal's analysis of what the Cigna tribunal decided was open to review in court as one of law.

(3) Whether, if the Cigna tribunal decided that the Unicover book was covered by the Cigna reinsurances and the Lincoln tribunal erred in law in reaching a contrary view, the Cigna tribunal's decision in this respect created an issue estoppel or was by way of obiter dictum or collateral rather than fundamental to its award.

(4) Whether Sun/Phoenix was bound in the Lincoln arbitration by any decision by the Cigna arbitration tribunal that the Unicover book was covered by the Cigna reinsurances.

I shall consider the parties' submissions under these four heads. However, it is convenient to start with issue (2).

ISSUE (2) - WHETHER THE LINCOLN TRIBUNAL'S ANALYSIS OF WHAT THE CIGNA TRIBUNAL DECIDED WAS OPEN TO REVIEW IN COURT AS ONE OF LAW.

9. This is a curious point. Mr Sumption acknowledges that the exercise of analysis, of the effect of the Cigna award, in which the Lincoln tribunal engaged required the Lincoln tribunal to direct itself correctly on at least three points of law, viz. that the Cigna award should be examined as a whole, that any particular part(s) of it should be interpreted in the light of the particular issue(s) to which those part(s) were directed, and that any decision extracted from the Cigna award had to be sufficiently clear if it was to have even potential relevance to the Lincoln arbitration. However, he went on to submit that there were "both elements of fact and elements of law involved" in the exercise undertaken by the Lincoln tribunal. The tribunal was not, he submitted, simply construing the language of the award. It had to examine the background to determine to what issues words in the award were directed; and that might involve looking at the pleadings, the argument and the evidence "or at least [at] what, judging by the award, the arbitrators in the earlier case thought was the effect of the pleadings, the argument and the evidence". There was, he submitted, "an exercise of judgment involved" on the part of the Lincoln tribunal. Further, at times in his submissions, he suggested that, although it was agreed that Toulson J should look at the Cigna award itself (and not just at the account of it contained in the Lincoln award), this was done in effect de bene esse or without prejudice to the principle that at the end of the day the Cigna award itself was inadmissible, since no evidence or material outside the award is admissible on an appeal to the court: see Arbitration Act 1996, s.69(1).
10. Under s.69(1), it was only open to Tomlinson J to give Lincoln leave to appeal to the Commercial Court "on a question of law arising out of an award". The issue about the correctness of the Lincoln tribunal's analysis of the Cigna award was treated by Toulson J as one of law. Toulson J raised with counsel on 3^rd December 2003 the question whether he should look at the Cigna award. After referring Toulson J to his power under s.70(4) of the 1996 Act to order arbitrators to state the reasons for their award in further detail, if an award "does not set out the tribunal's reasons in sufficient detail to enable the court properly to consider the …. appeal", Mr Kendrick agreed that this could be short-circuited by, as the judge put it, "the bundle being put in front of me which has got the Lincoln contract, the Cigna contract and the award".
11. However, Mr Kendrick QC, then leading for Sun/Phoenix evidently continued to submit that the Lincoln tribunal's analysis of the Cigna award in some way involved questions of fact or mixed fact and law. Toulson J records in his judgment (para. 44):

"Mr Kendrick also submitted that if, as I conclude, the Lincoln arbitrators misinterpreted the Cigna award (in finding that the Cigna arbitrators had not directed themselves to the question whether the Unicover whole account reinsurances were within the Cigna reinsurance, subject to Cigna's avoidance), this was a purely factual error. I do not accept that submission. The interpretation of the Cigna award involved legal analysis."

12. S.69(8) enabled Sun/Phoenix to appeal from Toulson J's decision on any question of law which Toulson J had determined under s.69(1) provided that he gave leave. That in turn depended upon him concluding that "the question is one of general importance or is one which for some other special reason should be considered by the Court of Appeal". When, after Toulson J's judgment had been given against his clients, Mr Kendrick sought permission to appeal to this court under s.69(8), the transcript shows that the judge was initially reluctant to give leave in relation to the correct analysis of the findings of the Cigna arbitrators. Mr Kendrick then said:

"My Lord, I should make it clear that it has never really been my case that the court should be looking at the pleadings [i.e. in the Cigna arbitration]. My position is that you should just look at the Award and construe it.

….

I wanted to show that I was not afraid in a forensic way of dealing with the other matters. My Lord, you yourself said in the judgment that there is a question of analysis of the findings of the arbitrators and their relevance in the Lincoln Arbitration.

….

…. I would say that the two are so closely bound up that you are not really asking them to go any further. I certainly make it quite clear that I would not be suggesting for a moment that the court should be looking at the pleadings or whatever. I would have to take my stand or fall by the Cigna award, para. 84, recitals I and L etcetera."

13. Toulson J was then persuaded to give permission to appeal both on the correct analysis and on the relevance of the Cigna award in the Lincoln arbitration. No suggestion was made at that stage of any challenge to the correctness of the judge's conclusion in paragraph 44 of his judgment (set out in paragraph 11 above) or to the permissibility of Toulson J looking at the Cigna award. As I see it, no permission was sought or granted in respect of any such challenge. On the contrary, Mr Kendrick seems to me to have used the judge's conclusion at paragraph 44 as the basis for now submitting that there was an easily identifiable question of legal interpretation of the document constituting the Cigna award which should go to the Court of Appeal, along with the issue of the relevance of any conclusion which that award did on analysis contain to the Lincoln arbitration.
14. In circumstances where Sun/Phoenix obtained leave to appeal from Toulson J on the simple ground that the effect of the Cigna award is one of straightforward legal analysis, I do not find it easy to see how they can now argue the opposite or suggest that the correctness of the Lincoln tribunal's analysis of the Cigna award may depend on unstated facts not reflected in the Lincoln award or not put before Toulson J by consent. Be that as it may, I am in any event satisfied that the question before us is, as Toulson J considered, one of law.
15. In Woodhouse A.C. Israel Cocoa Ltd. S.A. v. Nigerian Produce Marketing Co. Ltd. [1972] AC 741, the House of Lords held that the meaning of a letter was a question of construction in the surrounding circumstances as stated in the award, which it was for the court to determine, rather than a question of fact. Lord Hailsham added:

"It might conceivably have been different if all the circumstances had not been made part of the award, or if all the relevant circumstances had not been fully exposed. But, on the course which the umpire rightly decided to pursue, the meaning of the letter …. is a matter of construction in the light of the surrounding circumstances and not a matter of fact."

Here, so far as appears all the circumstances necessary to reach a conclusion on the correctness of the Lincoln tribunal's interpretation of the Cigna award were before the judge, either in the award or by consent. Further, if there had been any suggestion of further relevant circumstances, which either party was not willing to put before the judge by consent, there seems little doubt that he would, under s.70(4), have ordered the Lincoln arbitrators to amplify their reasons accordingly.

16. It follows that I would reject Mr Sumption's first submission, and consider that it is for us to analyse the Cigna award for ourselves and, in the light of the construction that we put upon it, to determine the answers to the first and third issues before us, to which I next turn.

ISSUE (1) – THE CORRECTNESS OF THE LINCOLN TRIBUNAL'S ANALYSIS OF WHAT THE CIGNA TRIBUNAL DECIDED.

17. It is necessary to set out more detail which can be derived from the Cigna and Lincoln awards. The predecessors to the Occ/Acc reinsurances placed with Cigna were more limited Occ/Acc reinsurances placed on Centaur's behalf by SCB with Odyssey Re (London) Ltd. as to 50% of each layer with effect from 1^st October 1996, the other 50% of each layer being placed with John Hancock Mutual Life Insurance Co. of Boston ("JEH"). These predecessor Occ/Acc reinsurances were however intended only to protect incoming business placed with Centaur by SCB, that being stated in the placing information; and the premium income estimate for such business was accordingly only US$17 million to $24 million. The Unicover book was in contrast placed with Centaur through Rattner Mackenzie Limited; and, by April 1998, Mr Cackett of Centaur was aware that the premium income on the Unicover book was very substantial (apparently, well over US$100 million), and further that it involved a large number of "buffer layer" (or fronting) contracts on which Unicover earned no overrider.
18. Mr Cackett was never happy with Odyssey as security, and had prior to March 1998 asked Mr Jeff Butler of SCB to get Odyssey replaced by security acceptable to Mr Sun/Phoenix. Contrary to the fact, Mr Butler of SCB had told Mr Cackett that JEH was taking over Odyssey's line. After Mr Cackett became concerned about the Unicover book, he formed the idea that the SCB Occ/Acc might be expanded to cover this and any other non-SCB business and so to take a total premium income of some $160 million (before specific facultative reinsurance). He wrote to Mr Butler by fax letter of 26^th June 1998, saying that he understood that SCB had the addenda replacing Odyssey by JEH and asking Mr Butler to arrange now to broaden the Occ/Acc reinsurance cover to include non-SCB business, giving a total premium in the order of $160 million. Mr Billyard of John Hancock agreed orally in late June 1998 to include non-SCB business, and indeed signed unqualified two endorsements (endorsements no. 2) to that effect in favour of Sun (though not Phoenix). However, the Cigna tribunal inferred that Mr Billyard only gave this agreement because he was told that the purpose was to allow non-SCB business to bring the income up to the original estimate of US $17-24 million. Mr Butler nevertheless told Mr Cackett both that JEH had signed an endorsement protecting all Centaur's Occ/Acc business and (contrary to the fact) that JEH had agreed to replace Odyssey. Then, as his next step, he approached Odyssey, who had not in fact been replaced by JEH. Odyssey flatly refused to accept a revised premium income of $160 million. In late August 1998 Mr Butler approached Mr Billyard to take over Odyssey's 50% lines, including the additional premium to be generated by the inclusion of the Unicover book, but he too refused. He refused even when Mr Butler (without telling him that Odyssey had declined the additional income that would result from the Unicover book) offered JEH a retrocession of any additional 50% line back to Odyssey.
19. Against this background, first Mr Riding of SCB and then on 28^th August 1998 his superior Mr Butler went to see Mr Minter of Cigna. The reasons incorporated in the Cigna award describe this meeting as "at the very core of the parties' dispute in this arbitration" and set the scene in their opening paragraph as follows:

"1. On 28 August 1998 Mr Jeff Butler, a director of the brokers Stirling Cooke Brown Reinsurance Brokers Limited (SCB), asked Mr Paul Minter, the senior underwriter of the respondent (Cigna), to do him a favour by agreeing to reinsure the claimants (Phoenix/Sun) in respect of certain occupational accident business. Mr Minter agreed orally to do so, later describing the transaction as a "big oblig. (favour) for SCB". The dispute at the heart of this arbitration concerns the scale of the favour which Mr Minter granted. Issues have arisen concerning the satisfaction of conditions precedent to the agreement becoming binding on Cigna, the right of Cigna to avoid the transaction for misrepresentation and non-disclosure and the scope of the business actually covered by the agreement to reinsure, assuming it to be valid and binding upon Cigna. …."

20. Then, after going through the history leading up to the meeting, the Cigna award said:

"29 – Certain things are clear about this meeting. Mr Butler did not broke the risk as an attractive proposition on its merits. He said that SCB needed a favour from Cigna and he would not take no for an answer. He wanted Cigna to take over Odyssey Re's lines on certain low-level protections of Centaur's principals, one of whom, Sun Life, could not accept Odyssey as security. Mr Minter refused. Mr Butler then asked him to consider a variation to his proposal, namely, that Cigna should replace Odyssey's lines and front them for Odyssey, to whom the risk would be ceded on a 100% facultative retrocession. One or other of them mentioned Unicover, and Mr Minter was not happy about the prospects of reinsuring Unicover business. A sister company of Cigna, Connecticut General Life, was a member of the Personal Accident Unicover Pool, and Mr Minter knew enough about Unicover to be aware that Unicover's premium income was growing, and that Centaur had underwritten three year whole account reinsurance protections for Unicover as of December 1997 at what he believed to be a cheap rate. There was a discussion about Unicover…. The meeting ended with Mr Minter's oral agreement to write the Centaur reinsurances subject to a 100% facultative retrocession of the risk to Odyssey…

"30 – Doubts and disagreements remain concerning what, if anything, was said about Centaur's EPI, and what was said concerning exposure to Unicover whole account protection…"

21. In September 1998 Mr Minter initialled four off-slips to reinsure Sun/Phoenix in respect of its 50% of losses on its Occ/Acc business, without any exclusion of Unicover business. He did so on the understanding that SCB would be arranging a retrocession to Odyssey, without Cigna receiving any overrider. Odyssey initialled corresponding slips at about the same time, purporting to convert its previous line into a retrocession. It appears that Odyssey has subsequently claimed that these slips were voidable, but we are not concerned with that.
22. The Cigna award recites that:

"J) Cigna sought declarations that:

1) [Cigna] never became bound to the Occ/Acc Covers;

2) In the alternative to (1), that [Cigna] was entitled to, and did properly, avoid the four reinsurance contracts by which it participated in the Occ/Acc programme, on grounds of misrepresentation and non-disclosure;

3) In the alternative to (1) and (2), that [Cigna] was not liable to indemnify the claimants against losses in respect of their reinsurances of the Unicover Personal Accident Pool Whole Account Protections;

4) In the alternative, that, at the least, [Cigna] was not liable to indemnify the claimants against losses in respect of their reinsurance of the said Unicover Pool's "burning cost policies".

K) At the hearing the respondent also submitted that the following risks, over and above the Unicover risks, had never been validly ceded by Centaur to the Occ/Acc Covers, so that no claims lay in respect of them: -

Risks numbered by Centaur as 121, 124-125, 132, 191-192, 199-210 and 216-218"

The "burning cost policies" referred to in the third declaration sought were policies placed with Sun/Phoenix to protect Unicover against the risk of having to pay extra premium under excess of loss policies which constituted the bulk of the Unicover book reinsured with Sun/Phoenix. The Unicover book consisted as a whole of risks which were numbered 158-164 by Centaur.

23. The Cigna award was in two parts, a dispositive part and accompanying reasons. The dispositive part includes the following:

"1. WE DIRECT that our REASONS for this Award are set out in the document annexed hereto and shall be considered as part of this Award.

2. WE HOLD that in August/September 1998 [Cigna] became bound to the Occ/Acc Covers by becoming a party to four contracts of reinsurance, two with the Phoenix Claimants and two with the Sun Claimant.

3. WE HOLD and DECLARE that [Cigna] has validly and properly avoided the contracts of reinsurance by which it participated in the Occ/Acc programme on the grounds of misrepresentation and non-disclosure in the placement of these Occ/Acc covers by SCB on behalf of the claimants.

4. WE HOLD that, subject to our decision in (3) above, all the numbered risks in respect of which the claimants sought an indemnity under the Occ/Acc covers were reinsured by [Cigna] with the exceptions of risks numbered 132 and 191-192."

24. In their reasons, the Cigna tribunal further describe the issues before it:

"39 – Against this factual background the broad issues which arose for decision were as follows: -

1) Was the fronting agreement concluded between [Sun/Phoenix] and [Cigna] through SCB voidable for misrepresentation or for non-disclosure?

2) If not, was the condition precedent to Cigna's participation fulfilled, namely that there should be a 100% facultative retrocession on back-to-back terms concluded between Cigna and Odyssey?

3) If so, what business written by Centaur on behalf of Phoenix and Sun was reinsured by Cigna under the occ/acc covers?"

25. The Cigna tribunal then considered those issues under paragraphs headed (1) Avoidance (paras. 40-49), (2) Subjectivity (paras. 50-75) and (3) "Allocation" (paras. 76-84). As to avoidance, the tribunal found this with respect to the meeting of 28^th August 1998:

"44 – Mr Butler could not, however, avoid all mention of Unicover. Mr Minter had performed a review of the reinsurance protections of the Unicover pool in which Cigna's sister company was involved. He knew that the Unicover whole account protections had been placed with Centaur under a three year programme w.e.f. 1 December 1997. He believed that it was on advantageous terms for Unicover and he knew that Unicover were writing a lot of business. It was inevitable that he would want to know whether the occ/acc covers were exposed to Unicover business. Consistent with his "one step at a time" approach, Mr Butler was able to tell him that the occ/accs did not include Unicover business, because at that stage neither John Hancock nor Odyssey had accepted the expansion of the occ/accs to take the extra income (and potential exposure) generated by it. He assured Mr Minter that they did not reinsure Unicover protection. Mr Minter most probably did request an exclusion of Unicover business, but Mr Butler could not accept this as it would then have to be noted on the slip, and would therefore direct Centaur's attention to the delicate question of JEH Re's supposed agreement to cover it. He then had to find a reason for rejecting an exclusion, and must have come up with a response on the lines that Centaur wished to retain the option to cede the odd Unicover facultative declaration to the occ/accs, adding that Mr Cackett was obtaining specific protections for his exposure to Unicover business.

45 – We also find that, consistent with his general approach, Mr Butler did tell Mr Minter that Odyssey were keen to stay on the risk and would therefore be willing to be fronted by Cigna. In a sense this was true, because EIU [Odyssey's underwriting agents] were content to stay on the Occ/Accs without any additional premium and were unaware that they should have been removed.

….

47 – To front for Odyssey on the basis broked to Mr Minter was not so big a favour as to agree to front them on a greatly increased premium income. Nonetheless it was still a favour, in Mr Minter's eyes. It was unusual for Cigna to front and to write retrocessional business, Mr Minter disliked the idea of reinsuring Cigna's leading competitors and especially Mr Cackett who had cost him money in the past. In waiving an overrider he had made it easier for SCB to put the front in place. He had trusted Mr Butler's assurance that only one or two facultative Unicover cessions might be ceded to the Occ/Accs"

26. While these paragraphs themselves contain a number of findings of fact, the Cigna award continues with the following paragraph 48 headed in bold type:

"48 - Findings concerning the meeting of 28 August 1998

(1) - Income. Mr Butler had misrepresented the reinsured's estimate of the premium income on the underlying business which the reinsured intended to be reinsured under the Occ/Acc covers. The true figure given by the reinsured was in the order of $160m.

(2) - Conversely, the estimate given by Centaur was not disclosed.

(3) - Unicover. Mr Butler had misrepresented the extent to which the Occ/Acc covers were intended by Centaur to protect Unicover derived business. The intention of Centaur was to use the Occ/Accs to reinsure the programme by which Centaur's principals reinsured Unicover's whole account. It was not just a question of the odd possible facultative reinsurance. In short, the business was misdescribed. No specific fac. reinsurance was later placed to cover Centaur's exposure to Unicover.

(4) - Odyssey. Mr Butler had stated that Odyssey were keen to remain on the Occ/Accs. This was true if the business ceded to the Occ/Accs remained at the same volume, but untrue if EIU were asked to reinsure the additional volume of business which the reinsured intended to cede. Odyssey had not yet been asked if they would accept to be fronted by Cigna.

(5) - A prudent reinsurer would wish to take these matters into account when deciding whether to reinsure the risk broked to him as a front for Odyssey. They were relevant to the potential levels of income and loss exposure of the business intended to be reinsured.

(6) - Mr Minter was induced to front for Odyssey by what Mr Butler said concerning estimated premium income, by his statement that the Occ/Acc covers did not involve exposure to Unicover at all (bar possibly one or two facultative reinsurances), and by his statement that Odyssey were keen to remain on the business. Had he known the true facts concerning Mr Cackett's intended use of the Occ/Acc covers and Odyssey's inability to accept further income on the business, he would have regarded the proposed business as too uncertain and risky to write. At the very least he would have insisted on a clear exclusion of all Unicover business or a maximum limit of premium.

(7) - It follows that Cigna were entitled to avoid the agreement to front for Odyssey on the Occ/Acc covers."

27. Having concluded that Cigna was entitled to avoid its agreement to front for Odyssey on the Occ/Acc covers, the tribunal said in paragraph 49 that issues (2) and (3) listed in paragraph 39 of their reasons (cf paragraphs 22 and 23 above) did not arise, but that they would "record their decisions on them in deference to the detailed submissions presented". The tribunal did this in paragraphs 50 to 84 of its reasons. It referred to designation in the slips initialled by Mr Minter of the business covered as:

"All business written by the reinsured and allocated to their Personal Accident Account in respect of Workers' Compensation and Occupational Accident business only."

It considered and rejected an argument by Cigna that, in order for the defendants to be able to claim losses arising on a risk reinsured by the Occ/Accs, some act of allocation of that risk to the Occ/Acc covers was required. It held that prima facie it was sufficient that the defendants had internally coded a risk as part of their Occ/Acc business. However, it continued:

"80 – We say "prima facie" because the coding cannot be conclusive. First, it must be a valid coding. A risk which was really wider than occupational accident or workers compensation would not qualify for protection, however coded. Secondly, good faith errors could be corrected under the generous E & O clause in the occ/acc contracts… Thirdly, it would be open to Centaur to agree with reinsurers on the Occ/Acc programme that a particular risk was not to be protected by the Occ/Acc or that business of a particular type falling within the general definition above was not to be protected under the covers. An example of the latter was the exclusion of Occ/Acc risks placed with Centaur by brokers other than SCB before Endorsement No. 2 was agreed. In other words, a qualifying risk coded PO or PW could be taken off cover by a specific decision of Centaur pursuant to the agreement of reinsurers."

28. Applying that analysis, the Cigna tribunal concluded that two risks in respect of which Sun/Phoenix had claimed indemnity were not protected by the Cigna Occ/Acc covers, viz. risk No. 132 (a whole account protection of JEH's personal accident account not restricted to workers' compensation and occupational accident business) and risks Nos. 191-2 (in respect of which Centaur had by fax dated 8^th September 1998 agreed with Odyssey that they should not be reinsured under the Occ/Acc covers, an agreement of which the tribunal considered that Cigna would have the benefit). But the tribunal rejected an argument by Cigna that the Unicover burning cost protection policies were outside the scope of the Occ/Acc covers. It said:

"It was argued by CIGNA that the burning cost protection policies described in paragraph 13 above were also outside the scope of the covers. The question turns neither upon whether these contacts are insurances or reinsurances, nor upon whether they are personal accident policies, which clearly they are not. It turns rather upon whether in this instance these additional premium protection policies can be described as business allocated by Centaur to their Personal accident Account" in respect of Workers' Compensation/Occupational accident business only". We think that it can. These were policies protecting Unicover against the contingency of having to pay additional premiums under their swing-rated excess of loss policies, which admittedly came within the definition. They were part and parcel of Centaur's Worker' Comp/Occ/Acc. Book of business."

29. Having thus identified those few risks which they found fell outside the scope of the Occ/Acc covers, the tribunal continued:

"84 - Apart from the above risks we hold that all the risks listed in paragraphs 132 –133 of [Cigna's] outline final written submissions were protected by the Occ/Acc covers subject to the earlier issues decided in this arbitration. If we had found that Cigna were bound to the Occ/Acc covers subject to an exclusion of Unicover whole account protections, then we should have regarded that as constituting an agreement to exclude that particular business from protection from under them."

When setting out this passage in his judgment, Toulson J recorded, correctly, that the risks listed in paragraphs 132-133 of Cigna's outline final submissions included the Unicover whole account book of reinsurances.

30. Toulson J expressed as follows his conclusion on the question whether the Cigna tribunal had decided whether the Cigna reinsurances would, apart from their avoidance, have covered the Unicover book:

"On my reading of their award, the Cigna arbitrators addressed that issue and reached an unambiguous decision on it. They concluded that the Unicover whole account reinsurances were reinsured by Cigna but that the Cigna reinsurance had been validly avoided. "

He explained this conclusion as follows:

"39 - The Lincoln arbitrators referred to recital K as supporting an argument that the Cigna arbitrators may not have considered whether the Unicover whole account reinsurances fell within the Cigna reinsurance, but for Cigna's successful avoidance. They did not refer to recital J (the recitals at J and K being a summary of submissions which had been advanced by Cigna in its pleadings and at the hearing) or more significantly to paragraph 84 of the Cigna arbitrators' reasons. Paragraph 84 shows that in arriving at their holding in paragraph (4) of their award the Cigna arbitrators had considered but rejected the conclusion that there had been an agreement to exclude Unicover whole account risks from the scope of cover written by Cigna.

….

40 - Mr Kendrick submitted that the sentence in paragraph 84

"If we had found that Cigna were bound to the occ/acc covers subject to an exclusion of Unicover whole account protections, then we should have regarded that as constituting an agreement to exclude that particular business from protection under them"

should be read as meaning that if the Cigna arbitrators had not found that the Cigna reinsurance had been properly avoided, they would have found that there was an agreement to exclude Unicover whole account protections from the scope of cover. He also pointed out that in paragraph 80, where the Cigna arbitrators referred to the possibility of Centaur and Cigna agreeing that a particular risk was not to be covered by the occ/acc covers, they were considering an allocation argument in relation to risks other than Unicover whole account risks. That observation is correct, but it is not in my view a persuasive reason for reading paragraph 84 in the way that Mr Kendrick suggests. If the experienced Cigna arbitrators had intended paragraph 84 to mean what Mr Kendrick suggests, I am sure that it would have been worded very differently and that paragraph (4) of the award would have identified the Unicover whole account reinsurances among the risks excepted.

41 - Paragraph 4 of their award and paragraph 84 of their reasons are also consistent with the Cigna arbitrators' findings about the meeting on 28 August 1998. As I read their reasons, they found that Mr Minter agreed to front for Odyssey on the occ/acc covers, amended to include business from all sources, without there being any exclusion of Unicover (which Mr Minter probably requested), because Mr Minter accepted Mr Butler's false representations about Centaur's intended use of the covers. Mr Kendrick made the point that in their reasons the Cigna arbitrators at times referred to this as an assurance (for example, in paragraph 47). The word "assurance" could mean either an assurance of intention (that is, a statement of intent) or a contractual promise. However, there would have been problems reconciling a contractual promise with a refusal to agree to an exclusion from cover, and paragraph 84 shows that the Cigna arbitrators had clearly in mind the question whether there was an agreement to exclude Unicover whole account risks from the cover.

42 - I was referred to underlying documents, including extracts from the pleadings and submissions in the Cigna arbitration. In my view they do nothing to weaken the natural reading of the award. They tend, if anything, to confirm it. For they show that Mr Kendrick in his closing submissions canvassed with the arbitrators the possibility of a finding that there was an agreement between Mr Minter and Mr Butler to exclude the Unicover whole account reinsurances from the occ/acc covers, and so it was a live issue for the arbitrators to decide.

43 - As to the alternative argument that there was a gap in the Cigna arbitrators' findings, Mr Kendrick put forward various forms of argument, which he says were not considered by the Cigna arbitrators, for holding that the Cigna reinsurance would not have protected Sun/Phoenix against losses under the Unicover whole account reinsurances; but they had the common element that there was an oral agreement between Mr Minter and Mr Butler that the Unicover whole account reinsurances were not to be covered by the Cigna reinsurance. Whether such an agreement, if there was one, was properly to be described as giving rise to an agreement to exclude or a collateral agreement or an estoppel by convention is a matter of legal labelling, about which there was discussion between Mr Kendrick and the Cigna arbitrators during his closing submissions. They considered whether there was an agreement to exclude the Unicover whole account reinsurances and, on my reading of their award, they concluded that there was not."

31. Mr Sumption's primary submission before us was that the Cigna award was simply silent on the question whether the Unicover book would have been reinsured under the Cigna Occ/Acc reinsurances, but for their avoidance. The conclusion that the Cigna reinsurances were avoided made it unnecessary to consider their scope in that respect. Alternatively, he submitted that the most that might be inferred is that the arbitrators held an opinion on a point which they did not decide in terms, and that that could not be good enough to found an issue estoppel even between the same parties. At this point, the argument on issue (1) tends to merge into the argument under issue (3) before us.
32. Much of the argument before us was directed to identifying the proper scope and effect of paragraph (4) of the dispositive part of the award. Mr Hunter argues that this paragraph was intended to respond to recitals J(3) and (4) as well as K, whereas Mr Sumption argues that it only responded to recitals J(4) and K. Mr Sumption cross-referred paragraph (4) to paragraphs 76-84 headed "Allocation", and drew attention to the absence in those paragraphs of any express reasoning addressing the Unicover book as a whole (as opposed to the Unicover burning cost policies in particular). The brief reference to the Unicover book in the second sentence of paragraph 84 should, he submitted, be understood as a neutral statement that the tribunal had not had to decide whether or not there was any such exclusion.
33. Mr Sumption's submission does not appear to me to meet the fact that the first sentence of paragraph 84 refers to "all the risks listed in paragraphs 132 to 133 of [Cigna's] outline final written submissions". These included the Unicover risks Nos. 158-164, i.e. the whole Unicover book. Further, the Cigna tribunal had in paragraph 79 made clear its view that the application of the Cigna reinsurances did not depend on any particular act of allocation to those reinsurances. Rather (subject to any specific agreement that particular risks would not be covered thereby) the Cigna reinsurances applied to all Occ/Acc business properly falling within the designation in the Cigna slips (that is business allocated to Cigna's Personal Accident Account as a sub-species of that general class of business called Workers Compensation/Occupational Accident business). Nevertheless, Mr Sumption is justified in drawing attention to the absence from paragraphs 76-84 of any reasoning explicitly addressing the Unicover book as a whole, in circumstances where it had been Cigna's alternative case that, if the Occ/Acc reinsurances were not avoided, the Unicover book fell outside or were excluded from their scope.
34. However, the reason for this omission seems to me to have been that the Cigna tribunal had already, in the course of making its findings regarding avoidance, expressed its view that the Cigna reinsurances would, but for their avoidance, have covered the Unicover book. The last sentences of paragraphs 44 and 47 are I consider particularly relevant in this regard. That still raises the question why the point was not more clearly addressed. The probable answer is that it had been in neither party's particular interest in the Cigna arbitration to tie their case too closely to a proposition that the Unicover book was not reinsured. The major losses suffered by Sun/Phoenix related to the Unicover book. Sun/Phoenix, relying on the slips procured by Mr Butler, were contending that the Cigna reinsurances covered their Occ/Acc business on an unlimited basis, and were attempting in this way to recover such losses from Cigna. Cigna's primary case was on the other hand that the Cigna reinsurances had been validly avoided for misrepresentation or non-disclosure relating primarily to the Unicover book; and the most obvious - although, as I shall demonstrate, not the only - way of running such a case, was on the basis that the Unicover book fell within the terms of the reinsurances, so that matters material to it should have been fully and accurately represented and disclosed. However, if for any reason their claim to avoid had failed, then it would have been in Cigna's interest to pursue an opposite analysis, namely that the Unicover book fell outside the scope of the Cigna reinsurances.
35. Consistent with this ambivalent approach, recital J(3) in the Cigna award records that Cigna's claim to a declaration that the Unicover book fell outside the Cigna reinsurances was as an alternative to their claim to a declaration establishing avoidance. Sun/Phoenix's position was also, or became also, ambivalent. The judge in paragraph 42 of his judgment refers to, and we were also shown, a transcript of Mr Kendrick's very last submissions before the Cigna tribunal. Mr Kendrick there sought to raise, from what may well by then have appeared to be impending shipwreck, an alternative case to the effect that, if Mr Butler misrepresented to Mr Minter that the Unicover whole account book (as opposed to a few other possible though as yet unspecified Unicover risks) was not intended to be and would not be reinsured under the Cigna reinsurances, the effect would be that the Unicover book was excluded from the scope of the Unicover risks. But the tribunal cannot be blamed if it did not treat this unpleaded, last minute alternative seriously. The same transcript shows that Mr Legh-Jones QC's immediate reaction was that a conclusion that the Unicover book fell outside the scope of the Cigna reinsurances would not help Sun/Phoenix. Mr Butler would still have been placing the risk on a wholly different basis to that intended by Sun/Phoenix (and indeed by Mr Butler) – since both of them intended to treat the Unicover book as reinsured.
36. In this connection, the Cigna tribunal's formulation of paragraph 48 of its reasons, under the bold heading "Findings concerning the meeting of 28 August 1998", is in my view significant. These findings (set out in paragraph 26 above) focus on Mr Butler's misrepresentations and non-disclosure regarding the business intended by Sun/Phoenix to be reinsured (and intended by him in his dealings with Sun/Phoenix as well as, presumably, Cigna to be treated as reinsured) under the Cigna reinsurances. There is no express finding in paragraph 48 that the Unicover book would actually have been within the scope of the Cigna reinsurances. The last sentences of paragraph 48(6) are worth repetition:

"Had he [i.e. Mr Minter] known the true facts concerning Mr Cackett's intended use of the Occ/Acc covers and Odyssey's inability to accept further income on the business, he would have regarded the proposed business as too uncertain and risky to write. At the very least he would have insisted on a clear exclusion of all Unicover business or a maximum limit of premium."

That, it seems to me, is the common sense of the position. Regardless whether or not the Unicover book was within the scope of the Cigna reinsurances, Mr Butler's disingenuous intentions, and Sun/Phoenix's more innocent intentions arising as a result of their having also been misled by Mr Butler, were calculated to lead to major problems - indeed to precisely the dispute and arbitration that did arise between Sun/Phoenix and Cigna. No-one would sensibly write a reinsurance on that basis - and this is, as I see it, the gist of the last two sentences of paragraph 48(6) of the Cigna award (paragraph 26 above).

37. That said, I would agree that the Cigna tribunal was, when it came to express its views on points unnecessary for its decision, prepared itself to accept and treat the Unicover book as within the scope of the Cigna reinsurances. The tribunal could otherwise hardly have dealt with the question whether the Unicover burning cost policies were within such reinsurances in the terms it did. Nor do I think that it would otherwise have expressed itself as it did in the first and second sentences of paragraph 84 of its reasons. I would accordingly construe paragraph (4) of its dispositive award as embracing the Unicover book. But whether this paragraph should in these circumstances be treated as giving rise to any issue estoppel even between the parties is a different matter to which I next turn.

ISSUE (3) - WHETHER, IF THE UNICOVER BOOK WAS COVERED BY THE CIGNA REINSURANCES, THE CIGNA TRIBUNAL'S DECISION TO THAT EFFECT CREATED AN ISSUE ESTOPPEL OR WAS BY WAY OF OBITER DICTUM OR COLLATERAL RATHER THAN FUNDAMENTAL TO ITS AWARD

38. Mr Hunter's first submission is that the answer to this question is to be found in s.58(1) of the Arbitration Act 1996, which provides that:

"Unless otherwise agreed, an award made by the tribunal pursuant to an arbitration agreement is final and binding both on the parties and on any persons claiming through or under them."

However, I do not regard this as affecting the general principles by reference to which courts determine the extent to which judgments or awards are to be treated as final and binding in the instant or any future dispute. S.58(1) repeats in effect s.16 of the Arbitration Act 1950:

"Unless a contrary intention is expressed therein, every arbitration agreement shall, where such a provision is applicable to the reference, be deemed to contain a provision that the award to be made by the arbitrator or umpire shall be final and binding on the parties and the persons claiming under them respectively."

That was the provision in force in 1965, when in Fidelitas Shipping Co. Ltd. v. V/O Exportleb [1966] 1 QB 630 a Court of Appeal consisting of Lord Denning MR, Danckwerts LJ and Diplock LJ stated the principles governing issue estoppel, in considering whether an interim arbitration award issued in London in 1961 barred shipowners from subsequently relying on a new point in order to try to avoid the impact of a cesser clause on their claim against charterers.

39. Lord Denning stated the relevant principles as follows (p.640):

"The law, as I understand it, is this: if one party brings an action against another for a particular cause and judgment is given upon it, there is a strict rule of law that he cannot bring another action against the same party for the same cause. Transit in rem judicatam: see King v. Hoare (1844) 13 M & W 494, 504. But within one cause of action, there may be several issues raised which are necessary for the determination of the whole case. The rule then is that, once an issue has been raised and distinctly determined between the parties, then, as a general rule, neither party can be allowed to fight that issue all over again. The same issue cannot be raised by either of them again in the same or subsequent proceedings except in special circumstances, see Badar Bee v. Habib Merican Noordin [1909] AC 615, 623 per Lord Macnaghten."

He went on to consider the principle in Henderson v. Henderson (1843) 3 Hare 100 (with which we are not concerned on this appeal), and to say that "Like principles apply in arbitration". Diplock LJ said (p.643):

"Issue estoppel applies to arbitration as it does to litigation. The parties having chosen the tribunal to determine the disputes between them as to their legal rights and duties are bound by the determination by that tribunal of any issue which is relevant to the decision of any dispute referred to that tribunal. An arbitrator today has power to make an interim award determining particular issues separately from other issues in the arbitration."

40. Mr Sumption submits that it was not "necessary" or "relevant" for the determination of the dispute between Sun/Phoenix and Cigna for the Cigna tribunal to determine whether the Unicover book of business would or would not have been reinsured by the Cigna reinsurances, if these had not been avoided. He points out that the tribunal itself stated at the end of the section in its reasons dealing with avoidance that issues (2) and (3) before it "do not arise in view of our conclusion" on avoidance, and submits that paragraph (4) of the dispositive award was in the circumstances purely obiter. Further, in so far as it can be suggested that the Cigna tribunal, in the course of the reasoning leading to its conclusion that there had been a valid avoidance, expressed any view that the Unicover book would have been within the scope of the Cigna reinsurances, that was a passing and inessential view in the light of the considerations which I have mentioned in paragraphs 34 to 36 above.
41. Mr Sumption referred to Spencer Bower, Turner and Handley on The Doctrine of Res Judicata (3^rd Ed.) (1996), paragraph 201 of which states:

"Even where the court has expressly determined the same issue in the earlier proceeding an issue estoppel will not necessarily arise. Only determinations necessary to the decision, and fundamental to it, will found an issue estoppel. Other determinations, however positive, cannot."

An equivalent though slightly differently worded passage in the second edition of Spencer Bower was endorsed in dicta of May LJ and Balcombe LJ in In re Norway's Application (No.2) [1990] 1 AC 723, 743 and 751-2. The requirement that the prior determination should have been "necessary", rather than "merely collateral or obiter" was emphasised in the recent Court of Appeal decision in Good Challenger Navegante SA v. Metalexportimport SA (The Good Challenger) [2004] 1 LlR 67. In a passage from Blair v. Curran (1939) 62 CLR 464, 533, cited in Spencer Bower, Dixon J said:

"The difficulty in the actual application of these principles is to distinguish the matters fundamental or cardinal to the prior decision or judgment, decree or order or necessarily involved in it as its legal justification or foundation from matters which even though actually raised and decided as being in the circumstances of the case the determining considerations, yet are not in point of law the essential foundation or groundwork of the decision, decree or order."

Spencer Bower at paragraph 201 also cites Fullagar J's observation in Brewer v. Brewer (1953) 88 CLR 1, 15 that:

"Issue estoppel only applies to issues. There is no estoppel as to evidentiary facts found in the course of determining the affirmative or negative of an issue."

42. In In re State of Norway's Application (No.2) May and Balcombe LJJ indicated that the distinction between fundamental and collateral matters is not mechanical but flexible, and that it is to be drawn according to the circumstances and justice of the particular case. May LJ cited Lord Denning's statement in Fidelitas that the rule was not inflexible, as well as Lord Upjohn's statement in Carl Zeiss Stiftung v. Rayner & Keeler Ltd. (No.2) [1967] 1 AC 858, 947 that:

"All estoppels are not odious but must be applied so as to work justice and not injustice and I think the principle of issue estoppel must be applied to the circumstances of the subsequent case with this overriding consideration in mind".

43. One test, but not a conclusive test, sometimes applied to determine whether a determination on an issue can have been fundamental is whether there could have been an appeal from that determination. Spencer Bower puts the matter as follows at paragraph 205:

"A decision of fact against the party who succeeded will not found an estoppel because it cannot be fundamental to the decision. It would be unjust to make such a decision the foundation of an estoppel, for no appeal is available to the person against whom it was given."

44. A similar passage was approved by Balcombe LJ in In re State of Norway's Application (No.2) at p.752. Woolf LJ at p.772-3 also regarded the absence of any realistic prospect of an appeal as inconsistent in that case with any issue estoppel. In The Good Challenger the Court of Appeal left open, in relation to decisions with two separate ratios leading to the same result, the correctness of the immediately following passage from paragraph 205 in Spencer Bower:

"A similar argument applies where several factual grounds are advanced as alternative bases for a cause of action and the court finds more than one in favour of the party who succeeds. No estoppel can be founded on any of the separate findings, for the party failing on such issues cannot appeal against any of them separately. To succeed on appeal he must succeed on all the issues, and if the finding on one is good, this will be fatal. There will be a cause of action estoppel, but the separate issues will not ground issue estoppel because none was fundamental to the decision."

45. Applying these principles to this case, paragraph (4) of the dispositive part of the award was not necessary for the Cigna tribunal's decision. It was directed to an issue which the Cigna tribunal correctly stated in paragraph 49 of its reasons did not arise in view of its decision on avoidance. Although expressed as part of the dispositive award, it was in fact obiter. Cigna, which won on avoidance, had no basis for appealing against it. If Sun/Phoenix had been able to appeal on the issue of avoidance, then the scope of the Cigna reinsurances might have become a necessary issue for determination, but there was no such appeal.
46. I have already considered the alternative way of looking at the Cigna tribunal's treatment of the question whether the Unicover book would have been covered under the Cigna reinsurances, but for their avoidance. In the light of what I have said in paragraphs 34-37 above, while I consider that the Cigna tribunal did hold and express the view that the Unicover book would have been covered by the Cigna reinsurances, I cannot regard its expression of that view as having been fundamental to its decision on avoidance, or as anything other than collateral. It is, I consider, clear that the Cigna tribunal would have arrived at precisely the same conclusion as it did regarding avoidance, in the light of the way it expressed its central findings of fact in paragraph 48 of its award, even if it had formed an opposite view to the effect that the Unicover book fell outside or was excluded from the scope of the Cigna reinsurances.
47. As will appear when considering issue (4) below, Mr Hunter's submissions in support of Toulson J's decision start from and depend upon a proposition that the Cigna award gave rise to an issue estoppel as between Sun/Phoenix and Cigna on the issue whether the Unicover book was covered under the Cigna reinsurances. It follows from this and my conclusions on issue (4) that Sun/Phoenix's appeal to this court succeeds.

ISSUE (4) - WHETHER SUN/PHOENIX WERE BOUND IN THE LINCOLN ARBITRATION BY ANY DECISION BY THE CIGNA ARBITRATION TRIBUNAL THAT THE UNICOVER BOOK WAS COVERED BY THE CIGNA REINSURANCES

48. Mr Hunter submits that, if, on a proper reading, the Cigna award determined that the Unicover book was covered by the Cigna reinsurances in a manner giving rise to an issue estoppel in any subsequent dispute between Sun/Phoenix and Cigna, then Lincoln, whose rights or liabilities depend upon the legal position as between Sun/Phoenix and Cigna, were entitled to rely against Sun/Phoenix upon that determination to establish that position.
49. The inspiration for Mr Hunter's submission is an obiter dicta passage in the judgment of Saville J in George Moundreas & Co. SA v. Navimpex Centrala Navala [1985] 2 Ll.R. 515. The relevant claim was by Moundreas for breach by Navimpex of a commission agreement relating to a separate shipbuilding contract between Navimpex as builders and Solway Maritime Ltd as buyers of hull 730. Saville J held that it was an implied term of the commission agreement that Navimpex would not break the shipbuilding contract, so as to deprive Moundreas of commission as and when future instalments of the price became payable. The issue thus arose whether Navimpex had broken the shipbuilding contract. Moundreas pointed to a finding to that effect in an arbitration between Navimpex and Solway. But they also called evidence and proved the breach independently. The second paragraph in the following passage from Saville J's judgment (p.520) is in these circumstances expressed provisionally and in obiter terms:

"The plaintiffs urged me to treat the award of establishing that the sellers were in breach of the shipbuilding's contract to a degree justifying the cancellation of the buyers. The defendants objected on the grounds that the award only represented the opinion of the arbitrators and was thus inadmissible as evidence: they relied upon Hollington v. Hewthorn, [1943] 1 K.B. 587.

It seems to me that where the rights or obligations of the parties to a contract are determined by the contractual machinery of arbitration under that contract there is something to be said for the view that the result that the arbitrators reach can (in the absence of special circumstances) be treated in effect as part of the contract and thus established by third parties in the same way as any contract can be proved. Thus in the present case the arbitrators have concluded that the sellers [sic] had a right to cancel the contract and to claim damages as the result of the failure of the buyers [sic] to perform their obligations under the contract. As between the parties that is now the contractual position as determined by the contractual machinery of arbitration – and it is difficult to see why a stranger to the contract cannot prove that contractual position by simply producing the award as he can prove other contractual rights and obligations by simply producing the contract.

Be that as it may, the plaintiffs took the precaution of calling much the same evidence on the buyers' major complaint as was presented to the arbitrators and on the basis of the evidence I have heard I am satisfied that the hatch covers and coamings of this vessel were not constructed in accordance with the standards stipulated in art. 1.4. of the shipbuilding contract and that the defects were so serious that the buyers were fully justified in refusing to accept delivery of the vessel and in eventually cancelling the contract for breach by the sellers."

50. Saville J's dicta give rise to some immediate questions. First, although Saville J spoke of a stranger to the contract proving the contractual position by producing the award (without addressing the problem that such documents are usually private), his expressed rationale was that the award crystallised or created what "is now the contractual position" between the contractual parties. This rationale would suggest that it should not matter in whose favour the award was; and that Navimpex just as much as Moundreas should have been able to "prove the contractual position" by producing the award. Mr Hunter's case is different. He submits that a third party has a one-way entitlement to rely on such an award, to the extent that it favours him. He submits that a third party would not be bound by a determination unfavourable to him; he could controvert it by any means or argument, factual or legal, available to him in the absence of any award. The rationale Mr Hunter advances is that the principle in Moudreas should itself be subject to the overriding principle that a stranger cannot be bound by a judgment or award to which he was not party or privy. Toulson J in the present case accepted this approach. He suggested that "the real considerations of fairness should be what is most fair to the parties and will avoid bringing the administration of justice in disrepute". Second, the situations in which Saville J contemplated that his suggested principle might apply are not defined. He suggested that it might be subject to "special circumstances". It is not clear how far he intended only to embrace situations where the evidential position was effectively identical in the two cases, or what the position would be in the case of a prior award or judgment which had gone effectively by default or by agreement. It can, however, be said in response to this uncertainty that a similar problem arises in relation to issue estoppel, where the courts have emphasised a need for flexibility and caution. Toulson J thus considered that it could make "a difference in terms of justice whether the earlier decision was the product of an informal arbitration, in which the arbitrator had not properly addressed the arguments", or whether "the issues had been fully and properly investigated and addressed in a reasoned decision", and that "the confidentiality of the arbitration proceedings could also be a relevant factor". One is, however, entitled both to question the basis for invoking in relation to arbitration the broad considerations of justice on which Toulson J relied, and to question how feasible and satisfactory such an approach would be. This is particularly so in the context of prior arbitration proceedings, which will have taken place separately from the proceedings in which a stranger later seeks to invoke their outcome, and in relation to which problems of privacy and confidentiality would be highly likely to arise.
51. The application of the principle suggested by Saville J seems to me particularly difficult in the context of the present case. The reasons are connected with the nature of the Cigna award and with what I have already said on issue (3) above. Had we been concerned with an award determining simply that the Unicover book was covered by the Cigna reinsurances (and therefore entitling Sun/Phoenix to claim their losses on that book from Cigna), Sun/Phoenix would never ordinarily have claimed to be reinsured by Lincoln in respect of such losses at all. Perhaps, if Cigna had become insolvent, Sun/Phoenix might have been tempted to assert that the Cigna award was wrong, and to claim against Lincoln. Had they done so, Lincoln might have argued with some force that, whatever the scope of the Cigna reinsurances apart from the Cigna award, after the award the Unicover losses could not constitute "retained" losses within the meaning of the Net Retained Lines clause. But that would not mean that Sun/Phoenix was bound by the Cigna award as against Lincoln. It is more that the Cigna award – and the effective reinsurance to which it gave rise - would itself have brought about a factual state of affairs to which the Net Retained Lines clause would apply. Had we, on the other hand, been concerned with an award declaring simply that Sun/Phoenix's had no right to be indemnified by Cigna in respect of losses on the Unicover book, because that book was outside the scope of the Cigna reinsurances, I would have seen force of Mr Hunter's submission that a stranger to the award in Lincoln's position should not be bound. The underlying reasons are identified in cases such as In ex p. Young, in re Kitchin (1881) 17 ChD 668 and Bruns v. Colocotronis (The Vasso) [1979] 2 Ll.R. 412 to which I will come (paragraph 55 below). The argument that the award declaring that there was no liability was a relevant "fact" would not assist, if it was open to Lincoln to look back at the position prior to and apart from the award, and to show that that involved a reinsurance by Cigna of the Unicover book. Accordingly, it may be that, if this case had concerned a simple award declaring that Cigna was or was not liable in respect of the Unicover book, the result would have been a legal position broadly to the effect advocated by Mr Hunter. But this result would not flow from any principle such as that suggested by Saville J.
52. The difficulty about issue (4) is that it arises on the basis of an award which was not a simple determination of the question of liability, but which is said to have determined that question either unnecessarily or when deciding the issue of avoidance on which it did focus. For most purposes it is I think possible to cut the Gordian knot. If it was unnecessary for the Cigna tribunal to determine the question whether the Unicover book was protected, or if it only did so collaterally, then issue (4) simply cannot arise. Issue (4) could only arise (a) if the Cigna tribunal were (contrary to my view) to be regarded as having determined the question as a necessary part of its reasoning in relation to avoidance, and (b) if, as a result, it could be said that an issue estoppel would exist as between Sun/Phoenix and Cigna, in any subsequent proceedings in which the same question in some way arose as between them. Only in this remote situation, could the further question arise in this case whether there exists any principle along the lines mooted by Saville J.
53. With this introduction, I start with some basic propositions. The principles of res judicata and issue estoppel apply between parties to the original proceedings or their privies. It is not, and could not be, suggested here that either principle has any direct application as between Sun/Phoenix and Lincoln. In Hollington v. Hewthorn [1943] KB 587 the Court of Appeal held that not even a criminal conviction for careless driving was admissible evidence to prove negligence in a subsequent civil action. The decision has been criticised, and has been qualified (for present purposes immaterially) by statute but otherwise remains law: cf Secretary of State for Trade and Industry v. Bairstow [2003] EWCA 321; 3 WLR 841. It is also relevant to bear in mind that the submission by Mr Denning KC in Hollington v. Hewthorn was not that the criminal conviction should be regarded as conclusive evidence, but that it should be admissible as prima facie evidence of negligence (cf p.589). S.11 of the Civil Evidence Act 1968 now achieves that result generally in civil proceedings (apart from for defamation proceedings, in which s.13 makes the conviction conclusive). But there is nothing to give a civil judgment, still less an arbitral award, evidential value in establishing facts needing to be proved in separate proceedings against a stranger to the original proceedings.
54. Further, Mr Hunter does not suggest that there is on this appeal any relevance in the principle commonly taken to have been established in Henderson v. Henderson (1843) 3 Hare 100, 114, that a party is normally expected to bring forward every point which would assist him in pursuing or defending a particular claim, and will be debarred from doing so later. That is a principle which applies between parties to proceedings. It has recently been explained as "separate and distinct from cause of action and issue estoppel" (although based on the same underlying public interest) and as depending upon the court's power to prevent misuse or abuse of its process: Johnson v. Gore Wood [2002] 2 AC 1. It is not suggested that the Lincoln tribunal relied or could have been relied upon any such principle when saying that it was bound by whatever the Cigna tribunal had decided.
55. There are a number of situations in which courts have considered attempts by A to rely as against B upon rights or liabilities established in separate proceedings between A and C. One familiar context is that of principal and surety. In In re Kitchin (above), the Court of Appeal held that, in the absence of explicit words, a judgment or award obtained by a creditor against the principal debtor does not bind, and is not evidence against, a surety, who is entitled to have the liability proved as against him in the same way as against the principal debtor. The court pointed out that the judgment or award might have resulted from the principal debtor's neglect to defend or admission. The principle was applied in the context of an arbitration award in The Vasso (above), where a guarantee had been given of the due performance and payment of all liabilities and obligations of shipowners arising under or out of certain written agreements which included an arbitration clause. Robert Goff J applied In re Kitchin, holding that the guarantee did not extend to the obligation to honour an award. He pointed out that an arbitration clause has special characteristics distinguishing it from the main obligations of the contract, as established by Heyman v. Darwins [1942] AC 356 – and confirmed, I would add, by subsequent cases. The distinction thus drawn, when considering the position of third parties, between on the one hand the main obligations of a contract and on the other hand an arbitration clause and any award relating to those obligations contrasts with, and in my view throws doubt upon, Saville J's assimilation of the two in Moudreas. The general principle in In re Kitchin and in The Vasso was further applied in Hayter v. Nelson [1990] 2 LL.R. 265, where a reinsurer, who had been held liable to original insurers under an arbitration award, claimed to recover the amount of such liability from his retrocessionaires. Saville J there said, citing In re Kitchin, that just as a surety could not be bound by such an award or judgment, absent agreement, so a retrocessionaire was not.
56. In Stargas SpA v. Petredec Ltd. (The Sargasso) [1994] 1 Ll.R. 412, it had been determined by arbitration award that charterers were liable to sub-charterers under a voyage charter for contamination of the cargo. Charterers claimed from disponent owners an indemnity by way of damages under the head time charter. Clarke J held that it was for charterers to prove both the breach of the head charter and that it put them in breach of the voyage charter, but that, once they had done this, the arbitrators' award (based on their conclusion that there had been a similar breach) could be regarded as caused by the disponent owners' breach and so as quantifying the damages flowing from that breach, unless it could be shown that the charterers had failed to mitigate their loss or that the award was perverse or unreasonable. Clarke J considered Moudreas as well as The Vasso and Hayter v. Nelson in the course of his judgment, but his decision appears to me to have turned on a relatively limited point of causation and not to bear on the issue that we have to decide.
57. In Hollington v. Hewthorn Goddard LJ, after giving his reasons for considering that a judgment was neither conclusive nor of evidential weight in subsequent proceedings, went on:

"A judgment, however, is conclusive as against all persons of the existence of a state of things which it actually affects when that state of things is a fact in issue. Thus, if A sues B, alleging that owing to B's negligence he has been held liable to pay £x to C, the judgment obtained by C is conclusive as to the amount of damages that A has had to pay C, but is it not evidence that B was negligent …. and B can show, if he can, that the amount recovered was not the true measure of damage."

That reasoning appears to me consistent with that of Clarke J in The Sargasso. The fact of the judgment for £x can be relied upon as the causative result of B's breach of duty, although it is open to B to try to show that (as between himself and A) his liability to A should give rise to damages measured on some other basis. Where the breaches of duty are effectively the same, as in The Sargasso, this may involve showing either that A failed to mitigate, by taking some obvious point in answer to C's claim, or that the amount awarded was for some other reason so erroneous as to break the chain of causation, or of course that the contract between A and B itself stipulated for some other measure or contained relevant limitations or exceptions.

58. In Sacor Maritima SA v. Repsol Petroleo SA [1998] 1 Ll.R. 518, shipowners had settled cargo-receivers' claim under bills of lading and then successfully pursued time charterers for an indemnity in arbitration. The award proceeded on the basis that the cargo contamination had occurred as a result of the acts or omissions of a surveyor appointed by voyage sub-charterers. Charterers then sought an indemnity from voyage sub-charterers in arbitration but failed because the arbitrators appointed under the sub-charter held (after hearing different evidence) that the contamination was due to the charterers' own fault in failing to instruct the master properly. I held on an appeal to the Commercial Court that the sub-charterers and the sub-charter arbitrators were in no way bound by the findings of the time charter arbitrator. As in The Sargasso, it was for the charterers to prove breach by the sub-charterers (and that it led to shipowners being entitled to claim an indemnity from charterers) before they could hope to rely on the time charter award to quantify their damages. I referred to Saville J's words in Moundreas. I said that, assuming that they represented the law, they did not assist the charterers. I added on the same basis that it made sense to treat the award in Moundreas as determining that the shipbuilders had been in breach of the shipbuilding contract with their buyers, but that in Sacor it was not sought to look to the time charter award to establish charterers' liability to owners, but to transpose findings of fact and conclusions of causation from it wholesale into the voyage charter arbitration. Since this was a passage based on an assumption that the dicta in Moudreas represented the law, it does not assist to resolve the issue whether and what validity and application Saville J's dicta may actually have.
59. As examples of a situation in which B and C commonly agree by contract to be bound by the determination by judgment or award of the legal position as between A and B, Mr Sumption referred us to third party liability insurance, and to the decisions in Post Office v. Norwich Union Fire Insurance Society Ltd. [1967] 2 QB 363 and Bradley v. Eagle Star Insurance Co. Ltd.[1989] AC 957. For a parallel, non-contractual, situation, he referred to Executor Trustee and Agency Company of South Australia Ltd. v. The Deputy Federal Commissioner (South Australia) (1939) 62 CLR 545. In that case the question arose as to the basis on which the Commissioner should tax trustees, whose rights and duties under a will had been determined in the Full Court in proceedings to which the trustees, beneficiaries and representatives of next of kin were party. The High Court of Australia held that the Commissioner was bound to tax on the basis of the Court's determination, not because the Commissioner could be said to be bound on any principle of estoppel by judicial decision, but because he was bound to take the rights of those involved "as they in fact actually exist between the parties", as Latham CJ put it at p.562.
60. We were also referred to the recent decision in Drake Insurance plc v. Provident Insurance plc [2004] 1 LlR 268. Mrs Kaur in that case had two insurances potentially available to cover her in respect of an accident: she had her husband's insurance with Provident taken out in respect of the car she was driving and her own with Drake covering any car she was driving with its owner's consent. Provident obtained an award declaring that it was entitled to avoid her husband's insurance. So she looked exclusively to Drake, which indemnified her in full and sought 50% contribution from Provident on the ground that Provident had not in reality been entitled to avoid. Among the points taken by Provident was that, if this was so, then Drake had paid as to 50% as a volunteer, because of a rateable proportion clause in Drake's policy. The Court of Appeal held that Provident had not in reality been entitled to avoid, but rejected the suggestion that Drake had paid as a volunteer. Rix LJ spoke of the award between Mrs Kaur and the Provident as "a fact in the world, binding as between her and Provident, to which, had Drake sought to enforce the [rateable proportion] clause against her, she would have been in a position to point on the issue whether there was another existing insurance". In other words, in the context of a suggestion that Drake had acted as a volunteer in paying Mrs Kaur 100% of her claim under her policy, the award was a powerful factual tool in Mrs Kaur's hands, both because of the legal arguments which it made possible (whatever their ultimate merit) and because of the business pressure under which it put Drake, not to act "contrary to business ethics and the good name of the industry" by seeking to enforce the clause. Pill LJ agreed with Rix LJ on this part of the case.
61. Clarke LJ put the matter somewhat differently. He thought that Mrs Kaur could in law have defeated any reliance on the rateable proportion clause:

"153. In such postulated proceedings between Mrs. Kaur and Drake, Drake would have said that there was existing insurance covering the same liability between her and Provident, whereas Mrs. Kaur would have replied that there was no such existing insurance because Provident had avoided the policy and had been held to be entitled to do so. As I see it that defence would have succeeded. It would have been no defence for Drake to say that, as between it and Provident or indeed, as between it and Mrs. Kaur, it was entitled to say that the policy had not been validly avoided because, in my opinion (as already stated), the question for decision was whether there was existing cover as between Mrs. Kaur and Provident and Provident had persuaded an arbitrator that there was not."

These sentences accept that it would have been open to Drake to show as against both Provident and Mrs Kaur that the Provident award was wrong. But they proceed, as I see it, on the basis that this would have been irrelevant, because, for the purposes of the rateable proportion clause, what governed was the position as it actually existed and had been determined as between Mrs Kaur and Provident. If this is right, then Clarke LJ's reasoning parallels that applicable under third party indemnity insurance or adopted by the High Court in Executor Trustee, as described in paragraph 59 above.

62. In the present case, there is nothing in the Lincoln reinsurance constituting an agreement that the operation and application of the Net Retained Lines clause should depend upon or follow any arbitral award or judgment as between Sun/Phoenix and any other potential reinsurer about the scope of any such other reinsurance. The only qualification that may have to be made to this proposition is that indicated in paragraph 51 above. If there had been an award between Sun/Phoenix and Cigna determining simply that losses on the Unicover book were recoverable under the Cigna reinsurances, their recoverability could be regarded as a relevant fact, which would mean that they were not being "retained" by Sun/Phoenix within the meaning of the Net Retained Lines clause.
63. The cases considered in paragraphs 53-58 above all concern the situation of a party to the judgment or award seeking to rely upon it as against a stranger. In the present case, Lincoln, the stranger, is seeking to rely on the Cigna award as against a party to it. Such attraction as there is in Mr Hunter's suggested principle derives from this distinction. The judge was clearly influenced by the feeling that Sun-Phoenix had had a full opportunity to argue the issue and lost, and that justice is not generally served by permitting re-litigation of previously decided issues. Such considerations underlie conventional principles of res judicata and issue estoppel. Even where those conventional principles do not apply, such considerations may have direct force in court, in the context of arguments based on abuse of process: cf Hunter v. Chief Constable [1982] AC 529 and Secretary of State for Trade and Industry v. Bairstow (above). In the latter case the Court of Appeal analysed the previous case-law in this area and concluded that

"it will only be an abuse of process of the court to challenge the factual findings and conclusions of the judge or jury in the earlier action if (i) it would be manifestly unfair to a party to the later proceedings that the same issues should be relitigated or (ii) to permit such relitigation would bring the administration of justice into disrepute" (cf per Sir Andrew Morritt V-C at para. 38).

But Mr Hunter did not, in the present private and arbitral context, make any submission to the effect that the Lincoln tribunal either had or exercised any similar jurisdiction to refuse to entertain any case by Sun/Phoenix contrary to the Cigna award on grounds of "abuse of process". I would add that, even if one were to assume that an arbitration tribunal could ever have any jurisdiction of this nature, its exercise on the facts of this case would in my view be excluded by the third, fourth and fifth considerations identified in paragraphs 66-68 below. The new principle which Mr Hunter seeks to develop from Saville J's dicta must, therefore, seek some foundation in legal principle other than the simple considerations of abuse of process which may apply in relation to the administration of justice in court.

64. However, as I see it, there is no foundation in legal principle for Mr Hunter's suggested new principle. First, as I have observed, Saville J's dicta are open to criticism for failing to distinguish between the relevance in relation to third parties of (on the one hand) the main obligations of a contract and (on the other hand) a judgment or arbitration award regarding such obligations.
65. Second, Mr Hunter's attempt to qualify Saville J's dicta so as to make them operate only one-way is contrary to ordinary principle. The principles of res judicata and issue estoppel commonly operate mutually. Saville J's dicta in Moundreas were themselves couched in terms suggesting an extended mutual principle. Spencer Bower puts the position in relation to estoppel arising from a judgment as follows at para. 213:

"Estoppels by res judicata in civil cases "must be mutual" (Petrie v. Nuttall (1856) 11 Exch. 569, 575). …. The established requirement is that res judicata estoppels in civil cases are mutual, and a party is only bound in favour of another if a contrary decision would have bound him. This is important when determining whether a person is a privy."

Spencer Bower points out, however, that Lord Denning disagreed with this, and purported to overrule Petrie v. Nuttall in McIlkenny v. Chief Constable of the West Midlands Police [1980] QB 283, 321, but that George Baker LJ expressed no view on the point while Goff LJ held that mutuality was basic. The House of Lords did not as I see it express any view on this point, although it agreed with Goff LJ on the different issue of abuse of process: see Hunter v. Chief Constable. However, the importance of finality, as a condition for issue estoppel, was emphasised in the speeches in the Houser of Lords in Carl Zeiss Stiftung: see pp. 909-910, 928, 935, 942-3 and 968-9; and the Court of Appeal's decision in Secretary of State for Trade and Industry v. Bairstow (above) has recently underlined the existence of a clear distinction between (on the one hand) the finality or evidential admissibility of a prior decision as between parties or privies to that decision and (on the other hand) the wider jurisdiction that courts have in litigation to exclude or control collateral challenges to an earlier decision by one party to that decision as an abuse of process.

66. Thirdly, I do not consider that the arguments based on general justice have the force in the present context which Mr Hunter suggests and which Toulson J accepted. I do not think that it is obviously just, or even convenient, to allow a stranger to enjoy a one-sided entitlement to hold a party to the award or judgment to its terms, with a concomitant right to challenge its correctness whenever it appeared favourable to do so. In the context of arbitration, the limitation of appeals to points of law in addition to the need for leave for appeal (cf. Arbitration Act 1996, s.72) suggests a general need for caution, quite apart from the specific difficulty facing any attempt by Sun/Phoenix to appeal on an unpleaded alternative that they had not raised until the last minute (cf paragraph 35 above). Further, although there is no doubt in this case that it suits Lincoln to hold Sun/Phoenix to what it submits was determined by the award, it is not difficult to think of qualified scenarios. It could depend on potentially changing or developing facts whether the implications of an award between A and B were favourable to C at a particular time, or even from time to time. There could also be awards with both favourable and unfavourable aspects (for example a determination that certain risks were reinsured by A with B, but subject only to certain qualifications or monetary limits, with which C might wish to take issue). There would thus appear to be no guaranteed way of knowing when and how far an award would be binding. It would depend upon what suited the stranger to the award at any particular time.
67. Fourthly, and linked with the third point, there is a strong element of fortuity about the one-sided benefit for which Mr Hunter contends. Why should Lincoln gain any benefit from an award to which they were not party, particularly in the present context? Sun/Phoenix could not be said to have gained any benefit against anyone - let alone as against Lincoln - from any conclusion by the Cigna tribunal that, but for the avoidance, the Unicover book was protected. Further, if Sun/Phoenix had realised the hopelessness of their case on avoidance and had conceded avoidance or compromised their claim, without any award ever being issued by the Cigna tribunal, Lincoln would have had to arbitrate the scope of the Cigna reinsurances in relation to the Unicover book with Sun/Phoenix without the benefit of any of the present submissions based on the Cigna award. Toulson J was impressed by the supposed difficulty that Lincoln would face: if Sun/Phoenix were permitted to engage in the volte face of calling Mr Minter to prove the existence of an exclusion against which they had previously argued before the Cigna tribunal, then Lincoln would have to call Mr Butler of SCB to try to prove that there was no exclusion, when his evidence before the Cigna tribunal had been that there was not even a misrepresentation or non-disclosure. But I do not see why that would create any special difficulty, let alone one requiring a departure from basic principles. Each witness would be expected to repeat his previous evidence. The Lincoln tribunal would be likely to be given access to transcripts of their previous evidence. It would have to make up its mind about the proper factual findings to make after hearing their further oral evidence. However unlikely it might be, Mr Butler might even convince the Lincoln tribunal that there was not only no exclusion but also no misrepresentation or non-disclosure - which would be to Lincoln's advantage.
68. Fifthly, and more fundamentally, the solution for which Mr Hunter contends appears to me to overlook or obscure important differences between arbitration and litigation. In the context of litigation, problems of potentially conflicting judgments arrived at between different parties to the same overall complex of disputes are met by provisions for joinder of parties or proceedings or for trial together, if necessary on a mandatory basis using the courts' compulsive powers. Even in circumstances in which there has been no such joinder, and where neither res judicata nor issue estoppel has any application, the court may intervene to prevent abuse of its process, as stated in paragraphs 63 and 65 above. All this is facilitated by the public nature of litigation, the public interest in the efficient administration of justice and the courts' coercive powers. Considerations of general justice of the sort to which Toulson J referred thus have relevance and can be given effect in the context of litigation. Arbitration is in contrast a consensual, private affair between the particular parties to a particular arbitration agreement. The resulting inability to enforce the solutions of joinder of parties or proceedings in arbitration, or to try connected arbitrations together other than by consent, is well-recognised - though the popularity of arbitration may indicate that this inability is not often inconvenient or that perceived advantages of arbitration, including confidentiality and privacy are seen as outweighing any inconvenience. Different arbitrations on closely inter-linked issues may as a result lead to different results, even where, as in the present case, the evidence before one tribunal is very largely the same as that before the other. The arbitrators in each arbitration are appointed to decide the disputes in that arbitration between the particular parties to that arbitration. The privacy and confidentiality attaching to arbitration underline this; and, even if they do not lead to non- parties remaining ignorant of an earlier arbitration award, they are calculated to lead to difficulties in obtaining access, and about the scope of any access, to material relating to that award.
69. The conclusion that I would reach is that Mr Hunter's suggested principle has no sound basis, and that the dicta of Saville J in Moudreas cannot be regarded as reflecting or as based on any general principle of law in the arbitral context to which they were directed.

CONCLUSION

70. I would therefore allow this appeal and restore paragraph (4) of the dispositive part of the Lincoln award to the form in which it was made by the Lincoln tribunal.

Lord Justice Longmore:

71. I entirely agree with the judgment of Mance LJ on the first three issues to which the appeal gives rise. That means that the ratio of our decision is contained in his judgment on issue (3) to the effect that the Cigna arbitrators' conclusion (that the Unicover book was, apart from Cigna's successful avoidance, reinsured by Cigna) was not a necessary part of their decision, was collateral to it and thus not sufficient to constitute an issue estoppel in subsequent litigation. I add a few words on issue (4) not because I disagree in any way with Mance LJ's reasoning or conclusion but because we are differing from a considered dictum of Saville J and the judgment of Toulson J in the present case.
72. In the Moundreas case [1985] 2 Lloyds Rep. 515 the claim was made by a shipbroker for his commission against the builders/sellers of the ship. The commission was payable on the happening of various stages of the shipbuilding contract which had not occurred because, as the arbitrators decided, the buyers were entitled to terminate the shipbuilding contract by reason of breach of contract on the part of the builders/sellers. The broker contended that it was a breach of the agency contract for the builders/sellers to deprive him of the opportunity of earning his commission at the relevant stages of the shipbuilding contract. He had, therefore, to prove that the builders/sellers had so deprived him which they would have done if (as the arbitrators had held) they were guilty of a repudiatory breach which the buyers were entitled to (and did) accept as terminating the contract. It was in these circumstances that Saville J said (page 520):-

"As between the parties that is now the contractual position as determined by the contractual machinery of arbitration - and it is difficult to see why a stranger to the contract cannot prove that contractual position by simply producing the award as he can prove other contractual rights and obligations by simply producing the contract."

73. There are, as it seems to me, four difficulties about this passage, which was not essential to his decision since the shipbroker in fact proved, by satisfactory evidence, that the builders/sellers were in repudiatory breach. The first (and purely formal) difficulty is that a stranger to a contract with an arbitration clause will often be in considerable difficulty in "simply producing the award". Arbitrations are usually private and awards can normally be referred to in other proceedings only if the parties to the award consent that it should be so referred to. I do not, however, wish to engage with possible exceptions to that general principle since Mr Ian Hunter QC for Lincoln asserted (and Mr Sumption QC for Sun/Phoenix appeared to accept) that Saville J was not confining his observations to cases of arbitrators' awards. On this basis, Saville J's observations, if correct, would apply equally to a stranger producing a judgment, as to which there is no difficulty because it is a public document.
74. It is, of course, possible that Saville J was intending to confine his remarks to cases of arbitration awards on the basis that it is an implied term of every contract containing an arbitration clause that the parties will honour an award. But even if that were to have been the basis of Saville J's observation, there would still be the difficulty (more acute in 1985 than now) that a third party cannot normally rely upon or otherwise take advantage of a term in a contract to which he is not a party. No argument was addressed to us on the Contracts (Rights of Third Parties) Act 1999 and I leave that matter there.
75. The second difficulty, on the assumption that the observations were intended to be general in character, is that while it makes sense to say, in an appropriate case, that a stranger to a contract can prove the existence of "contractual rights and obligation by simply producing the contract" that is not sufficient for the purpose of the stranger. He needs to prove not the existence of "contractual rights and obligations" but that the person whom he is suing is in breach of the relevant obligations. He cannot do that merely by producing an award or a judgment, unless he has a contract with the person whom he is suing to the effect that he agrees that that question can be regarded as determined by a prior award or judgment. This has long been held to be the law in cases of contracts of guarantee, see Ex parte Young, Re Kitchin (1881) 17 ChD 668 (prior judgment) and The Vasso [1979] 2 Lloyds Rep. 412 (prior award). It has also been held to be the case for reinsurance, see Hayter v Nelson [1990] 2 Lloyds Rep. 265. If the pre-1985 authorities had been cited to Saville J in Moundreas I doubt if he would have expressed himself in such general terms.
76. It can, of course, be said that these two pre-1985 authorities are cases where a claimant who was successful in previous proceedings is seeking to rely on that success to preclude a stranger to those proceedings from putting forward the defence that the original defendant was not liable in a second set of proceedings to which the erstwhile stranger is now the defendant. That would obviously be unjust. Mr Hunter argued that it was different where it was the now defendant who had been party to the previous proceedings; then, he said, it would be eminently just and reasonable that the now defendant, who has raised in previous proceedings the very same issue as arises in the current proceedings and has had that issue decided against him, should not be allowed to re-argue the very same matter in the new proceedings. He submitted that Moundreas was an example of just such a case and should be followed. This submission gives rise to the third and fourth difficulties with the observations of Saville J which is that they seem to me to be contrary to both principle and authority.
77. The long-established principle is that estoppels are mutual. The principle for which Mr Hunter contends is curiously one-sided in that it is the now defendant who is bound by a decision against him but the now plaintiff would not be bound by a decision in the now defendant's favour. That is contrary to principle, see for example Spencer Bower Res Judicata 3rd ed para. 213. It is because of this requirement of mutuality that the further requirement of issue estoppel has evolved, that the estoppel is only effective as between the same parties or their privies. But if it would be unfair to bind someone who is not a party, the principle of mutuality requires that someone not a party cannot take advantage of a decision in proceedings made when he was not there.
78. Fourthly there is the matter of authority. It is true to say that the majority of the Court of Appeal in the Birmingham bombing case McIlkenny v Chief Constable of the West Midlands [1980] QB. 283 supported the reasoning for which Mr Hunter was contending. In that case, the alleged bombers (they have now had their convictions quashed) had argued at trial that they had been assaulted by the police and that their confessions were not, therefore, admissible in evidence. On a voir dire Bridge J had stated that he was satisfied beyond reasonable doubt that the police had not assaulted the defendants; the result was that their confessions were held admissible. The same evidence was then deployed before the jury who nevertheless convicted the defendants and must, on any view of the matter, have likewise been convinced beyond reasonable doubt that they had not been assaulted by the police. The alleged bombers then sued the police for assault; both the Court of Appeal and the House of Lords held that it was an abuse of process for them to claim damages in civil proceedings for assaults which a criminal court had concluded beyond reasonable doubt had not taken place. In the Court of Appeal, however, both Lord Denning MR and Sir George Baker concluded that the claimants were prevented by issue estoppel from raising the same points as had been concluded against them in the criminal proceedings despite the fact that the defendant chief constable in the civil proceedings was not the same party as the Crown in the criminal proceedings. They relied on the advice of the judges to the House of Lords in the Duchess of Kingston's case (1776) 2 Smith LC (13th Edn) 644, 647-8:-

"But in all these cases, the parties to the suits, or at least the parties against whom the evidence was received, were parties to the sentence, and had acquiesced under it."

From these words, the principle was derived that a party, against whom an issue had been decided, could not raise the self-same issue in subsequent proceedings.

79. As Goff LJ pointed out, this was contrary to more recent authority, particularly the Court of Appeal case of Mills v Cooper [1967] 2 QB 459 which had decided that the assertion which a party to new proceedings was in law estopped from asserting had to be the same assertion as was made

"in his previous cause of action or defence in previous civil proceedings between the same parties or their predecessors in title."

80. When the Birmingham case went to the House of Lords, sub nom Hunter v Chief Constable of the West Midlands Police [1982] AC 529, the House decided that the institution of the civil proceedings was indeed an abuse of process of the court and that

"the question whether it also qualifies to bear the label "issue estoppel" is a matter not of substance but of semantics".

The House declined to engage in that matter of semantics but Lord Diplock said at page 540:-

"Nevertheless it is my own view, which I understand is shared by all your Lordships, that it would be best, in order to avoid confusion, if the use of the description "issue estoppel" in English law, at any rate . . . . were restricted to that species of estoppel per rem judicatam that may arise in civil actions between the same parties or their privies, of which the characteristics are stated in a judgment of my own in Mills v Cooper [1967] 2 QB 459 that was adopted and approved by this House in DPP v. Humphrys [1977] AC 1 . . . . "

81. In these circumstances it seems to me too clear for argument that in this court we remain bound by Mills v Cooper and cannot be seduced along the path proposed by Mr Hunter and adopted by the majority of the Court of Appeal in the Birmingham bombing case. Toulson J was not told that this court had come to the same conclusion in Secretary of State for Trade and Industry v Bairstow [2004] Ch 1 by which, in my view, we are now bound.
82. I would add that even if it were a principle of English law that a party against whom an issue had been decided in earlier proceedings could not reopen that issue in subsequent proceedings, it could only doubtfully apply in the present case. The earlier proceedings decided the issue of cover in favour of Sun/Phoenix; it can only be said to have been decided against them in the sense that for the purpose of the subsequent proceedings Sun/Phoenix would prefer that the decision had gone the other way. But I doubt if the principle on which Mr Hunter relied (if it exists) was intended to cover such a case.
83. The sad truth is that in the absence of any third-party or consolidation procedure in arbitration, parties may be put into the position of making inconsistent cases in different proceedings. In litigation it is possible to make inconsistent cases in the same proceedings; doing so later, in different proceedings, may come under the head of abuse of process. But that is no reason to extend the law of issue estoppel in arbitration proceedings beyond its proper sphere.
84. All the above is not to deny that there may be cases in which an award can be evidence in subsequent proceedings even though it will not necessarily be conclusive evidence. It may, to use Rix LJ's expression in Drake Insurance Plc v Provident Insurance Plc [2003] EWCA Civ 1834, be a "fact in the world". A good example of this is to be found in The Sargasso [1994] 1 Lloyds Rep. 412 where a charterer had been held liable by an award in favour of a sub-charterer who had sued to recover damages for damage to cargo. The charterer then sued the shipowner and proved breach of contract; the measure of damages to which he was entitled was governed by the award pursuant to which he had been held liable to the sub-charterer. It quantified the loss which he had actually suffered; he was entitled to put it in evidence for that purpose and say he should be able to recover not less than the amount of the award; the shipowner would also be entitled to say that the charterer should not recover more than the amount of award. That would not have prevented the shipowner from arguing that the charterer had not taken the right points and that he had thus failed to mitigate his damages or, indeed, that the award against him had been made by reason of some fact which was not a breach of contract on the owners' part.
85. In The Sargasso Clarke J referred to the observations of Saville J in the Moundreas case and in that context they are uncontroversial. But I do not consider that they form a safe basis on which to found any extension of the existing law of issue estoppel, and they should not be followed in future for that purpose.

Lord Justice Jacob:

86. I agree with both judgments. It is worth standing back from the detail. What Lincoln seek to do is to rely upon a non-operative (in the sense that no actual consequences flow from it), opinion expressed by the Cigna arbitrators. The opinion is in its nature private. Moreover it was unappealable. Lincoln seek more than just to rely upon the opinion –they say it is conclusive for all purposes and so conclusive in the later arbitration.
87. I think such a result would be obviously wrong for the following reasons:

(a) An arbitration is an essentially private matter between the parties to it. Only some consequence of an award (e.g. that A should pay B money) can go further and extend beyond the privacy of the arbitration itself – so as to become a "fact in the world." (Rix LJ's phrase).

(b) Because the determination of arbitrators is itself a private matter it is in its nature not intended to be available to third parties for any purpose. A third party's rights against one of the parties to an earlier arbitration cannot depend on the happenstance of the availability of the details of that arbitration in a later arbitration involving that third party. In this connection I note that the position may be different if the earlier decision is that of a court. In particular a decision of a court as to the construction of a contract is a matter of law – with the consequence that the further principle of judicial precedent on such a question may come into play.

88. Where a party seeks to re-litigate in subsequent proceedings against Y a point he fought fully in earlier proceedings against X, it may be that, notwithstanding a lack of mutuality, he can be prevented from doing so on the grounds of abuse of process. As to that I express no concluded opinion for, for the reasons given by Mance LJ, there is no question of abuse of process here.

ORDER: Appeal allowed, respondents to pay appellant's costs of the appeal and below; Application to appeal to the House of Lords refused.

(Order does not form part of approved Judgment)