Note: This file does not form part of the LCD or CS sites. The files on those sites are the only official versions of the CPRs and related material. Please also note that the cross-references are not claimed to be comprehensive.

(In this YAWS version a few editorial alterations have been made to the description of the CPR and Practice Directions, eg references to Part 3.8 have been changed to Rule 3.8, to ensure consistency with the remainder of the site.)

Click here to reload this page into top frame


This Report is referred to in: 03a1368 [97], [97], Aldi Stores Ltd v WSP Group Plc [16], Collier v Williams [91], [94], Garratt v Saxby [23], Sinclair Roche & Temperley v Simatra [26], Stuart v Goldberg [81].
Neutral Citation Number: [2002] EWCA Civ 1642
Case No: A3/2001/2450

IN THE SUPREME COURT OF JUDICATURE
COURT OF APPEAL (CIVIL DIVISION)
ON APPEAL FROM THE HIGH COURT OF JUSTICE
COMMERCIAL COURT
The Hon Mr Justice Morison

Royal Courts of Justice
Strand, London, WC2 A 2LL

13 November 2002

Before:

LORD JUSTICE WARD
LORD JUSTICE CLARKE
and
SIR CHRISTOPHER STAUGHTON

 

Between:

ASSICURAZIONI GENERALI SpA
Claimant/ Respondent

- and -

ARAB INSURANCE GROUP (B.S.C.)
Defendant/ Appellant

 

Mr Julian Flaux QC and Mr Simon Picken
(instructed by Holman, Fenwick and Willan) for the Appellant
Mr Stewart Boyd QC and Mr Richard Millett
(instructed by LeBoeuf, Lamb, Greene and MacRae) for the Respondent
Hearing dates : 18th to 24th June 2002

HTML VERSION OF JUDGMENT : APPROVED BY THE COURT FOR HANDING DOWN (SUBJECT TO EDITORIAL CORRECTIONS)
Crown Copyright ©
   Lord Justice Clarke:
   Introduction
1.   This is an appeal by the defendant retrocessionaires ("ARIG") from a judgment or order of Morison J, which was finally made on 9 November 2001, in which he gave judgment for reinsurers ("Generali") for US$8,751,000 plus interest of US$1,450,000 after the trial of an action in which he held ARIG liable under the terms of two quota share retrocession contracts ("the contracts"). He also ordered ARIG to pay Generali’s costs, part on the standard and part on the indemnity basis, and refused permission to appeal. Tuckey LJ subsequently granted permission to appeal on the merits, but refused permission to appeal on indemnity costs.
   Background
2.   Generali is and was an Italian insurer and reinsurer with a branch office in London. ARIG is and was an insurer and reinsurer based in Bahrain with an established place of business in London. The judge held that the contract between the parties was made on 9 September 1992. It related to the year from 2 July 1992 to 2 July 1993, but in March or April 1993 was extended for two months until 2 September 1993.
3.   The background to the contract was as follows. In about 1989 Alexander Howden and others developed packages of insurance risks, which essentially concerned builders’ risks in relation to small to medium sized construction projects in the USA, notably in California. Initially Generali acted as direct insurer of the contractors’ risks. It was, however, decided that it was desirable, if not essential, to insert a US based insurance company between the insured and Generali. As a result, an insurance company called United National Insurance Group of Pennsylvania ("United") entered into the direct insurance contracts with the insured as a front for Generali. United thus reinsured the risks with Generali, but subject to such reinsurance, or strictly retrocession, as it sold on the market. In the event Generali retained part of the risk but retroceded part of the risk to others, including ARIG which took 7.5 per cent.
4.   As effectively the direct insurer or lead reinsurer, Generali had to have a system for investigating and evaluating the claims which it received from the insured contractors. To that end it had previously entered into a contract dated 1 July 1991 with Gay & Taylor Inc ("G&T"), which was a company of loss adjusters, to evaluate and process the claims. That contract was still in force at the relevant time. It included an obligation upon G&T to establish a claim file and "to investigate all claims and to recommend the amount of loss Reserve to be established with respect to such claim".
5.   Generali paid claims under the insurances and made claims against ARIG under the contract. ARIG continued to accept premium and to pay claims until 7 February 1999 but thereafter refused to pay any more claims. As a result Generali commenced this action against ARIG, which relied upon a number of defences. All the defences failed at the trial. By way of appeal ARIG raises some of those defences again.
   Approach of the Court of Appeal
6.   CPR rule 52.11 provides, so far as relevant, as follows:
  
"(1)   Every appeal will be limited to a review of the decision of the lower court unless –
  
 (a)   a practice direction makes different provision for a particular category of appeal; or
 (b)    the court considers that in the circumstances of an individual appeal it would be in the interests of justice to hold a re-hearing.
 (3)   The appeal court will allow an appeal where the decision of the lower court was
  
 (a)   wrong; …
 (4)   The appeal court may draw any inference of fact which it considers justified on the evidence."
This Paragraph is referred to in: 03a1368 [97].

7.   Mr Boyd correctly observed that ARIG has not suggested that this appeal should be by way of rehearing and not by way of review. He further drew attention to the contrast between CPR rule 52.11(1) and RSC Order 59 rule 3, which it replaced and which provided that "an appeal to the Court of Appeal shall be by way of rehearing". Mr Boyd submitted that it follows both from the contrast between "review" and "re-hearing" within CPR rule 52.11(1) itself and between "review" in CPR rule 52.11(1) and "rehearing" in RSC Order 59 rule 3 that the approach of the Court of Appeal when conducting a review must be different from its approach when conducting a rehearing, either under CPR rule 52.11(1)(b) or under RSC Order 59 rule 3.
8.   There is plainly force in that submission, although the position is complicated by the fact that "rehearing" does not always have the same meaning. Thus, under the RSC, an appeal to the judge from a master under RSC Order 58 rule 1 proceeded by way of rehearing, but it was a rehearing of a different kind from that conducted by the Court of Appeal under RSC Order 59 rule 3. This can be seen from Tanfern Ltd v Cameron-MacDonald [2000] 1 WLR 1311 per Brooke LJ, giving the judgment of the court, at paragraph 30.
9.   That was a case in which the court gave guidance in relation to a number of aspects of appeals conducted under the CPR. However, as I read it, neither the guidance nor the case was concerned with appeals on the merits after a trial. Brooke LJ said this:
  
   "The appellate approach: the general rule
 30.   As a general rule, every appeal will be limited to a review of the decision of the lower court. This general rule will be applied unless a practice direction makes different provision for a particular category of appeal, or the court considers that in the circumstance of an individual appeal it would be in the interests of justice to hold a rehearing: CPR r 52.11(1). The appeal court will only allow an appeal where the decision of the lower court was wrong, or where it was unjust because of a serious procedural or other irregularity in the proceedings in the lower court: CPR r 52.11(3).
 31.   This marks a significant change in practice, in relation to what used to be called "interlocutory appeals" from district judges or masters. Under the old practice, the appeal to a judge was a rehearing in the fullest sense of the word, and the judge exercised his/her discretion afresh, while giving appropriate weight to the way the lower court had exercised its discretion in the matter. Under the new practice, the decision of the lower court will attract much greater significance. The appeal court’s duty is now limited to a review of that decision, and it may only interfere in the quite limited circumstances set out in CPR r 52.11(3).
 32.   The first ground for interference speaks for itself. The epithet "wrong" is to be applied to the substance of the decision made by the lower court. If the appeal is against the exercise of a discretion by the lower court, the decision of the House of Lords in G v G (Minors: Custody Appeal) [1985] 1 WLR 647 warrants attention. In that case Lord Fraser of Tullybelton said, at p 652:
  
    "Certainly it would not be useful to inquire whether different shades of meaning are intended to be conveyed by words such as ‘blatant error’ used by the President in the present case, and words such as ‘clearly wrong’, ‘plainly wrong’, or simply ‘wrong’ used by other judges in other cases. All these various expressions were used in order to emphasise the point that the appellate court should only interfere when they consider that the judge of first instance has not merely preferred an imperfect solution which is different from an alternative imperfect solution which the Court of Appeal might or would have adopted, but has exceeded the generous ambit within which a reasonable disagreement is possible".
10.   Mr Boyd submitted that the same approach should be adopted to an appeal of this kind. However, for my part I would not accept that submission. As Brooke LJ makes clear, Lord Fraser was considering the correct approach where the appeal is against the exercise of a discretion by the lower court. An appeal of the kind with which we are concerned challenges the judge’s conclusions of fact, not the exercise of a discretion.
11.   Brooke LJ was not considering an appeal of this kind beyond observing that it is by way of review under CPR rule 52.11(1) and that the question is whether the judge was wrong under rule 52.11(3). An appeal to the Court of Appeal under RSC Order 59 rule 3, although expressed to be by way of rehearing, was not a "rehearing in the fullest sense of the word" because (for example) the Court of Appeal never exercised its discretion afresh unless and until it had held the judge to have exercised his discretion wrongly. Moreover it did not hear the oral evidence again. In these circumstances it seems to me that, although expressed to be by way of rehearing, it was very much an appeal by way of review of the decision of the trial judge.
12.   Thus, for example, in cases in which the court was asked to reverse a judge’s findings of fact which depended upon his view of the credibility of the witnesses, it would only do so if satisfied that the judge was plainly wrong. This can be seen from many cases, as for example The Ikarian Reefer [1995] Lloyd's Rep 455, where the court reversed the decision of the trial judge that the plaintiff insured shipowners had not deliberately scuttled their vessel or cast her away. Giving the judgment of the court, Stuart-Smith LJ addressed the correct approach as follows (at pp 458-9):
  
"(1)   The burden of showing that the trial Judge was wrong lies on the appellant. …
 (2)   When questions of the credibility of witnesses who have given oral evidence arise the appellant must establish that the trial Judge was plainly wrong. Once again there is a long line of authority emphasizing the restricted nature of the Court of Appeal’s power to interfere with a Judge’s decision in these circumstances though in describing that power different expressions have been used. In SS Hontestroom v SS Sagaporak [1927] AC 37 at p 47 Lord Sumner said:
  
    "None the less not to have seen the witnesses puts appellate Judges in a permanent position of disadvantage as against the trial Judge and unless it can be shown that he has failed to use or has palpably misused his advantage, the higher Court ought not to take the responsibility of reversing conclusions so arrived at merely on the results of their own comparisons and criticisms of the witnesses and of their own view of the probabilities of the case."
    ….
    Finally in Mersey Docks & Harbour Board v Proctor [1923] AC 253 at p 258, Viscount Cave LC said:
  
    "In such a case … it is the duty of the Court of Appeal to make up its own mind not disregarding the judgment appealed from and giving special weight to that judgment in cases where the credibility of witnesses comes into question, but with full liability to draw its own inferences from the facts proved or admitted and to decide accordingly.
 (3)   When a party has been acquitted of fraud the decision in his favour should not be displaced except on the clearest grounds. This proposition is not in contest and is supported by the House of Lords in Akerhielm v De Mare [1959] AC 789 at p 806, where the earlier authority of Glasier v Rolb (1889) 42 Ch D 436 is cited."
13.   In my view, although it is true that RSC Order 59 rule 3 expressly refers to a rehearing, the exercise upon which the court was engaged was essentially one of review. That is I think so, even though the extract from the speech of Viscount Cave in Mersey Docks & Harbour Board v Proctor might suggest that that was not so because of the reference to the court having "full liability to draw its own inferences from the facts proved or admitted". However, I observe that CPR rule 52.11.1(4) expressly gives the appeal court (which of course includes this court) power to draw any inference of fact which it considers justified on the evidence. There is no suggestion that that rule applies only to appeals by way of re-hearing under rule 52.11(1)(b), so that the court has that power when conducting a review. In these circumstances, it seems to me that in the type of appeal in which the court is asked to reverse findings of fact based upon the credibility of the witnesses, the same approach should be adopted in this court whether the appeal is by way of review or re-hearing.
14.   The approach of the court to any particular case will depend upon the nature of the issues kind of case determined by the judge. This has been recognised recently in, for example, Todd v Adam [2002] EWCA Civ 509, [2002] 2 All ER (Comm) 1 and Bessant v South Cone Inc [2002] EWCA Civ 763(Bailii). In some cases the trial judge will have reached conclusions of primary fact based almost entirely upon the view which he formed of the oral evidence of the witnesses. In most cases, however, the position is more complex. In many such cases the judge will have reached his conclusions of primary fact as a result partly of the view he formed of the oral evidence and partly from an analysis of the documents. In other such cases, the judge will have made findings of primary fact based entirely or almost entirely on the documents. Some findings of primary fact will be the result of direct evidence, whereas others will depend upon inference from direct evidence of such facts.
15.   In appeals against conclusions of primary fact the approach of an appellate court will depend upon the weight to be attached to the findings of the judge and that weight will depend upon the extent to which, as the trial judge, the judge has an advantage over the appellate court; the greater that advantage the more reluctant the appellate court should be to interfere. As I see it, that was the approach of the Court of Appeal on a "rehearing" under the RSC and should be its approach on a "review" under the CPR.
16.   Some conclusions of fact are, however, not conclusions of primary fact of the kind to which I have just referred. They involve an assessment of a number of different factors which have to be weighed against each other. This is sometimes called an evaluation of the facts and is often a matter of degree upon which different judges can legitimately differ. Such cases may be closely analogous to the exercise of a discretion and, in my opinion, appellate courts should approach them in a similar way.
This Paragraph is referred to in: Collier v Williams [91].

17.   In Todd v Adam , where the question was whether a contract of service existed, Mance LJ drew a distinction between challenges to conclusions of primary fact or inferences from those facts and an evaluation of those facts, as follows, at paragraph 129:
  
   "With regard to an appeal to this court (which would never have involved a complete rehearing in that sense), the language of ‘review’ may be said to fit most easily into the context of an appeal against the exercise of a discretion, or an appeal where the court of appeal is essentially concerned with the correctness of an exercise of evaluation or judgment – such as a decision by a lower court whether, weighing all relevant factors, a contract of service existed. However, the references in r 52 11 (3) and (4) to the power of an appellant court to allow an appeal where the decision below was ‘wrong’ and to ‘draw any inference of fact which it considers justified on the evidence’ indicate that there are other contexts in which the court of appeal must, as previously, make up its own mind as to the correctness or otherwise of a decision, even on matters of fact, by a lower court. Where the correctness of a finding of primary fact or of inference is in issue, it cannot be a matter of simple discretion how an appellant court approaches the matter. Once the appellant has shown a real prospect (justifying permission to appeal) that a finding or inference is wrong, the role of an appellate court is to determine whether or not this is so, giving full weight of course to the advantages enjoyed by any judge of first instance who has heard oral evidence. In the present case, therefore, I consider that (a) it is for us if necessary to make up our own mind about the correctness or otherwise of any findings of primary fact or inferences from primary fact that the judge made or drew and the claimants challenge, while (b) reminding ourselves that, so far as the appeal raises issues of judgment on unchallenged primary findings and inferences, this court ought not to interfere unless it is satisfied that the judge’s conclusion lay outside the bounds within which reasonable disagreement is possible. In relation to (a) we must, as stated, bear in mind the important and well-recognised reluctance of this court to interfere with a trail judge on any finding of primary fact based on the credibility or reliability of oral evidence. In the present case, however, while there was oral evidence, its content was largely uncontentious."
   In the same case Neuberger J stressed (at paragraphs 61 to 64) that the question whether there was a contract of service on the facts involved the weighing up of a series of factors. Thorpe LJ agreed with both judgments.
18.   Another example of a case which involved the assessment of a number of factors was Bessant v South Cone , where a key question was whether a trade mark should not be registered because it was to be registered for goods which were identical with or similar to those for which an earlier trade mark was protected within the meaning of section 5(2)(b) of the Trade Marks Act 1994Acts. Robert Walker LJ, with whom Buxton LJ and I agreed, referred to a statement of Buxton LJ in Norowzian v Arks Ltd (No 2) [2000] FSR 363 at p 370, which was approved by at least three members of the House of Lords in Designers Guild Ltd v Russell Williams (Textiles) Ltd [2000] 1 WLR 2416, to Re Grayan Building Services Ltd [1995] Ch 241, per Hoffmann LJ at p 254, and to Pro Sieben Media AG v Carlton United Kingdom Television Ltd [1999] 1 WLR 605, 612, which included the following:
  
   "On the other hand, the standards applied by the law in different contexts vary a great deal in precision and generally speaking, the vaguer the standard and the greater the number of factors which the court has to weigh up in deciding whether the standards have been met, the more reluctant an appellate court will be to interfere with the trial judge’s decision"
19.   In paragraph 20 Robert Walker LJ also quoted this passage from the speech of Lord Hoffmann in Biogen Inc v Medeva Plc [1997] RPC 1 at p 45:
  
   "The question of whether an invention was obvious had been called "a kind of jury question" (see Jenkins LJ in Allmanna Svenska Elektriska A/B v The Burntisland Shipbuilding Co Ltd (1952) 69 RPC 63, 70) and should be treated with appropriate respect by an appellate court. It is true that in Benmax v Austin Motor Co Ltd [1955] AC 370 (1955) 72 RPC 39, 42, this House decided that, while the judge’s findings of primary fact, particularly if founded upon an assessment of the credibility of witnesses, were virtually unassailable, an appellate court would be more ready to differ from the judge’s evaluation of those facts by reference to some legal standard such as negligence or obviousness. In drawing this distinction, however, Viscount Simonds went on to observe, at page 374, that it was "subject only to the weight which should, as a matter of course, be given to the opinion of the learned judge". The need for appellate caution in reversing the judge’s evaluation of the facts is based upon much more solid grounds than professional courtesy. It is because specific findings of fact, even by the most meticulous judge, are inherently an incomplete statement of the impression which was made upon him by the primary evidence. His expressed findings are always surrounded by a penumbra of imprecision as to emphasis, relative weight, minor qualification and nuance (as Renan said, la vérité est dans une nuance), of which time and language do not permit exact expression, but which may play an important part in the judge's overall evaluation. It would in my view be wrong to treat Benmax as authorising or requiring an appellate court to undertake a de novo evaluation of the facts in all cases in which no question of the credibility of witnesses is involved. Where the application of a legal standard such as negligence or obviousness involves no question of principle but is simply a matter of degree, an appellate court should be very cautious in differing from the judge’s evaluation."
20.   Robert Walker LJ also considered the overlap between the classes of case, with particular reference to what is meant by inferences of fact. He said (at paragraph 24):
  
   "It is worth reflecting on what judges mean when they speak of ‘inferences’ in this context. An inference from a number of primary facts may itself be a simple matter of fact. That is an inference from circumstantial evidence, or what might be called the ‘smoking gun’ kind of inference. (Inferences from a litigant’s failure to call a particular witness may be in this category.) In the present context, however, the inference is not a simple matter of fact because it involves a process of evaluation. It was put very clearly by Viscount Simonds in Benmax v Austin Motor Co Ltd [1955] AC 370, 373 (a patent case on obviousness):
  
    "I cannot help thinking that some confusion may have arisen from a failure to distinguish between a finding of a specific fact and a finding of fact which is really an inference from facts specifically found or, as it has sometimes been said, between the perception and evaluation of facts."
21.   Then, after a reference to Edwards v Bairstow [1956] AC 29., Robert Walker LJ said (in paragraph 26):
  
   "How reluctant should an appellate court be to interfere with the trial judge’s evaluation of, and conclusion on, the primary facts? As Hoffmann LJ made clear in Grayan there is no single standard which is appropriate to every case. The most important variables include the nature of the evaluation required, the standing and experience of the fact-finding judge or tribunal, and the extent to which the judge or tribunal had to assess the oral evidence."
22.   Thus the extent to which the findings of fact depend upon oral evidence or what Lord Hoffmann called the "penumbra of imprecision as to emphasis, relative weight, minor qualification and nuance" will vary from case to case. In the instant case, the judge had the considerable advantage of seeing the witnesses and of assessing their credibility, although, as ever, he did so against the documentary material that was available. In these circumstances we should, I think, take particular care before holding that his conclusions of fact were wrong, especially since (as appears below) some of his conclusions depended to a significant extent upon the view which he formed of the witnesses. On the other hand this is not a case in which the judge was concerned to weigh a number of factors such that the judgment which he was called upon to make was a matter of degree.
23.   Finally, I would add that it seems to me that the approach of this court in this kind of case, where it is not suggested that we could rehear the evidence, is or should be the same whether it is conducting a "review" or "re-hearing" as those expressions are used in CPR rule 52.11. Where the court is concerned with the exercise of a discretion, the difference will be of considerable importance (see eg Audergon v La Baguette Ltd [2002 EWCA Civ 10 and Asiansky Television Plc v Bayer-Rosin [2001] EWCA Civ 1792), but in a case in which the appeal is after a trial on the facts where the judge’s findings of fact are challenged on appeal, this is an example of the kind of case which, as Jonathan Parker LJ (with whom Pill and Tuckey LJJ agreed) said in Audergon (at paragraph 85), a decision by the appeal court whether to hold a rehearing may make little practical difference.
   The appeal
24.   Before the judge ARIG said that it was entitled to avoid the contract for both non-disclosure and misrepresentation on a number of grounds. It also asserted a breach of warranty. It no longer asserts any breach of warranty. Nor does it rely upon all the grounds upon which it previously alleged non-disclosure or misrepresentation. It does, however rely upon one alleged non-disclosure and two alleged misrepresentations as follows:
  
 i)   failure by Generali to disclose that the reserving policy adopted by G&T was imprudent or unusual;
 ii)   presentation by Generali of materially inaccurate loss statistics; and
 iii)   misleading ARIG as to the participation of Munich Re by stating or implying that it was participating in the whole package and not (as was the case) in only one section.
25.   The judge treated points i) and ii) as effectively raising only one point. He made findings of fact as to what the reserving policy was, although he said that the methodology used was never really in dispute. He rejected ARIG’s case that the policy was unusual or imprudent. He therefore rejected point i) and, with regard to point ii), held that ARIG had not shown that the figures with which it was presented were materially inaccurate. He further said that he was not satisfied on the balance of probabilities that, had the files been reserved as ARIG’s reserving expert Mr Beckerman said they should have been, it would have made any difference to ARIG’s decision to participate, either initially or for the additional two months in 1993.
26.   As to point iii), which relates to the participation of Munich Re, he held that there was no relevant misrepresentation and, in any event, he did not consider that any such representation as there was played any part in ARIG’s decision to take part in the programme.
27.   It is convenient to consider points i) and ii) under the heading of reserving policy and point iii) under the heading of Munich Re but, before doing so, I should identify the terms of the relevant insurance arrangements and the role of the participants in them. I shall at the same time consider the date upon which the contract was made which is relevant to the Munich Re point.
   The insurance, reinsurance and retrocession
28.   As stated earlier, in about 1989 Alexander Howden and others developed packages of insurance risks, which essentially concerned builders’ risks in relation to small to medium sized construction projects in the USA, notably in California. There were two facilities, nos 0956 and 0957. No 0956 was known as the WK Cooper "Con-Pac" and Allied Trades Package and no 0957 was known as the Alexander Howden Contractors and Allied Trades Package. The business was first written in 1990 and Alexander Howden initially acted as claims administrator. However, Generali subsequently entered into the contract with G&T dated 1 July 1991 to which I referred earlier. Thereafter it was G&T’s responsibility to administer the claims on behalf of Generali. I am not quite sure what, if any, participation others had in the 1990 and 1991 years, but the position from the beginning of July 1992 was essentially as follows.
29.   On 18 June 1992 Alexander Howden sent ARIG a fax inviting participation in the two facilities. The fax described the placings in general terms. It included the following:
  
   "As you will see these placings are for Contractors involved principally in Residential, Commercial and associated trades coverage is provided on an all jobs basis with no single project being more than $10,000,000 in Contract.
    The estimated premium income over both placings in the next twelve months is estimated at approximately US$20,000,000.
    As respects the Liability coverage, please note that any account with a known products exposure is automatically disqualified and not even submitted to the Underwriters for an indication of terms.
    Each account is subject to a vigorous pre-qualification procedure in the USA and in London before it is even submitted to Underwriter.
    This process has contributed to the success of the placing which is demonstrated by the enclosed figures and the premium volume.
    The leading Insurers in London are the Assicurazioni Generali Spa with support from the Commercial Union and the Munich Re.
    We believe you will find our submissions self explanatory but if you have any questions pleas contact the undersigned.
    We now await details of your proposed participation which can be on a reinsurance basis if you wish."
30.   Attached to the fax were a series of documents which gave details of the proposed business. Some of the documents are of some significance to the issues in this appeal. One such document made a number of statements with regard to the business under the heading "Introduction 1992". It included the following:
  
   This placing has a proven record in attracting contracting risks.
    …. one of the reasons for its success has been the low handling costs for both Producers and Underwriters.
    In order that risks may be handled swiftly and economically by both Leading Underwriters and Brokers, business will be handled in the following manner:
  
 1)   Summary sheets will be completed by the Agent, providing details of risk.
 2)   The Leading Reinsurer will indicate terms for the risk and these will be submitted by Alexander Howden to their Agent.
 3)   If a firm order is obtained, the Underwriter and their Leading Reinsurer will simultaneously bind the risk.
 4)   The Agent of Alexander Howden will then issue a Certificate of Insurance to the Assured. ….
 5)   From renewal 1992, premium payment and claims settlement to be by monthly bordereaux."
   The introduction concluded by saying that no major changes were planned for 1992-3, that steps would be taken to simplify policy issuance and that the placing would be strengthened by improving its geographic base into the Pacific North-west, the South-west and the North-east.
31.   A draft slip was attached setting out the terms of the cover, which included two sections as follows. Section A was essentially two classes of property cover with a limit in each case of US$10,000,000 each and every loss each and every location. Section B, on the other hand, was liability cover described as comprehensive general liability insurance with a limit of US$1,000,000. The draft slip set out the general conditions of the cover in detail. They included condition I, which provided that claims handling procedure was "as per attached agreement". No such agreement was attached. There followed a schedule of agents which included a number of well known brokers.
32.   The next section of material attached to the fax of 18 June was entitled "Premium and Loss Exhibits". It included the crucial document, so far as the reserving policy point is concerned, namely a table of figures expressed to be "as at 8th June 1992" as follows:
  
   AHL CONTRACTORS PREMIUM AND LOSS EXHIBIT SECT A & B
FACILITY NO GROSS PREMIUM INCURRED LOSSES LOSS RATIO
0919089 $ 157,653 $ 52,275 33.15%
0956090 $ 1,649,525 $ 19,347 1.17%
0956091 $ 3,869,003 $ 140,034 3.62%
0956090 $ 8,581,632 $ 3,385,943 39.45%
0956091 $11,315,479 $ 165,899 1.46%
  $25,573,292 $ 3,763,498 14.71%
  
  
33.   The figures for 0956091 were unclear on the fax but the judge reproduced them in his judgment and there is no evidence that they were not sufficiently clear to ARIG at the time. The only other figures which Generali provided were figures for section A only, in respect of the two facilities for 1991 as follows:
0956091 $ 636,516 $ 12,084 1.90%
0957091 $2,000,000 $ 156,572 7.60%
  $2,697,116 $ 168,656 6.25%
34.   The next document set out details of the reinsurance available to ARIG. There followed 12 pages of forms to be filled out by prospective insured which set out detailed information required including five years’ loss history. In all there were 28 pages of various kinds of information attached to the fax. The judge made these findings of fact as to what ARIG did on receipt of the fax. The fax was addressed to Mr Hans Gewalt, who was ARIG’s senior underwriter but who was about to take his summer break. He left a note in Swedish dated 26 June on the fax addressed to his deputy, Mr Ture Anderberg, saying "sorry, Ture but I did not have time to look at these offers". Mr Anderberg subsequently passed the fax to Mr Rehman, who was an underwriter in ARIG’s Engineering Department with experience of writing international facultative reinsurance business, including major construction risks. By the time he looked at the fax the proposed inception date of 29 June had already passed.
35.   It was regarded as an important offer because Mr Anderberg and Mr Rehman felt it necessary to consult Mr Gewalt. It was also regarded as an unusual risk because ARIG did not ordinarily write small contractors or US based business risks. Having reviewed the documents, Mr Rehman spoke to Mr Wardlaw and Mr Toms of Alexander Howden on the telephone on 3 July. During that conference call he made a number of notes on the documents. In this regard the judge said this in paragraph 16 of his judgment:
  
   "He recalls discussing the capacity which was on offer at that time and the split of the forecast premium of US$20 millions for the next year between sections A and B, and between the AH package and the WK Cooper package. He discussed the ‘pre-qualification procedures’ and was led to believe, accurately, that AH would initially assess the suitability of the proposal and weed out those which they believed would be unacceptable to underwriters and that the Lead Reinsurance underwriters at Generali would carry out their own assessment and quote terms, where the risk was acceptable. He also says that he was led to believe that Munich Re was taking 35%, Generali was taking 40% and Commercial Union 20%. He says he ‘understood’, from the fax and the telephone call that these percentages applied to both sections A and B. He says that he formed a very favourable impression of the potential business as the brokers had been able to answer all their questions and ‘give the necessary comfort we were looking for’. He recalls a further telephone conversation with AH on 9 July 1992, when he chased them up for the ‘wordings’ for which he has asked in the earlier call."
   On his copy of the fax of 18 June Mr Rehman noted 40% against Generali, 20% against Commercial Union and 35% against Munich Re.
36.   On 9 July Alexander Howden sent ARIG four faxes. The first attached 33 pages of policy wording for section B and for section C, which set out the general conditions applicable to sections A and B. The second attached 11 pages of additional broking information and graphics which broke down the account into various classifications. The third attached 21 pages of standard endorsements and the fourth attached 29 pages which comprised an index of the policy, the policy schedule and the policy wording for section A.
37.   On the next day, 10 July, Mr Rehman, on behalf of ARIG, faxed Mr Toms of Alexander Howden as follows:
  
   "Derek, thanks for your patience and input on this placement in view of which, based on your offer 18-6-92 and faxes 9-7-92, we are prepared to accept a 7.5% line of whole to be signed in full subject the CGL R/I available and total deductions to ARIG 30% in all.
    Please forward final slip(s) for our signature after finalising all points such as fronting, protection, etc.
    Pls note that our liability only incepts from inception of this facility for projects to be declared from now; no run off liability is accepted from previous periods or contracts already having incepted and/or completed. Pls confirm. Best rgds."
38.   On 14 July Mr Rehman received a fax from Mr Toms and Mr Wardlaw of Alexander Howden dated 13 July which included the following:
  
   "… we confirm that we will be using your line of 7.5% on each sections with respect to declarations attaching on or after the 2nd of July 1992.
    … we confirm that we have arranged excess of loss re-insurance in respect of your 7.5 % line on the CGL section – section B. Your re-insurance cover note will be sent shortly.
    … we anticipate that you will be writing as a retrocession of the Generali United Kingdom Branch. Will advise again if this is not the case. Will send copy of our slip in next day or so. … we confirm total deductions to ARIG are 30%.
    Lastly, we confirm that this placing only apples to occurrences under declarations bound during the twelve months @ July 2nd 1992. Liability from back years is not transferred …
    Trusting all in order and many thanks."
39.   On 21 July Mr Toms sent Mr Rehman "the retrocession slips, the reinsurance slips which front them and the direct slips" attached to a fax which contained the following:
  
   "We believe that you will find these to be in order and would ask you to authorize your line, where shown on each retrocession slip and fax them back to us immediately.
    We believe you will find everything in order but if you have any questions please contact the undersigned immediately."
   If, which is by no means certain, I have understood the position, the direct slips evidenced the contract between the insured and United, the reinsurance slips fronting the insurance evidenced the contract between United and Generali and the retrocessions slips evidenced the contracts of retrocession between Generali and the retrocessionaires including ARIG.
40.   The judge held that Mr Rehman carefully examined the retrocession slips. The judge was to my mind right to reach that conclusion; the care that Mr Rehman took can be seen from the questions he asked and the notes he made on the various documents which he reviewed. He observed that Alexander Howden had carelessly recorded the reinsured’s retention as nil, which would have meant that Generali’ retention was nil. Mr Rehman pointed out the error to Alexander Howden on the telephone and was told that Generali was retaining 26.3175% for section A and 39% for section B. He noted those percentages on the slip before reviewing the slip with Mr Gewalt in early September 1992.
41.   As to participation, the documents sent to Mr Rehman seem to evidence the following. So far as section B is concerned the position appears to have been this. Generali wrote a 96.5% line. The remaining 3.5% was written as to 1.25% by Commercial Union, 1% by Indemnity Marine and 1.25% by Ocean Marine, although it may be that Indemnity Marine and Ocean Marine were part of Commercial Union so that the reinsurers of United were Generali as to 96.5% and Commercial Union as to 3.5%. Of Generali’s 96.5% line, Generali retained 39% and retroceded the remainder as follows: ARIG 7.5%, GAN 5%, Star 5% and HIH/Wellington 40%. Thus Munich Re did not participate in section B in any guise.
42.   The position was somewhat different in the case of section A. Generali agreed to write a 41% (or possibly 40%) gross line. Great Lakes, which as I understand it was part of Munich Re, took a 35% line, Commercial Union took 20%, DP Spyer took 1%, Commercial Union took a further 1.25%, Ocean Marine took 1.25 % and Indemnity Marine took an illegible percentage which may have been 1%. Generali retroceded 7.5% to ARIG. It is not at all clear (at least to me) what other retrocession was in place for section A risks. In these circumstances I am not sure how the figure of 26.3175% given to Mr Rehman on the telephone was arrived at, but it is not necessary to analyse the percentages in any detail in order to resolve the issues in the appeal.
43.   So far as the Munich Re issue is concerned, the only point of significance is that a careful perusal of the documents sent to ARIG attached to the fax of 21 July would have shown Mr Rehman that Munich Re were participating as to 35% in section A and not at all in section B. It will be recalled that the original fax of 18 June had stated:
  
   "The leading Insurers in London are [Generali] with support from the Commercial Union and the Munich Re."
   It will also be recalled that Mr Rehman was told on the telephone that their relative percentages were 40%, 20% and 35% respectively.
44.   The relevant slips included the following condition:
  
   "Claims Handling Procedure:- as per the attached agreement. Claims settlement authority is US$25,000 per claim"
   No agreement was attached, but on 26 August Alexander Howden sent Mr Rehman a fax to which were attached "the Claims Handling Procedures which have been agreed on the direct placings" and which were stated to be given to ARIG to complete their file because they were "participating on a reinsurance basis". The document attached gave details of various procedures agreed with G&T but did not make any express reference to any reserving policy or methodology.
45.   On 7 September Alexander Howden faxed Mr Rehman asking him to sign and stamp the reinsurance slips as a matter of urgency. Mr Rehman discussed the matter with Mr Gewalt. As the judge put it, he had carefully compared the conditions on the draft slips sent in June with the retrocession slips sent in July for signature and noticed some minor discrepancies which he corrected before signing and stamping the slips and returning them to Alexander Howden. He did so by attaching signed copies of the following documents to a fax dated 9 September: the retrocession slips on the basis that ARIG’s participation was 7.5% of the whole, the CGL reinsurance terms and the original/main placing slips. I note that ARIG’s copy of the fax contains a box stating that it was sent on 8-9-92 and that on its copy of the Alexander Howden fax of 21 July Mr Rehman noted "Replied 8-9-92".
46.   The documents as signed by Mr Rehman do not explain how the 26.3175% is arrived at. On the retrocession slips prepared by Alexander Howden ARIG’s participation was stated to be 7.5%, although in one case the words "NB Signing 95.70%" were added. In each case in the box marked "Acceptance" Alexander Howden had written "Fax 10-7-92", which was presumably a reference to the fax sent by Mr Rehman to Alexander Howden referred to above. By contrast, in each of the four boxes signed by Mr Rehman, which were of course in respect of sections A and B of each of the facilities, the percentage stated was "7.5% line of whole" and the date was 07-09-92, although the fax was sent on 9 September.
47.   There was an issue between the parties at the trial as to when the contracts were made. This was relevant only to the question whether ARIG was induced to enter into them by a misrepresentation by or on behalf of Generali as to the participation of Munich Re. It was said by ARIG that in the fax of 18 June and the subsequent telephone conversation on 3 July Alexander Howden represented on behalf of Generali that Munich Re participated as to 35% in the whole cover, that is in sections A and B, that that representation was material, that it was untrue because Munich Re was participating only in section A, that ARIG was induced to enter into the contracts by the misrepresentation and that the contracts were made by its fax of 10 July.
48.   Generali’s case, on the other hand, may be summarised as follows. Alexander Howden made no representation as to the participation of Munich Re in either the fax of 18 June or the subsequent telephone conversation or, if it did, nothing written or said amounted to a representation that Munich Re’s participation was 35% in both sections A and B. If that was wrong, no representation made was material and in any event it did not induce ARIG to enter into the contracts. In any event the contracts were made on 9 September, by which time any misrepresentation had been corrected because by then ARIG had been told that Munich Re’s participation was restricted to section A.
49.   I shall return to the judge’s findings on those issues below, but it is convenient to note here that the judge held that the contract was made on 9 September and not earlier. ARIG challenges that conclusion in this appeal. The judge set out his reasons in paragraph 23 of his judgment as follows:
  
"(1)   The fax of 18 June 1992 was an invitation by AH to ARIG to treat. It did not amount to an offer since, amongst other things, it did not specify the respect in which ARIG could participate – whether a co-reinsurer of United or as a retrocessionaire of Generali.
 (2)   ARIG’s fax of 10 July 1992 was not an acceptance of an offer, because no offer had been made and in any case the acceptance of a 7.5% line could have been on the basis of a co-insurance [where the contracting party would include United] or a retrocession [where the contracting party would be Generali].
 (3)   The reference by ARIG in that fax to the need for AH to send final slips for signature after finalising points on matters such as fronting, protection etc is significant for two reasons: first, it shows that as at that date ARIG did not regard the contract as concluded because there were terms and conditions still to be sorted out and agreed; and second, it shows that ARIG regarded the formal act of scratching the slip as the last step in the contractual chain.
 (4)   The fax from AH saying that ‘we will be using your line of 7.5% still left open the question whether ARIG was contracting as retrocessionaire or as co-insurer. That issue remained to be resolved and, until it was, there could not have been a concluded contract. When AH had decided in which capacity ARIG was to contract they did so at the same time as sending ARIG the final slips to scratch which they asked to be returned signed and stamped as a matter of urgency. It was only when these slips were sent that ARIG’s contractual position had been clarified.
 (5)   From an objective analysis of the contractual documents, it seems to me clear that both parties intended the contract to be concluded when the slips were scratched and returned by ARIG to AH. That would be the usual method by which insurance, reinsurance or retrocession was concluded. Thus, in my judgment, the contract was only concluded in September when that step was taken by ARIG. "
50.   Mr Flaux submitted that the judge was wrong to hold that the contracts were not made until 9 September and that he should have held that they were made on 10 July. Although the parties had reached a large measure of agreement by the time Mr Rehman sent the fax of 10 July, I agree with the judge that no binding agreement was made until 9 September. In particular, no agreement had been reached as to who the other party to the contract was going to be and, as I read the fax, Mr Rehman was saying that once points of that kind were resolved the agreements would come into existence when ARIG signed the slips.
51.   In short, I agree with the judge that an objective analysis of the documents supports the conclusion that both parties intended that the contracts would be concluded when the slips were scratched and returned to Alexander Howden. Mr Flaux criticised the judge’s statement that that would be the usual method by which, insurance, reinsurance or retrocession was concluded. While it may be going too far to say that it would be the usual method of contracting, it is certainly a method often used. In any event, as I read his judgment, his conclusion was essentially based upon an analysis of the documents. As already stated, I agree with that analysis.
   Legal principles
52.   Before considering the issues of fact, it is convenient to discuss briefly the relevant legal principles.
   Non-disclosure
53.   It is common ground that the relevant principles are stated by the House of Lords in Pan Atlantic Insurance Co Ltd v Pine Top Insurance Co Ltd [1995] 1 AC 501. In order to be entitled to avoid a contract of insurance or reinsurance on the ground of non-disclosure the insurer or reinsurer must show that the fact not disclosed was material and that its non-disclosure induced the contract.
   Misrepresentation
54.   In order to be entitled to avoid a contract for misrepresentation an insurer or reinsurer must prove that he was induced to enter into the contract by a material misrepresentation made by or on behalf of the insured or reinsured. That principle is not in dispute.
   Materiality
55.   In Pan Atlantic the House of Lords, by a majority, held that to be material a fact did not have to have a decisive influence on the mind of the prudent underwriter. The test is that stated in subsections 18(2) and 20(2) of the Marine Insurance Act 1906, which relate to non-disclosure and misrepresentation respectively and which the House of Lords held set out the common law principles relevant to non-marine (as well as marine) insurance. The subsections are in the same terms and provide:
  
   "(2) Every circumstance is material which would influence the judgment of a prudent insurer in fixing the premium, or determining whether he will take the risk."
   Inducement
56.   In Pan Atlantic the House of Lords held that the material non-disclosure or misrepresentation must induce the contract. It was expressly held that it is not sufficient that the non-disclosure or misrepresentation is material. Lord Mustill expressed his conclusion in this way at page 549D-E:
  
   "For these reasons I conclude that there is to be implied in the Act of 1906 a qualification that a material representation will not entitle the underwriter to avoid the policy unless the misrepresentation induced the making of the contract, using "induced" in the sense in which it is used in the general law of contract."
57.   It was not necessary for Lord Mustill to consider the precise meaning of "induced" in the general law of contract. Lord Mustill did, however, refer to a presumption of inducement which has been given some consideration subsequently: see eg Clarke on The Law of Insurance Contracts, 3rd edition paragraph 22-3C, Arnould on The Law of Marine Insurance, 16th edition, volume 3, paragraph 611, and St Paul Fire & Marine Insurance Co (UK) Ltd v McConnell Dowell Constructors Ltd [1995] 2 Lloyd’s Rep 116.
58.   In paragraph 611 of Arnould the editors say:
  
   "Under the general law, a misrepresentation need not be the sole inducement, but must have been a real and substantial cause affecting the decision of the representee to enter into the contract, or to do so on the terms agreed."
   They cite as examples JEB Fasteners Ltd v Marks, Bloom & Co [1983] 1 All ER 583 and Edgington v Fitzmaurice (1885) 29 ChD 459. In paragraph 22-3C of Clarke it is said that what is usually required is what is identified as a type B representation, which is defined as follows:
  
   "Type B is a representation such that, if the recipient had known the truth, he would still have been willing to make the contract, but only on different terms, notably, but not only as to premium."
59.   It seems to me that the true position is that the misrepresentation must be an effective cause of the particular insurer or reinsurer entering into the contract but need not of course be the sole cause. If the insurer would have entered into the contract on the same terms in any event, the representation or non-disclosure will not, however material, be an effective cause of the making of the contract and the insurer or reinsurer will not be entitled to avoid the contract. Thus I agree with Sir Christopher Staughton, whose judgment I have seen in draft, that, in this context at least, causation cannot exist when even the ‘but for’ test is not satisfied; cf the recent decision of the House of Lords in a very different context in Fairchild v Glenhaven Funeral Services Ltd [2002] UKHL 22. (Bailii)
60.   Those principles seem to me to be consistent with the approach of this court in St Paul Fire & Marine v McConnell: see per Evans LJ (with whom Rose and Nourse LJJ agreed) at pages 124-5, where he discussed the general principles, and at page 127, where he held that, if the three underwriters who gave evidence had been told the truth, on no view would they have underwritten the insurance at the same premium on terms which included subsidence risk. Evans LJ also considered the role played by presumption in this class of case. He did so in the context of a fourth underwriter who was not called to give evidence, no doubt because the trial took place before the decision of the House of Lords in Pan Atlantic.
61.   Evans LJ put the position thus at page 127:
  
   "The existence of such a presumption is recognised in the authorities; see Halsbury’s Laws vol 31 par 1067 where the law is stated as follows:
  
    Inducement cannot be inferred in law from proved materiality, although there may be cases where the materiality is so obvious as to justify an inference of fact that the representee was actually induced, but, even in such exceptional cases, the inference is only a prima facie one and may be rebutted by counter evidence.
    The authorities cited include Smith v Chadwick (1884) 9 App Cas 187 and in my judgment they justify the above statement of the law. This provides a reminder of the need to distinguish "materiality" from "inducement", although inevitably the two overlap.
    Here, the evidence of the three underwriters who did give evidence and of the expert witnesses was clear. If the underwriters had been told the true state of the ground conditions, as revealed by the 1982 report, and of the conflicting views expressed by the authors of that report and by Worleys, then they would have called for further information and in all probability either refused the risk or accepted it on different terms. In fact, all four underwriters including Mr Earnshaw accepted it without any relevant enquiries. There is no evidence to displace a presumption that Mr Earnshaw like the other three was induced by the non-disclosure or misrepresentation to give cover on the terms on which he did. In my judgment, these insurers also have discharged their burden of proof."
   It appears to me that a presumption of this kind really amounts to no more than this. It simply operates where the evidence before the court is enough to lead to the inference that the insurer or reinsurer was, as a matter of fact, induced to enter into the contract.
62.   In all the circumstances I would summarise the relevant principles of inducement in this context in this way:
  
 i)   In order to be entitled to avoid a contract of insurance or reinsurance, an insurer or reinsurer must prove on the balance of probabilities that he was induced to enter into the contract by a material non-disclosure or by a material misrepresentation.
 ii)   There is no presumption of law that an insurer or reinsurer is induced to enter in the contract by a material non-disclosure or misrepresentation.
 iii)   The facts may, however, be such that it is to be inferred that the particular insurer or reinsurer was so induced even in the absence from evidence from him.
 iv)   In order to prove inducement the insurer or reinsurer must show that the non-disclosure or misrepresentation was an effective cause of his entering into the contract on the terms on which he did. He must therefore show at least that, but for the relevant non-disclosure or misrepresentation, he would not have entered into the contract on those terms. On the other hand, he does not have to show that it was the sole effective cause of his doing so.
   Correcting a misrepresentation
63.   Where the insured or reinsured corrects the misrepresentation or discloses the material fact before the insurer or reinsurer enters into the contract, the latter will not be entitled to avoid the contract for misrepresentation or non-disclosure. In such circumstances it may be said that there was no longer any or any material misrepresentation or non-disclosure or it may be said that there was no inducement. Perhaps it does not matter.
64.   The correction must be fairly made to the insurer or reinsurer such that the corrected picture is fairly presented on behalf of the insured or reinsured and comes to the knowledge of the insurer or reinsurer. It is not sufficient to say that he would have discovered the true position if he had acted with all due care: see eg Chitty on Contracts, 28th edition, vol 1 paragraph 6-036 and Redgrave v Hurd (1881) 20 ChD 1. As I see it, it will in each case be a question of fact whether the misrepresentation was corrected so as to ensure that the corrected facts came to the knowledge of the insurer or reinsurer or whether, when the contract was made, the insurer or reinsurer was induced to make it by the original material misrepresentation or non-disclosure.
   Munich Re
65.   Although this is the third of the three points identified above, it is convenient to consider it first because it naturally follows from the primary facts set out above. I have set out the issues between the parties which are relevant to the Munich Re point above. The judge’s conclusions, which are to be found in paragraphs 28 to 31 of his judgment, may be summarised as follows:
  
 i)   The fax of 18 June did not contain any representation that Munich Re participated in the whole package. The words "with support from the Commercial Union and the Munich Re" in the fax of 18 June could not mean more than that Munich Re were participating in the packages and did not necessarily mean that they were participating in both sections A and B. Mr Rehman misunderstood the position because he wrongly thought that it was not possible to participate in only one of the packages, although (as I read his judgment) the judge expressed considerable scepticism as to whether Mr Rehman was really of that view.
 ii)   As to the telephone conversation on 3 July, the judge said this in paragraph 31:
  
   "I am not prepared to place any great weight on the oral conversation between Mr Wardlaw and Mr Rehman at the beginning of July 1992. Mr Wardlaw had no recollection of the conversation; the witnesses were dealing with something that occurred 9 years ago and for which there are no notes relevant to this issue. Mr Wardlaw says he thinks he would have explained the full position about Munich Re’s participation. I doubt he did so because Mr Rehman had obviously talked himself into believing that participation could only be across the board. The most obvious explanation for how this issue came to the fore is simply that Mr Rehman has misled himself and thought, through his own ignorance, that support from Munich Re must have involved their participation in both Sections of both packages. The truth was staring at him in the face. It is most improbable that the brokers would have negligently or deliberately misrepresented the position when the true position was there for all to see."
 iii)   In any event the judge did not consider that what was said about Munich Re played any part in the decision taken ARIG to participate in the programme. In reaching that conclusion he placed particular reliance upon the oral evidence of Mr Anderberg to which I refer below.
   Misrepresentation
66.   I agree with the judge that the fax of 18 June does not itself contain a representation that Munich Re and Commercial Union were participating in both sections of the cover. It merely asserts that the leading insurers were Generali with support from Munich Re and Commercial Union. That statement seems to me to be consistent with their participating in only part of the cover, especially where any consideration of the documents would show that the business was divided into two sections and there was no reason at all why it should not be possible to participate in one and not the other.
67.   However, for my part, I do not think that it is possible to dismiss the evidence as to what was said in the telephone conversation on 3 July in quite the summary way in which the judge did in the passage quoted above. In particular, it is not correct to say that no notes relevant to this issue were made at the time. On the contrary, on his copy of the fax Mr Rehman noted 40%, 20% and 35% against Generali, Commercial Union and Munich Re respectively. That seems to me to afford strong support for his evidence that he was told that those were the percentage participations of each and for the conclusion that, if he had been told that any of the percentages related to only one section of the business, he would have noted that fact on the fax.
68.   That seems to me especially to be so given that one of the percentages related to Generali. It must have been clear that Generali was participating in the whole of the cover and it seems to me to be a reasonable inference from the fact that Mr Rehman noted one percentage against each that he understood that they were each participating on the same basis, albeit in different shares. I would therefore hold that, although the fax itself did not contain the representation alleged, it was represented to Mr Rehman in the course of the telephone conversation that Generali were participating as to 40%, Commercial Union as to 20% and Munich Re as to 35%.
69.   It is of course possible that Mr Rehman was told that those were the percentages of their respective participation in section A, since those were indeed the true figures for section A. However, if that was indeed what he was told, it is difficult to understand why he did not note that fact on the fax. To my mind it is more likely that he understood from what he was told that those were the percentage participations of each in the whole business. In these circumstances, I would hold that it is more probable than not that that is what he was told.
70.   It is common ground that any such representation was a misrepresentation because it was not true, since Munich Re was only participating in section A and not in the much more risky section B. Given my earlier conclusion that the contracts were not made until September, the next question is whether the misrepresentation was corrected by the slips attached to Alexander Howden’s fax of 13 July, which show that Munich Re’s participation was limited to section A.
71.   Mr Flaux submitted that the slips were not sufficient to correct the misrepresentation. Although he accepted that the slips attached to the 13 July fax show that Munich Re (ie Great Lakes) only participated in section A, he submitted that simply to include that information in the middle of a considerable amount of complex documentation was not enough to correct the untrue representation made on the telephone on 3 July. He submitted that ARIG’s attention should have been specifically drawn to the fact that the documents contained different information as to the participation of Munich Re from that which he has been given earlier.
72.   Mr Flaux placed some reliance upon the decision of Hobhouse J in Abrahams v Mediterranean Insurance & Reinsurance Co Ltd [1991] 1 Lloyd’s Rep 216, but he properly accepted that the case is not directly in point and I have not found it to be of particular assistance in this context. I have already expressed my view that each case depends upon its own facts. On the facts here, there is I think force in Mr Flaux’s point that, having misrepresented the true position on 3 July, it was incumbent upon Generali, or its brokers Alexander Howden, expressly to bring the true position to the attention of ARIG, that they failed to do so by doing no more than including the participation of Great Lakes on the slips and that, since no-one at ARIG in fact appreciated the true position, ARIG was induced to make the contracts by the original representation. It is not, however, necessary finally to resolve this question in order to determine the Munich Re point because of my conclusion on inducement, to which I now turn.
   Materiality and inducement
73.   The judge did not consider materiality separately but expressed his conclusion that ARIG was not induced to enter into the contracts by any misrepresentation as to the participation of Munich Re in paragraph 30 in these terms:
  
   "But in any event I do not consider that what was said about Munich Re played any part in the decision taken by ARIG to participate in the programme. Although Mr Anderberg suggested in his witness statement that but for Munich Re’s participation ARIG would not have participated, his position modified when he gave evidence. Quite rightly, I think, he said that the decision taken by ARIG whether or not to participate was taken on the merits of their appraisal of the risk but that the fact Munich Re was also a participant was a comfort factor after the decision had been taken."
74.   Mr Boyd submitted that that conclusion was justified on the evidence. He relied in particular upon the following exchange during the cross-examination of Mr Anderberg:
  
"A.   Am I right in thinking that you are asking me a question -- because it is quite a long question -- that the Munich Re decision did not have any impact on our own decision?
 Q.   On your underwriting judgment, on deciding whether or not to accept this risk, that is what I am asking.
 A.   You know, we are making an independent underwriting.
 Q.   I am sure you are.
 A.   On all risks. All the other factors involved here -- the excellent figures, et cetera -- indicated that this was a very profitable account and I am sure we would have written it without the participation of Munich Re, I am sure. It was a comfort factor but Munich Re is not guiding our underwriter, we are guiding our underwriting ourselves.
    MR JUSTICE MORISON:The position is that you take an independent underwriting decision, having made the decision to participate, you are feeling comfortable with that decision having regard to the participation of Munich Re in the programme.
 A.   Yes."
75.   It appears to me that on that evidence the judge was entitled to conclude that ARIG would have entered into the contracts whether Munich Re was a participant or not. Mr Flaux submitted, however, that the judge should have had regard to other evidence given by Mr Anderberg, including his evidence in re-examination as follows:
  
"Q   If you had been told that Munich Re was on section A and section B at the time you assessed the risk, and then just before you accepted the line you were told that Munich Re was not, in fact, on section B, what view would you have taken?
 A   I would have been very puzzled, asking "why", because this was a profitable account and I would have gone back to the producer and asked them for a clarification and why Munich Re did not participate in section B, which was actually, as far as I remember, when we were asking for the premium split it generated, on the advice premium volume of 20 million, it actually generated 75 per cent. Why should Munich Re not participate on this 75 per cent of the 20 million if it was a profitable account with good prospects? That would have been my immediate reaction.
  
    If I would not have got any clarification which I could buy, I would have declined participation."
76.   Mr Anderberg did not speak with one voice throughout his evidence. As the judge pointed out, his evidence in his witness statement was different from that which he gave in cross-examination. The judge had to weigh up the whole of his evidence including what he said in both cross-examination and re-examination. There was no evidence as to why Munich Re participated only in section A or as to what they (or indeed Generali) would have said if asked the question suggested by Mr Anderberg in re-examination. I have reached the clear conclusion that the judge was entitled to reach the conclusion which he did and to hold that what was said about Munich Re played no part in ARIG’s decision to participate.
77.   Since preparing this judgment in draft I have seen Ward LJ’s draft judgment. I recognise that he has formed a different view on this point. He has also set out more of the evidence than I have. I have reconsidered the question of inducement in the light of his judgment. In doing so I have reconsidered all the evidence including the evidence which he has set out.
78.   I agree with Ward LJ that in determining whether the insurer or reinsurer was induced to enter into the contract the court does not embark upon the exercise of finding the decisive cause or the main reason. However, I remain of the view that the non-disclosure must be an effective (or as Arnould puts it in the passage quoted in paragraph 58 above) a real and substantial cause of the decision to enter into the contract. That conclusion seems to me to be supported by the passages from the judgments in Edgington v Fitzmaurice quoted by Ward LJ.
79.   Having reconsidered the evidence I also remain of the view that it was open to the judge to conclude that what was said (or written) about the participation of Munich Re played no part in ARIG’s decision to participate. In short, it was open to the judge to hold that ARIG had not shown that, if it had known that Munich Re was participating only in section A, it would not have entered into the contracts or would have taken some other share and I can see no basis upon which this court could properly interfere with that conclusion.
80.   It follows that it is not necessary for us to consider materiality separately, any more than it was necessary for the judge to do so. Finally, I would simply observe that, even if the misrepresentation was material, there is no inconsistency between such a conclusion and the conclusion that it did not induce the contract. As was expressly held in Pan Atlantic , in order to show that a misrepresentation is material, it is not necessary to show that it would have a decisive influence on the mind of a prudent underwriter, whereas in order to show that a misrepresentation induced the contract it is necessary to show (at least) that, but for the misrepresentation, the particular underwriter would not have made the contract, either at all or on the terms on which it was in fact made.
81.   For these reasons, in agreement with Sir Christopher Staughton, I would dismiss the appeal in so far as it relies upon the Munich Re point. I turn to reserving policy.
   Reserving policy
82.   The only figures presented to ARIG by Generali before the initial contracts were made were those attached to the fax dated 18 June which I have set out in paragraphs 32 and 33 above. The only other figures were those similarly produced as at 18 March 1993 when ARIG was considering renewing the contracts for two months from 2 July to 1 September 1993.
83.   They were as follows:
  
   "CONTRACTORS PREMIUM AND LOSS UPDATE SECTIONS A & B
FAC NO GROSS PREMIUM PAID OUT-STANDING INCURRED
Paid & O/S		 LOSS
RATIO
910989
 $ 233,991 $ 31,772 $ NIL $ 31,772 13.57%
956090
 $ 1,666,516 $ 21,833 $ 456,732 $ 478,565 28.70%
956091
 $ 4,596,397 $ 223,529 $ 249,454 $ 472,983 10.29%
956092
 $ 3,245,289 $ NIL $ 10,500 $ 10,500 00.32%
957090
 $ 10,237,000 $ 2,661,930 $ 1,783,384 $ 4,445,314 43.42%
957091
 $ 13,334,976 $ 330,181 $ 1,162,273 $ 1,492,454 11.20%
957092
 $ 11,986,078 $ 4,027 $ 160,223 $ 164,250 01.37%
           
TOTAL $ 45,300,347 $ 3,273,272 $ 3,822,566 $ 7,095,838 15.66%
   Those figures excluded a particular loss in respect of which coverage was not renewed and were stated to be provided by the third party claims administrator ("TPA"), which was of course G&T, as being in excess of deductibles and self insured retentions ("SIRs").
84.   As in the case of the earlier figures, Alexander Howden also provided the figures referable to section A only, as follows (ignoring the cents):
"FAC NO GROSS PREMIUM PAID OUT-STANDING INCURRED
Paid & O/S		 LOSS
RATIO
956091
 $ 749,418 $ 96,264 $ 76,633 $ 172,897 23.07%
956092
 $ 394,640 $ NIL $ 6,000 $ 6,000 01.52%
957091
 $ 2,332,347 $ 148,970 $ 366,789 $ 515,7594 22.11%
957092
 $ 2,217,193 $ 4,027 $ 78,223 $ 82,250 03.71%
TOTAL $ 5,693,598 $ 249,261 $ 527,645 $ 776,906 13.65%
           
85.   The judge observed that it was common ground that both sets of figures were largely accurate as a statement of what was shown on the books of G&T. They were the result of G&T’s reserving policy. So far as necessary, I shall return below to what, if anything, they reveal but it is first appropriate to identify ARIG’s case on both non-disclosure and misrepresentation. By a re-amended defence and counterclaim dated April 2000 ARIG alleged that
  
   "… [Generali] failed to disclose to [ARIG] that the claims administrator at all material times … [G&T] operated an unusual and/or imprudent reserving policy."
   It was further alleged that that failure was a failure to disclose a material fact and that, as a result, Generali presented materially inaccurate loss statistics in the two sets of figures set out above. Finally it was said that ARIG was thereby induced to enter into the contracts.
86.   The judge described the position thus in paragraphs 37 and 38:
  
   "What is complained of is [the figures’] misleading character. ARIG say that the figures would have looked quite different had the TPA [ie G&T] not operated an unusual and imprudent reserving policy.
    Through Mr Simon QC, ARIG put their case on the need to disclose the reserving policy in two separate ways. He submitted that the reserving policy was imprudent and unusual and, ipso facto, should have been disclosed as a material fact as a part of the fair presentation of the risk. Second, he submitted that without disclosure of it, the figures presented in the fax of June 1992 and in March 1993 were materially misleading. I agree with Mr Collins QC on behalf of Generali that there is effectively only one point. The reserving policy is not itself a material fact which requires to be disclosed. The evidence was clear on this point. An underwriter would only be expected to be informed of such a policy if it were unusual and imprudent so that it rendered the figures showing the ratio of losses to premiums unreliable. Was the policy unusual and imprudent so that the figures disclosed were misleading? Mr Simon’s two points merged into just one."
87.   ARIG challenges that conclusion in this appeal but, before giving such consideration as may be necessary to the question whether the judge was correct, it is convenient to consider whether G&T’s reserving policy was unusual or imprudent as alleged by ARIG. This involves a consideration of what that policy was, what the relevant standards were at the time and whether the policy met those standards. It also involves a consideration of the views formed by the judge as to the expert witnesses called on both sides.
   The expert witnesses
88.   In considering G&T’s reserving policy, the judge had the benefit of the evidence of two expert witnesses called on behalf of ARIG and one expert witness called on behalf of Generali. ARIG called Mr Kiverstein and Mr Beckerman. Mr Kiverstein is an accountant and statistician and Mr Beckerman is a US attorney, with considerable claims experience. Mr Kiverstein carried out an analysis of the results of the coverage as at 1999 and 2000. He prepared a schedule which formed the basis of Appendix A to ARIG’s pleaded case. As explained in the judgment, Appendix A contained some 702 entries out of an available 954 claims reported up to 18 March 1993.
89.   Mr Kiverstein calculated two figures for "expense reserves", by which he meant both a reserve against the expense of investigating and resisting the claim and a reserve against liability in respect of the claim. The "expense reserves" thus included an element for indemnity reserves as well as expense reserves properly so called. The two figures were included in Appendix A, which ARIG initially used to identify the loss ratios which it said that the figures presented to it on inception and renewal would have shown if Generali had not had an unusual and imprudent reserving policy. They were mostly US$10,000 for section A and US$40,000 for section B. Those figures were then used in many cases instead of nil which appeared in the G&T documents. The problem with them was, however, that they were based upon the actual results in 1999 and 2000. They were thus calculated statistically looking backwards. As the judge put it, they were formulaic in the sense that they were calculated by taking the losses and applying a mathematical averaging process and then rounding the derived figure downwards.
90.   Mr Kiverstein accepted that the reserves so calculated were not reserves which the TPA would make if approaching the reserving process on a case by case basis. The judge expressed his conclusions as to the value of this part of Mr Kiverstein’s evidence in this way in paragraphs 43 and 44 of his judgment:
  
   "It seemed to me that Appendix A did not provide the particulars which it purported to provide. As he said, "my reserve of $40,000 cannot be compared to the reserve needs on any individual file." He also explained that his Appendices omitted any deductions for deductibles and … SIRs. … But, crucially, the claims population upon which Mr Kiverstein prepared his work, namely some 5,452 claims, was different from the claims which were known about in June 1992 and March 1993, namely 512 claims and 954 claims respectively.
    It became apparent to ARIG’s advisers that valuable as Mr Kiverstein’s work may have been to ARIG he was not the witness who could talk about the Gay & Taylor reserving policy in response to Generali's expert, Mr Palange."
91.   Mr Kiverstein was not in the event used as ARIG’s reserving expert, since he approached the problem statistically, looking back in the light of hindsight, and not looking forward as G&T had to do at the time. It was no doubt in part at least for that reason that ARIG instructed Mr Beckerman, who gave evidence critical of G&T’s system or methodology. He was instructed to advise ARIG in response to the opinion of Mr Palange.
92.   In the context of this appeal it is to my mind important to note that the judge formed an unfavourable view of Mr Beckerman and a favourable view of Mr Palange. The advantage of a trial judge in assessing the evidence of expert witnesses with regard to expert evidence was noted by Brandon LJ in Joyce v Yeomans [1981] 1 WLR 549 at p 556:
  
   "In my judgment, even when dealing with expert witnesses, a trial judge has an advantage over an appellate court in assessing the value, the reliability and the impressiveness of the evidence of the experts called on either side. There are various aspects of such evidence in respect of which the trial judge can get the ‘feeling’ of a case in a way in which an appellate court, reading the transcript, cannot. Sometimes expert witnesses display signs of partisanship in the witness box or lack of objectivity. This may or may not be obvious from the transcript, yet it may be quite plain to the trial judge. Sometimes an expert witness may refuse to make what a more wise witness would make, namely, proper concessions to the viewpoint of the other side. Here again this may or may not be apparent from the transcript, although plain to the trial judge. I mention only two aspects of the matter, but there are others."
   Brandon LJ was there considering the evidence of medical experts, but in my view the same principles apply to all experts including experts on reserving.
93.   The judge expressed this view of the evidence of Mr Beckerman in paragraph 44 of his judgment:
  
   "I have to say at once that I did not regard Mr Beckerman’s evidence as particularly helpful or reliable. In the most general terms he tended to advocate ARIG’s case, no doubt through considerable experience of advocating claims in the US courts, often, I suspect, before a jury. He did not quite have the presence, authority and balance to be expected of an independent expert. But he was labouring, I think, under the disadvantage of being instructed late. … He was adamant that a claims handler MUST post a reserve within 30 days of a claim being made. He went on to say that he did not regard a 90 day delay as acceptable. 30 days was a US industry standard. There was no mention of this standard in his report, nor in the joint memorandum prepared by the experts nor in a book on the subject to which he helpfully drew our attention. He was of the view that there were really no exceptions to this standard. He felt able to make judgments about the adequacy of the reserves in relation to files for which he had only one piece of paper extracted. I consider that he exaggerated the need for a reserve in 30 days and was not prepared to look at the matter from the perspective of a claims handler in 1992/1993. Broadly speaking, I felt unable to rely upon Mr Beckerman’s evidence unless it was consonant with the evidence of Mr Palange."
94.   By contrast, the judge expressed this view of Mr Palange in paragraph 45:
  
   "Mr Palange gave me confidence about the reliability of his evidence. He is well qualified to give his opinion in this case. He has conducted "literally hundreds" of reserve and technical audits of US Insurers’ claims departments and TPAs. What he did was to carry out a sampling of 144 original files which he selected from Appendix A. He said, and I accept, that he selected samples of large and small incurred losses, large and small paid losses and a sampling of all sizes of loss where Appendix A had used the figure of US$40,000 (repetitively). He reached a number of conclusions, based on his own analysis of complete files chosen at random within the parameters set out, which I accept and find as follows."
   The judge then made a series of findings of fact to which I shall return below. For present purposes it is important to note that the judge’s conclusions were influenced to a considerable degree by the adverse view which he formed of Mr Beckerman as compared with the favourable view which he formed of Mr Palange. In considering the arguments advanced in this appeal, that is to my mind an important factor which we should keep in mind.
   G&T’s reserving policy
95.   Generali called the then Claims Director of the Program Risk Division of G&T, Mr Junior Reynaud, to describe the reserving policy carried out by G&T. The judge plainly regarded him as an impressive witness. In paragraph 46 the judge summarised the principal features of his evidence as to the policy adopted by G&T as follows:
  
"(1)   In almost every case, a claim under the programme would be in the form of a law suit. Claims were attended to immediately on receipt. He would look at the claim, as he looked at all of them in the early years. It would be allocated to a claims examiner, of whom there were 33 and entered in the system.
 (2)    A nil reserve was set until Gay & Taylor were in a position "to set a loss reserve that properly represented underwriters’ most likely financial exposure in the context of coverage provided by the insurance policy". This was rarely possible before extensive investigation had been carried out in the circumstances surrounding the event which had given rise to a claim." The court papers served on the insured would not usually contain sufficient information to set a reserve; for example, the formal claim might or might not contain a date of loss.
 (3)    Every claim made in the form of legal proceedings would be reviewed with coverage and defence counsel. In almost every such case, the files in the Court Depository would have to be looked at by coverage counsel whom Mr Reynaud often accompanied. The purpose of the search was to establish dates of ‘loss’ and ‘occurrence’ [this was prior to the decision in the Montrose case]; an indication of the damages being sought; a listing of all the insurance carriers and a listing of all sub-contractors. Once the sub-contractors had been identified it was usually possible to say from whom indemnities might be sought and whether the loss fell outside the insurance year. If the claim was large, then the amount of documentation would increase in size. However, completion of this process would not usually provide Mr Reynaud with sufficient information to set a reserve. It might take anything up to three years from first notification before sufficient information was available to set appropriate loss reserves.
 (4)    In summary, it was not possible to set a loss reserve which properly represented underwriters’ most likely financial exposure "until I had all of the information regarding the circumstances surrounding events that gave rise to the claim".
 (5)    As regards expense reserves it was not unusual to set a nil reserve because the insured had a minimum deductible/SIR of $2,500 (up to $50,000).
 (6)    As and when necessary, a reserve (and the file) would be reviewed."
96.   The judge added that Mr Reynaud said that the reserving policy used by G&T was not unique in the United States or in the insurance industry. He said that it was the policy used by his former employers, who were a large insurance group and by the Alexander Howden Property Claims Division, by whom he was employed for a short period. No-one had complained about it. The judge accepted this evidence, as in my opinion he was entitled to do.
97.   It was always recognised by Mr Reynaud that G&T’s system did not involve a set period within which a reserve must be put on a claim. His evidence was in effect that it was not sensible, or indeed possible, to do that because of the complexity of some of the claims advanced. The judge made further findings as to the system adopted in paragraph 45 of his judgment which are set out below. Those findings included the following specific finding as to the system adopted:
  
   "Gay & Taylor adopted a form of case reserving. That is, the reserves were applied on a case by case basis. Initially a nil reserve was set when a new file was opened and the claim would be investigated both as to ‘coverage’ and liability, often through outside attorneys or independent adjusters. Once the likely outcome was capable of being assessed, Gay & Taylor did not adopt a policy of stair step reserving, but pursued a policy of ‘one-time’ reserving "that is to say, to establish a reasonably accurate reserve for the first time it was set once adequate and accurate investigation information was received."
98.   It was submitted to the judge that in expressing his opinions Mr Palange treated G&T’s policy as being different from what it was because, it was said, Mr Palange knew that the system described by Mr Reynaud was an imprudent one and therefore, in effect, pretended that it was a different system. The judge rejected that submission. Mr Flaux submitted that he was wrong to do so and that he should, in any event, have held that Mr Reynaud’s evidence showed that the rigid system adopted by G&T was imprudent because it was based on the philosophy that no reserve would be set until all the information was available, which might take years.
99.   It can be seen that there is a close overlap between the question what G&T’s system was and whether it was unusual and imprudent. I shall therefore return to it after considering the standards of reserving policy.
   Relevant standards
100.   The general principle governing reserving policy was not in dispute. As pleaded on behalf of ARIG, Generali or its agent must:
  
   "make every effort to establish the full circumstances surrounding an event that has given rise to a claim. Having done so, it is incumbent on the claimant to set a reserve that properly represents its most likely financial exposure in the context of coverage provided by the policy(ies) and for that reserve to be fixed as soon as is reasonably practicable, including reserving for legal and associated expenses. The reserves should be reviewed on a regular basis (including liaison with and obtaining regular updates from external advisers, such as loss adjusters and/or lawyers involved in actual or potential claims)".
101.   The experts, namely Mr Kiverstein, Mr Beckerman and Mr Palange, prepared a joint memorandum which identified points of agreement and points of disagreement. The points of agreement included the following:
  
   "Matters of Market Practice
    It is the custom, practice and procedure in the United States insurance industry to post reasonably accurate loss and expense reserves for primary construction, liability and construction defect liability claims, in as timely a fashion as the facts and circumstances of each individual claim reasonably allow.
    Reserving
    One of the duties of a TPA is the posting (establishment and maintenance) of reserves. Reserves are used to help an insurance carrier predict the amount of money it will have to spend on paying claims and lawsuits.
    Case reserving is the most common form of reserving practice within the insurance industry in the United States. "Case" reserving means, simply, that each claim is reserved individually, on its own facts, circumstances, coverage and merits, i.e., case-by-case. Substantive changes in the facts surrounding a claim will justify increases or decreases in the case reserves set.
    Reserving done on a case by case basis, involves the claims handler (or TPA) managing the file and making their best educated appraisal of what each claim will ultimately cost.
    It is desirable to avoid "stair stepping" reserves and increasing reserves merely to cover individual partial payments (as made). However, this does not mean that reserves should not be adjusted. It merely means that reserves should be posted based on the most accurate information available at the time, acknowledging the fact that all claims and lawsuits evolve."
102.   There followed lists of factors to be taken into account when setting an initial reserve for construction claims including construction defect liability claims and when setting expense reserves respectively. The points agreed as set out in the joint memorandum continued:
  
   "Expense reserving should be based upon the reasonable likely cost of the investigation and of the defense of the claim. …
    When a new claim is received, there may be times when information is so limited that a determination of the indemnity or expense reserve amount cannot immediately be made.
    Normally, defence counsel will provide the claims handler with the information obtained from the discovery process. Defense counsel may also suggest a future course of discovery. However, defence counsel will not normally recommend a specific reserve amount, although he or she may be invited to do so by the claims handler.
    The retention of defence counsel, and the initiation of the discovery process, should result in the prompt receipt of at least some liability and damages information. Information discovered should be transmitted promptly by the defence attorney to the claims handler in writing.
    No insurer (or TPA) relative to large construction defect claims could establish reserves that would "properly represent[s] its most likely financial exposure in the context of coverage provide by the policy(ies)" until such time as they received adequate investigative data concerning coverage, liability and damages as the facts and circumstances of each individual claim reasonably allowed.
    Large losses are more likely to involve complex coverage issues and consideration. Once a coverage position is adopted by an insurer that is "adverse" to an insured, and the larger the potential damages exposure, the more likely it is that the insured will contest such a coverage position(s).
    It is generally accepted within the US Insurance Market that a large loss is so designated when the incurred value of such loss reaches or exceeds a "significant amount". Some insurers may select some other value, but US$100,000 is a somewhat common level utilised for a