AXA Versicherung Ag v Arab Insurance Group

Axa does not seek to appeal the primary facts found by the judge, as opposed to the inferences to be drawn from them. Its second ground of appeal is that there was no proper basis for the judge’s hypothetical broke and that the judge’s findings on inducement (that Mr Holzapfel would still have written the business following a fair presentation) was erroneous, whether the hypothetical broke is taken to be the one suggested by the judge to Mr Holzapfel in evidence or the one contained in his judgment i.e. including the voluntary particulars table.

In addition, Axa contends that the judge’s finding is a product of procedural unfairness. The hypothetical broke put by the judge to Mr Holzapfel when he was giving evidence was not pleaded; nor was it supported by evidence from the brokers that that is what they would have presented to him. It was also not part of Arig’s case. And when it came to the judgment various factors were introduced into the hypothetical broke as to what could or would have been said that had not been put to Mr Holzapfel nor been the subject of evidence by the brokers. In addition, the loss statistics suggested to him in evidence were not the same as those that the judge found should have been disclosed. As a result, Axa and Mr Holzapfel were deprived of the chance of dealing with, and responding to, a broke which only appeared in the judgment. This was unfair. Further the judge seems to have rejected his evidence in response to the hypothetical broke put to him by the judge without having cast any doubt on the answer when he gave it. Most importantly of all the judge’s conclusion cannot stand with important unchallenged evidence of Mr Holzapfel to which I shall refer.

Axa’s case was that the hypothetical broke should have included the statistics in the voluntary particulars. But the broke referred to in paragraph 163 of the judgment was based on the Swire Fraser statistics above and there are significant differences between the two: see [14] above. Further no one except the judge suggested that it was the Swire Fraser statistics that should have been presented.

Further the hypothetical broke at [163] included saying that:

“there had been a change of underwriter at Arig and the new underwriter was adopting a more rigorous and selective approach to the risks which he was prepared to accept, so that quite a number of the risks which the previous underwriter had been prepared to accept would probably not have been accepted under the new approach”.

That Arig could honestly have said this is difficult to square with the judge’s findings in paragraphs 94 - 101. In those paragraphs he held that the change of strategy was less clear cut than Arig suggested [98]; that the position was more nuanced than Mr Hylander’s witness statement suggested [99]; that during his tenure Arig had continued to write business with low deductibles and from time to time to write risks with each of the characteristics which he would have preferred to avoid albeit not necessarily on the same scale as before [99]; and that a memorandum written by Mr Hylander in January 1995, which was intended to form part of the placing documents for the renewal of the energy quota share and which said that Arig had “adopted a slightly different underwriting policy in 1993” was a fair summary, although by 1996 Arig was once more seeking to increase its book of construction risks as the premiums available for operational risks were becoming more competitive [101] i.e. was writing riskier risks. He concluded that there was genuine change in direction “but that it was the kind of intangible and gradual change which although real would have been difficult to demonstrate in any quantifiable way as part of any broking exercise” [111].

In addition, as the judge noted at [42] in relation to the information provided by Arig to Swire Fraser:

“the positive points made by Arig with a view to persuading prospective reinsurers that these historic results did not reflect the outcome which might be expected on the new risks for which Arig was seeking reinsurance had nothing to do with the fact that there was a new underwriter now in place and that a radically different underwriting strategy was now being followed. Instead Arig focused on the higher deductibles and higher premiums which it suggested could be achieved in 1995 market conditions”.

Further, as the judge found, Mr Bader and Mr Stow did not, when considering the broke for the first treaty, discuss the change of underwriting policy as something to be presented to putative reinsurers [55]. Mr Hylander’s memorandum of 7 July 1996 which the judge set out in [55] - [56] revealed certain important matters to which the judge drew attention in [60] – [65]. It showed that Arig expected a poor loss development on the risks which it had already written and expected to write in 1996, because it had been and would be unable to achieve deductibles as high as it wished. The reinsurance sought would give Arig “a good chance of making an overall profit” which without the benefit of such reinsurance it was much less likely to achieve, [61]. Second, the memorandum referred to Arig’s knowledge based on its own experience that energy construction business was highly exposed. [62]. Third, that loss experience was regarded by Mr Hylander as highly relevant to the decision to seek first loss protection [63]. Fourth, and most significantly, there was no suggestion by Mr Hylander that Arig’s past experience was irrelevant because a different underwriting strategy was being pursued. [64], Fifth the memorandum recognised that it would be very difficult to obtain the first loss protection which Arig sought [65].

In those circumstances, it was not likely that the hypothetical broke would have contained reference to the change of underwriting strategy. Nor was there any, or any sufficient, evidence that, in addition to the 1989 – 1995 statistics, which should have been revealed, Arig would have presented the risk to Albingia in the terms suggested or relied on the considerations that the judge took into account.

As to inducement the judge placed some reliance on the fact that Albingia had entered into a quota share treaty with Arig. However, under that form of treaty there was a sharing of risk, which there is not under a first loss treaty. Further the statistics presented in respect of underwriting years 1990 to 1995 showed loss ratios of less than 100%: see [18] above. And these ratios were reached after including two large losses to most of which the first loss treaty would not respond. Further 74.5% of the risks were said to be operating risks. That Mr Holzapfel was prepared to write that business does not tell one much about his preparedness to write first loss business on construction risks with the loss statistics with which he should have been presented.

The judge also referred to four first loss treaties which Arig entered into between late 1995 and the first treaty. They are (a) the AIG Oil Treaty [48]; (b) the Copping syndicate [49]; (c) the Atkin syndicate [50]; and (d) a further treaty with AIG [51]. Here the relevant loss ratios were markedly different to those of Arig.

In respect of treaty (a) the loss ratio for 1991, the most mature year, was only 33.2%. In respect of treaty (b) both the 1991 and 1992 years had been loss making, but this was explained as the consequence of two substantial losses in 1991 and one in 1992. If those losses were taken out of the reckoning insofar as they exceeded the first loss limit, the loss ratios on both years was around 30%. In relation to treaty (c) the brokers worked out the “as if” to the treaty terms and the loss ratio was an average of 2%. In relation to treaty (d), as the judge records, no historic loss statistics were provided but the risk was presented by NMB on the basis that Albingia already had a very good relationship with AIG and that the proposed treaty was very similar to the AIG oil treaty. The judge found that Mr Holzapfel took a 20% line subject to suitable excess of loss reinsurance probably because of the existing relationship with AIG. None of these statistics were anything like those from Arig.

In relation to the claim of procedural unfairness it is necessary to consider some of the history. In the original Particulars of Claim Axa had pleaded that Arig had failed to disclose material loss statistics and that, if it had done so, Mr Holzapfel would have declined the first treaty; and that, in the alternative, he would have asked for “as ifs” applying those loss statistics to the proposed treaty, and, having seen them, would then have declined the risk. I call these the “as if to the treaty” statistics.

The loss statistics were those in the voluntary particulars table. The as if to the treaty statistics were before the court. These were the 1989 to 1995 loss statistics (i.e., the premium and losses in relation to the years 1989 to 1995 as at the end of 1995) and the 1989 to 1996 loss statistics (i.e. the premium and losses in relation to the years 1989 to 1996 as at the end of 1996) if, in each case, the premiums and losses were those which would apply under the terms of the 1996 and 1997 treaties.

The 1995 figures are as follows:

Table 3 – the 1995 As If Calculations (i.e. the 1995 Loss Statistics applied to the 1996 Treaty on as “as if” basis)

In paragraphs 189 and 190 of his first witness statement Mr Holzapfel said:

“189.

Had ARIG's relevant historic loss record as at 31 December 1995 been included in the NMB offer fax 17 July 1996 then I would probably have instructed Ms Jerabek to decline the offer. The only other possibility is that I would have instructed Ms Jerabek to request further information from NMB, including relevant "as if" calculations specific to the proposed treaty, to see how the limits and premium amounts associated with the 1996 First Loss Treaty as proposed, affected the performance of the construction account. These "as-ifs" would include the application of the proposed premium scale to the historical account, taking account of any deductions to premium under the First Loss Treaty, and limiting the individual losses to the US $500,000 proposed treaty limit.

190.

Upon receipt of such "as-if" calculations (as set out in the table at 25A of AXA's Re-Re-Re-Amended Particulars of Claim), I would certainly have instructed Ms Jerabek to decline the 1996 First Loss Treaty or, if such figures were not provided until after my return to the office on 29 July 1996, I would have declined the offer myself.”

Arig’s first line of defence was that the hypothetical prudent underwriter would not have regarded any loss statistics as being material to the writing of the 1996 treaty: paragraph 27d of the Defence. In relation to the change in underwriting strategy it was not until its Re-Re-Amended Defence and Counterclaim in March 2015 (paragraph 27 e) that Arig pleaded that Mr Hylander had adopted a change in underwriting strategy, even though Mr Hylander and Mr Bader had been contacted by Arig and questioned in relation to these proceedings from the start. That averment formed part of the claim that there was no breach of the duty of disclosure because, in the light of the change of policy, there was nothing material to disclose.

As to inducement Axa was simply put to proof. No positive case was pleaded - in the alternative to a denial of any failure to disclose material information - as to what else a fair presentation would have included if the 1989 onwards loss statistics had been disclosed, or what else Arig/NMB would have said if they had been included. Mr Charles Kimmins QC, for Axa, submitted that, if it had been pleaded that a fair presentation would have included or been accompanied by a statement as to what the change of strategy amounted to, Axa might well have turned round and said that what it was claimed would have been said would, itself, have been a misrepresentation (either as to the past or as to Mr Hylander’s intentions for the future); and would, in any event, have wished to investigate and seek discovery in respect of the change and to consider what evidence there was of it.

As to the evidence Mr Stow said in paragraph 9 of his second witness statement

“9.

I do not think that an underwriter of Thomas Holzapfel's experience would have accepted the absence of loss statistics without a satisfactory explanation. He could, if he wished to, have requested statistics in respect of Arig's energy portfolio, notwithstanding that they would not have reflected Arig's underwriting philosophy going forward. If he had, we would have requested the historical statistics from Arig. However, I expect that Arig would have explained that the statistics related to risks accepted before Mr. Hylander took over as underwriter and changed Arig's underwriting philosophy. Arig would probably also have produced 'as-if' statistics showing the results of the construction portfolio 'as-if' the new underwriting philosophy had been in place during prior years. Of course, Thomas Holzapfel or any other reinsurers could have declined the risk if there were not satisfied with the placing information.”

That passage, which went to whether Albingia had waived any entitlement to see loss statistics, does not say anything about what Mr Stow, himself, would have done. Nor does it say that in the hypothetical broke he would have made some play of the fact (a) that the quota share was oversubscribed, amounting to a vote of confidence in Arig by the market; (b) that he would have emphasized that Mr Hylander had adopted an approach to energy risks which would have satisfied Mr Holzapfel; (c) that he would have emphasized that Albingia would receive some immediate cash flow; or (d) that he would have sought to explain away the 1989-1990 figures on the basis that the market in general had suffered adverse market conditions – although these are matters that the judge found would or could have been said in the hypothetical broke. Nor does one gain much from an expression of view on his part as to what Arig would have done.

In paragraph 28 of his first statement Mr Hylander said:

“28.

Since NMB had provided Albingia with the information that would be needed to decide whether to underwrite the 1996 Treaty I would have been surprised if it had also asked for historical loss statistics - Albingia already knew how many risks to expect, how much premium to expect and how many risks there would be. Had it done so, however, I would have provided them together with a clear warning that they were not reflective of my underwriting intentions going forward, which accorded with the information contained in the slip”

That passage, which appears again to be directed to the question of waiver, is, Mr Kimmins submits, the meagre extent of Mr Hylander’s evidence on what he might have said and does not cover the various points that the judge found in his hypothetical broke.

When Mr Holzapfel was cross examined Counsel for Arig did not put any statistics from 1989 to 1995 (as opposed to 1991 to 1995) to him. (We were told that by the time of the trial Arig did not have the information necessary to carry out an as if exercise in relation to 1989 and 1990 statistics). What was put to him was the Arig reconstruction as ifs with figures for “only risks of a type that would be ceded to the FLT”. This was the exercise that the judge characterised as of little value. The figures put to Mr Holzapfel contained a claims size and frequency analysis. It was in relation to these figures that Arig secured the favourable answer to which the judge referred in [152] of his judgment: see [43] above.

When it came to re-examination there occurred this exchange:

Counsel for AXA

“Q. I see. Could you go to {L1/13/182}? [This was one of the versions of the Arig reconstruction “as ifs” in relation to 1991-5 put to Mr H.] You were asked a number of questions about the loss ratios which appear here. I want to ask you a hypothetical question: assume that you were aware that there were in fact loss statistics for 1989 and 1990 which were in existence --

MR BRYAN: My Lord, we didn't ask any questions in cross-examination about 1989 and 1990, so it's not appropriate to examine those years, it's a matter that should have been in his witness statement if he can deal with 1989 and 1990.

  ……..

MR JUSTICE MALES: Just a moment. Do you accept that you haven't challenged paragraph 193 (Footnote: ²) of Mr Holzapfel's witness statement? {B1/1/29}

MR BRYAN: My Lord, we would have to look that up.

MR JUSTICE MALES: If you don't accept that, then I will allow the question.

MR BRYAN: My Lord, the question was asked in relation to 1989 and 1990, not 1991.

MR JUSTICE MALES: Yes, I know.

MR KIMMINS: It's entirely fair when --

MR JUSTICE MALES: I will allow the question.

MR KIMMINS: Thank you, My Lord.

Assume you had been told that there were in existence loss records for 1989 and 1990, would you have asked to see them?

A. Yes, of course.

Q. Why?

A. Because I wanted to see the performance, how the company managed the business over a longer period of time.

Q. I see.

A. And I think it's their duty to disclose whatever they have in terms of loss history.

Q. I see. The figures that we are looking at, could we go back to {L1/13/182}, are the figures on ARIG's inwards book, that is the position. Now I want to take you to one paragraph in your witness statement, please, which is {B1/1/28}, and I want to ask you about -- hold on, let me find the reference -- 189:

"Had ARIG's relevant historic loss [history] as at 31 December been included in the NMB fax… I would probably have asked Ms Jerabek to decline the offer. The only other possibility is that I would have instructed Ms Jerabek to request further information from NMB including relevant 'as-if' calculations specific to the proposed treaty, to see how the limits and premium amounts associated with the 1996 first loss treaty as proposed, affected the performance of the construction account. These 'as-ifs' would include the application of the proposed premium scale to the historical account, taking into account deductions…"

If we could have up on the screen, please. {A1/9/111}. Now, Mr Holzapfel, is this a document you have seen before?

This is Table 3

A. I do not recall.

Q. These are as-if calculations using the ARIG inwards book loss data and applying it to the treaty.

A. Mm. They are widely different from what Mr Hylander has calculated.

Q. Now, what I want to ask you is this: the documents you were shown were on ARIG's inward book, and let's look at 91 to 95, which is what appeared in the document you were shown, 91 to 95 on the document in the L bundle that we looked at, one saw certain percentages. If one looks at the as-if percentages for 91 and following, you can see they are 395, 239, 42, 0 and 0.

MR BRYAN: My Lord, none of these as-ifs were put into cross-examination.

MR JUSTICES MALES: No, they weren't.

MR KIMMINS: My Lord, the difficulty is this: what my learned friend did is cherry-pick a scenario of being shown 91 onwards. Now, I ought to be entitled to say to my witness "okay, fine, if you are going to be cross-examined on the position, were you shown 91 to 95 only, what would your reaction have been if you looked at those on an as-if basis?" I have to be able to put that in circumstances where it was only in cross-examination that my learned friend put the 91 to 95 figures in isolation to Mr Holzapfel.

MR JUSTICES MALES: Well, I don’t think this does arise out of cross-examination. It's simply not a point that Mr Blackwood went to, and you are left, therefore, with the submissions that you can make based on paragraph 189.” 

So, Mr Kimmins was making the point that Mr Holzapfel had not been cross examined by reference to the 1991 to 1995 figures on an as if to the treaty basis (as, indeed, he had not). And, on that basis, he was not allowed to re-examine on the 1989 – 1995 statistics.

At the very end of his evidence the judge asked Mr Holzapfel a series of questions in which he put forward what Mr Kimmins characterised as hypothetical broke mark 1. It is, in effect, summarised in paragraph [163] of the judgment. For ease of reference the questioning is set out in an appendix to this judgment. The loss records that it was suggested (hypothetically) would be shown to Mr Holzapfel are the Swire Fraser documents. They are D1/4/23 in the trial bundle which is the document which produces the figures in the table at [12] and D 1/4/27 which is a year-by-year breakdown of the figures for premiums and losses attributable to underwriting years 1989 – 1993. They are the two documents referred to in [13] above.

As I have already observed these figures differ from those in the voluntary particulars table in that they did not include the numbers of risks or the number of claims both of which, particularly the latter, are material. In addition, it is to be noted that the Swire Fraser figures came with Loss Ratios which showed for 1989 to 1993 the ratio of losses up to $ 5 million to net premiums. This produced the following ratios:

Mr Hartigan, Axa’s expert, was cross examined by reference to the 1991 – 1995 figures, and not the 1989-1995 ones. Even then he indicated that he would want to make sure that the risks of a type that supposedly would not be on the treaty would not in fact be ceded by requiring a list of exclusions. Mr Simon Bryan QC, Counsel for Arig made plain that, so far as he was concerned, what was being put did not include the 1989 and 1990 statistics.

In its closing submissions Arig invited the court to find that, if NMB had presented loss statistics to Albingia in July 1996, that would have involved production of loss ratios which were the same or similar to those in the 1991 – 1995 statistics shown to Mr Holzapfel in cross examination, and not the Swire Fraser statistics. The submissions also referred to the five-point question on inducement asked by the judge and said that “this hypothetical question is not in point” because the questions were put by reference to statistics which assumed a treaty limit of $ 5 million and assumed that the 1989 and 1990 statistics were relevant when they were not. No suggestion was made that what Mr Holzapfel said in paragraphs 189 and 190 in relation to his reaction if the 1989 -1995 statistics had been presented to him was wrong.

In their closing submissions Axa pointed out that Mr Holzapfel’s evidence in these paragraphs that he was “induced by the [non-disclosure of] the 1995 Statistics including the 1989 and 1990 years’ results was never challenged”.

As I have said, the judge indicated [152] that a fair presentation would have involved at least some “as if” statistics without indicating which ones those might be. The candidates include Arig’s reconstruction “as ifs”. Another candidate is table 3 which would seem unlikely. It is, also, in my view, not what the judge can have in mind, because it is not an exercise applying any change in underwriting strategy. A third is statistics with a higher premium or a lesser deductible such as Mr Bader’s “as ifs”. But examples of these were never put to Mr Holzapfel. On the whole it seems to me that the judge must have been contemplating some form of Arig reconstruction as ifs even though he thought these to be of very little value [106] and incapable of demonstrating in any concrete or quantifiable way what impact the new policy would have had on risks in place during the period from 1989 to 1992 [111].

In response to the submission that the constituents of the hypothetical broke and the considerations which informed the judge’s conclusions at paragraphs 172 ff were unsupported by evidence Mr Bryan submitted to us that the matters in question were all dealt with in the evidence and he directed us to passages in it.

Thus, the fact that Arig had already entered into four first loss treaties was established, as was the existing relationship between Arig and Albingia in the quota share treaty, which the judge described as “an obvious element of any broke” and the oversubscription of which represented a vote of confidence in Arig by the market, of which, the judge said, NMB would have been able to make some play. The judge also found, in the light of his evidence, that Mr Holzapfel knew that energy underwriters were moving away from operational risks in favour of construction risks, even though the latter were a higher risk: [166]. He was entitled to find, as he did, that the existing relationship would have been an important factor if a fair presentation of the risk had been made.

Similarly, the judge was entitled to regard it as significant that Mr Holzapfel had not made any inquiry about deductibles, despite the position that he had taken in relation to the Copping treaty: see [53] above. Mr Kimmins submitted that this missed the point that, in view of the statement “This is a new Treaty for the Reassured and as such does not have a corresponding loss record”, Mr Holzapfel had thought that there were no loss statistics to reveal, a position which the judge found understandable. But, as Mr Bryan pointed out, fifteen risks had already been written which were to be ceded to the first loss treaty and Mr Holzapfel could easily have inquired as to what the deductibles, if any, were on them or would be on future business. The judge was entitled to think that the fact that he did not was illustrative of how apparently categorical evidence given by Mr Holzapfel in good faith may be wrong. In addition the judge found it impossible to determine whether Mr Holzapfel in fact had that understanding at the time.

The judge was equally entitled to find that cash flow was a factor of some importance to Mr Holzapfel at the time, as is apparent from the contemporary documents which the judge saw, and which we have seen, and from passages in Mr Holzapfel’s oral evidence. He was entitled to find that cash flow would not have been a major factor but would have been of some benefit. Mr Kimmins observed that the cash flow from the already written risks to be ceded to the treaty would only be $ 230,000, of which Albingia would only get $ 115,000. But every little helps, especially when there are excess of loss premiums to be paid.

In my view, it was open to the judge to treat the fact that Albingia had already written two first loss treaties, that Mr Holzapfel made no inquiry about deductibles, and that Albingia was interested in cash flow as relevant factors in determining whether Mr Holzapfel would have accepted the risk. Mr Holzapfel’s response when the omission to ask about deductibles was pointed out to him was recorded in the judgment [168]:

“When asked about this apparently surprising omission, Mr Holzapfel said that he had to leave it to the ceding company to decide whether a risk was adequate in terms of pricing, conditions and deductibles. He added that with hindsight "it would have been nice" to have information about Arig's deductibles, but that "it was sort of a carte blanche to say, Yes, we want to support them" and was "a shot in the dark, insofar we had no, say, information about deductibles or whatever.”

That is evidence that, in relation to Arig, he was indeed prepared to take a shot in the dark in order to support his quota share partner.

That there would be some immediate cash flow from premiums on the treaties already written which were to be ceded to the first loss treaty was obvious. The judge found that NMB would have been able to point to the cash advantage. There is no evidence from Mr Stow of which I am aware in which he said that he would have done so but the judge was, it seems to me, entitled to infer that that is what Mr Stow is likely to have done; and, in any event, the benefit to cash flow was evident.

In relation to the seven matters set out in paragraphs [172] to [178] of the judgment - see [55] above - the first is a conclusion, which, in the light of his non-acceptance of Mr Holzapfel’s evidence about findings about deductibles and cash flow the judge was entitled to reach, about the safety of reliance on Mr Holzapfel’s assertion about what he would/would not have written in the various scenarios “about which he was asked”.

The second observation is based on the absence of any information about deductibles, which I have addressed above.

The third point is the significance to be derived from the fact that Mr Holzapfel wrote a 20% quota share of Arig. On the documentary and oral evidence before him the judge was, in my view, entitled to conclude that Mr Holzapfel regarded Arig as a high quality reinsured, that he wrote a quota share for that reason and that the fact that he did so was significant in determining whether he would have subscribed to the first treaty following a fair presentation. The effect of the quota share treaty was that Albingia was already sharing the fortunes of Arig which would extend to any construction risks it wrote. In addition, Mr Holzapfel indicated in his oral evidence that if Albingia supported Arig in the quota share treaty which was predominantly on production risks business it might just as well follow them on the construction business, i.e. the first loss treaty, just as it did with AIG on a much larger scale. A similar combination of quota share and first loss occurred with the Copping syndicate.

The fourth is a finding as to Mr Holzapfel’s fair mindedness such that he would have been willing to consider on its merits any explanation given to him of poor loss records. A finding of fair mindedness is not something of which Mr Holzapfel can complain and his willingness to consider any explanation given to him of its loss records was apparent from his evidence.

The fifth point is that NMB would have explained that the poor results achieved by Arig were on risks written for the most part by a former underwriter and that Mr Hylander had adopted a more rigorous and selective approach to energy risks in general which had satisfied Mr Holzapfel for the purpose of writing the quota share. Axa submits that this finding was not open to the judge. There was no reference to a change of strategy in the Swire Fraser broke; nor in Mr Hylander’s memo of 7 July 1996 in which he said that Arig would find it difficult to get cover. Mr Bader and Mr Stow did not discuss it when NMB tried to get cover either for the quota share or the first loss treaty; or, on the judge’s finding the change was only “slightly different”.

However, the judge had found in [111] that Mr Hylander had “sought to introduce a more rigorous and conservative underwriting policy than his predecessor and that he was successful in doing so” even though that it was not demonstrable in any concrete or quantifiable way what impact that policy would have had during the period from 1989 to 1992. In the light of that finding it seems to me that the judge was entitled to find that, if the first treaty fell to be broked with the 1989 statistics the likelihood was that reference would be made to the change of policy. The judge’s formulation of what would have been said (“Mr Hylander has adopted a more stringent and selective approach… so that quite a number of the risks which the previous underwriter had been prepared to accept would probably not have been accepted under the new approach”) must have represented the judge’s assessment of the appropriate limit of what could properly be said in the light of what change he found had come about.

I do not regard the judge as disabled from reaching a conclusion as to what would be a fair presentation (an objective question) by the absence of direct evidence from the broker about what he would have said or disabled from inferring what the broker would have said (in hypothetical circumstances) in the absence of such evidence. An inference can be drawn from surrounding circumstances. Further, evidence of what the broker did not say when making an unfair presentation is not necessarily a reliable guide to what he would have said or added when making a fair one. Limited though his evidence was, Mr Hylander’s statement was to the effect that if figures had been asked for they would have been given but he would have provided Axa with “a clear warning that they were not reflective of my underwriting intentions going forward, which accorded with the information contained in the slip”: see [83] above. In those circumstances the judge was entitled to find that, if loss statistics had been provided they would have been accompanied by a reference to the change of underwriting policy. That evidence was given in relation to whether Arig should have given historic loss statistics but it can be relied on as covering the situation where (in the hypothetical broke) they are in fact provided. The same goes for the evidence of Mr Stow to the effect that, if Mr Holzapfel had requested historical statistics, he would expect Arig to have explained that they related to risks accepted before the change of underwriting philosophy and that Arig would probably have produced as if statistics showing the results of the construction portfolio as if the new underwriting philosophy had been in place during prior years. Mr Stow confirmed to the judge that he himself had performed such an exercise in relation to other brokes.

The sixth point was that the particularly bad results achieved in the 1989 and 1990 years reflected adverse market conditions not comparable to the situation as it was believed to be in 1996.

As to the adverse conditions in the market in 1989 and 1990 there was no direct evidence from the brokers that they were aware of the fact that the condition of the market was very different in those years, or that they would have made that point in any broke or that Mr Holzapfel was aware of market conditions in those years. But 1989 and 1990 undoubtedly had adverse market conditions in the energy market and it was legitimate for the judge to infer that the contrast between those years and 1996 would have been drawn. Mr Outhwaite’s evidence was that he doubted that any underwriter was not hit by losses in 1989 and 1990.

The seventh point was that, although first loss reinsurance on energy construction risks was a particularly hazardous type of risk, Mr Holzapfel was not averse to running these risks. The fact that he had done so was proved in evidence and was a relevant factor.

For all those reasons the judge was left in doubt as to what Mr Holzapfel would have done if there had been a different presentation and such explanatory points as could have been made.

In the light of the material to which I have referred above, I accept that, subject to the point to which I am about to come, it was open to the judge to be left in the state of doubt in which he ended up, having regard to the factors to which he referred.

There remains, however, for consideration the question of procedural error.

In summary, Axa submits, the position stands thus. Axa pleaded in clear terms that the 1989 – 1995 loss statistics as set out in the voluntary particulars table should have been disclosed. The judge agreed [158]. Axa also pleaded that, if those statistics had been disclosed, Mr Holzapfel would have declined the risk, either when those statistics were disclosed or after he had been provided with the “as if” calculations specific to the proposed treaty. Mr Holzapfel, whom the judge found to be an honest man, gave evidence to that effect [189] – [190]. He was not cross examined by counsel for Arig by reference to the 1989 – 1995 figures and counsel for Arig did not challenge what he there said. The judge declined to allow any re-examination by defence in relation to “as if to the treaty” statistics for 1991-1995. The judge then put hypothetical broke mark 1, which is not the same as the voluntary table. His questions met with the clear response that Mr Holzapfel would have declined the risk. He was not questioned or challenged as to his answer; nor was it suggested to him that what he said was difficult to understand or wrong or misguided. In Arig’s final submissions, it was not suggested that what Mr Holzapfel said about his response to the 1989 – 1995 statistics was wrong; but that the judge’s five stage questioning was “not to the point”.

Nevertheless, the judge said he was left in doubt as to what Mr Holzapfel would have done if a fair presentation had been made to him.

The sequence of events set out in the previous paragraph amounts, Axa submits, to a serious irregularity. Axa succeeded in establishing that the 1989 – 1995 statistics should have been disclosed but failed on the question of inducement in circumstances where Arig did not challenge Mr Holzapfel when he gave his evidence as to what would have happened if those statistics had been disclosed. That evidence was not particularly surprising given the latter figures showed large loss ratios for 1992 even after Mr Hylander was in charge. The table below compares the loss ratios in the voluntary particulars table with table 3:

What the judge said was this:

“160 In his first witness statement Mr Holzapfel said that if Arig's historic loss record as at 31 December 1995 had been included in the NMB offer fax of 17 July 1996, he would probably have instructed Ms Jerabek to decline the offer. That evidence is not difficult to accept in isolation, but it assumes that the figures were presented without explanation, which is not the relevant scenario. However, his witness statement went on to say that if Arig had told him that it had changed underwriter in 1991 and that the new underwriter wrote a different kind of book, this would not have made any difference: he would still have declined the treaty irrespective of this explanation. In his second statement Mr Holzapfel insisted that such an explanation would not have affected his usual approach to assessing risk, which "always involved a careful assessment, based on the documents presented to me by the broker, of the profitability of the account over time". This, however, was an over statement. While I accept that examination of loss statistics was Mr Holzapfel's usual practice, it was not invariably so. There is at least one case in evidence where he did not do so, which was a case where Albingia already had an existing relationship with the reinsured (AIG: see [51] above).”

But the judge did not refer to Mr Holzapfel’s evidence that had he not told Ms Jerabek to decline the risk immediately, the only other possibility was that he would have asked for the as if to the first treaty figures and, having received them, would certainly have refused the risk. The figures in Table 3, which would appear to show something significantly wrong even in 1992, a year with more development to come, were neither referred to nor addressed.

Mr Bryan submitted that this point is misplaced. It is necessary to look at what Mr Holzapfel said when cross examined because it is apparent that what he said orally was not consistent with what he was saying in his written statement. He was asked what would be the position if it had been explained to him that there had been a change of underwriter and in underwriting philosophy and a simplified version of the Arig as if reconstruction figures had been produced; and it was suggested that he would have written the first treaty in the same terms. To that he replied:

“No. I don’t think so. Preparing statistics with the benefit of hindsight is not acceptable. I want to see the full picture and then we can go into details, but show the full picture first, please. Put facts on the table and then we can say what are they doing to implement, how are they going to do this, do another as if exercise and whatever but not on the basis of as if.”

A little after that he was asked to assume that he was shown a table which showed the actual loss ratios for 1991 onwards of 109%, 65% and 12% (i.e. as in the voluntary particulars table), received an explanation of the change of underwriting strategy, and was told of the difference between those loss ratios and the lower ones produced by the as if calculations. On those assumptions, he was asked whether he would have written the business and he said that he would.

It is not wholly clear to me to exactly what “full picture”/“facts on the table” Mr Holzapfel was referring in the answer that I have quoted in [117] above. In particular it is not clear whether he was referring to as if to the treaty statistics: but, if he was, his subsequent acceptance that he would have written the treaty given the explanation for the change of underwriting philosophy would indicate that he would not have declined the risk on account of the absence of such statistics.

Axa’s case on procedural error was forcefully and attractively presented by Mr Kimmins. But I am not persuaded that it should carry the day. The judge was plainly aware of paragraphs 189 and 190 of Mr Holzapfel’s statement. He referred to paragraph 189 specifically when he cut short Mr Kimmins’ re-examination and, as appears from what he said, expected submissions to be made on the basis of it (as they were). It is that paragraph to which he refers at [160] of his judgment: see [115] above. Whilst he does not refer to the possibility [sic] of Mr Holzapfel asking for as if to the treaty figures one of the prime bases of his decision was that categorical statements by Mr Holzapfel as to whether or not he would have written the risk in the various scenarios about which he was asked could not safely be relied on [172].

This was apparent from the fact that, although Mr Holzapfel said that his approach to assessing risk “always involved a careful assessment based on the documents presented to me by the broker of the profitability of the account over time”, this, the judge found, was an overstatement and was not his invariable practice: [160]. It was also apparent from the contrast between what he said about the critical importance of deductibles in relation to avoidance of the Copping treaty and his absence of inquiry of Arig about deductibles on its risks that “apparently categorical evidence given in good faith may be, if not dangerously wrong at least not the whole story” and that was, as the judge found, something which could be equally true in relation to loss statistics [169]. Further Mr Holzapfel had said that cash flow was not an important factor whereas the contemporary documents showed that it was, as the judge put it [170], “indeed a factor of some importance”.

It was in that context that the judge approached the question of what Mr Holzapfel would have done with the hypothetical broke. He first considered Mr Holzapfel’s evidence that, if in addition to being given the actual figures going back to 1991 he had been told that Mr Hylander had taken over in 1991, and that he had applied a new strategy which if applied to the 1991 and 1992 years would have produced loss ratios below 30%, he would probably have written the risk on the same terms [162].

The significance of Mr Holzapfel’s answer was that he would have been influenced by an explanation of the change of policy such that he probably would have written the risk on the same terms, despite what he had originally said in his witness statement which was that he would still have declined a treaty irrespective of any explanation about the change of underwriter and underwriting policy.

It was in the light of that that the judge addressed the question of what would have happened if Mr Holzapfel had been shown the 1989-1995 figures and had received the explanation of the change of underwriter to which he referred at [163]. Given:

I do not regard the judge as having been precluded from reaching that conclusion by the way in which Arig conducted its case. Arig did not run a case based on the statistics from 1989. That meant that there was no cross examination from its counsel as to what would have happened if the 1989-1995 figures had been produced. Here such “cross examination” as there was came from the judge and involved no reference to the possibility of as if to the treaty statistics or consideration of Table 3.

However, in circumstances where:

Mr Holzapfel’s written evidence had been that, presented with the loss statistics for 1989 onwards, he would probably have instructed Ms Jerabek to decline the offer and that the only other possibility was that he would asked for as if to the treaty statistics and would then have declined the offer. Thus his primary position was that he would simply have declined the offer without more ado. It is noticeable, also, that, when the different scenarios were put to him by counsel and by the judge he did not refer, at any rate with any clarity, to the need for as if to the treaty statistics.

The judge plainly thought that there was a very real possibility that Mr Holzapfel would not have made a careful assessment of the statistics at all such that he was left in doubt as to what Mr Holzapfel would have done. He was, also, entitled to express a want of understanding as to why, if the statistics started with 1989 the position would have been the opposite of what it would have been if they started in 1991.

As the judge put it:

“Mr Holzapfel's answer was that he did not think that he would have entered into a relationship with Arig on the first loss treaty, because Arig had demonstrated that it was not able to manage the business properly over time. The scenario was that there had been a period of poor underwriting decisions with an underwriter following what turned out to be a bad policy, but that the situation had been managed by the introduction of a new underwriter who was successfully implementing a more rigorous approach”

The judge cannot have meant that he did not understand what Mr Holzapfel was saying rather than that he was not persuaded that what he was saying was an accurate representation of what would have been his attitude to the hypothetical broke suggested to him. That his reaction would have been as he described was not without foundation. Mr Holzafpel said repeatedly in his evidence that he would have looked to see how well the company was managed and he could well have been concerned about how Arig dealt with underwriting after Mr Simenstand fled Bahrain. But the judge’s conclusion was, in my view, one that was open to him. He was entitled to think that the matter of principal significance to Mr Holzapfel would have been the change of underwriter/underwriting policy rather than the previous history, particularly when he found [155] that Mr Holzapfel knew or should have known that 1989 and 1990 had been disastrous years for the energy market generally. That tallied with the fact that Mr Holzapfel had said that he would have taken into account and been influenced by what he was told about the new underwriter and would have written the risk on the same terms if presented with the 1991-1995 statistics.

Nor am I persuaded that the procedure has gone so awry that we must set aside the judgment. The critical question is whether Axa have suffered such an injustice that we must set aside the decision. The procedural irregularity which is sought to be relied on is, in essence, twofold: (1) that the judge declined to allow Mr Kimmins’ re-examination; and (2) that the judge did not accept paragraphs 189 and 190 when, in relation to the possibility of a request for as if to the treaty statistics, these were not formally challenged.

As to the former, I am not persuaded that the judge was not entitled to restrict the re-examination intended, which was based in part on leading the witness by reference to his evidence in chief in respect of a scenario on which he had not been cross examined.

As to the latter, the judge was concerned with events which had happened over 20 years ago, of which Mr Holzapfel had no recollection, and with what he would have done in hypothetical circumstances. I quite accept that if it is suggested that a witness is telling an untruth as to what he says did or did not happen, it is incumbent on counsel to put that to him and, if he does not, his client may find himself bound by the evidence which the witness has given. The position is not the same when the court is concerned with a parallel universe in which the question is what would have happened in hypothetical circumstances 20 years ago where the underwriter concerned has no recollection of the actual broke and where the onus of proof is on the reinsurer.

The judge gave Mr Holzapfel the opportunity to deal with his hypothetical broke and it was Mr Holzapfel’s answers to the judge, when taken with all the other considerations to which the judge legitimately referred, that were of most significance in contributing to the judge’s doubt. I do not regard the fact that Mr Holzapfel was not specifically challenged in relation to possible as if to the treaty statistics as justifying or requiring an allowance of the appeal or rendering that scepticism unwarranted or unfair. Nor do I regard the fact that what the judge referred to in his cross examination was the Swire Fraser statistics as invalidating his conclusion on account of the difference between them and the voluntary particulars table.

In short, I am not persuaded that, in deciding that the onus of proof had not been satisfied the judge was plainly wrong or reached a decision on no evidential basis or one to which he was not entitled to come. In the light of all the considerations to which he referred, it was open to him to be left in doubt. Such scepticism was not, in my view, unfair either to Mr Holzapfel or Axa.

For these reasons, I would dismiss the appeal.

The facts of this case raise a question as to what needs to be pleaded or put in evidence, and by whom, in inducement cases. In the usual case the insurer will plead that there has been a material non-disclosure because some identified matter has not been disclosed; that he was induced to enter into the insurance on account of the non-disclosure; and that he has validly avoided it. The insured will not admit that the insured was induced to accept the risk by the non-disclosure.

It is for the insurer to prove inducement. That has to be determined by reference to what a fair presentation would require to be disclosed (which may be more than the inclusion of that fact the non-disclosure of which was complained of by the insurer) and is, in my view an objective question. In this hypothetical world, the question is what a reasonably prudent underwriter would think needed to be disclosed in order to give a fair presentation of the risk. In addition, there arises for consideration what additional matters the insured or his broker would have urged upon the insurer as reasons for writing the business. The dividing line between what a fair presentation required and what additional matters the broker would have urged may be difficult to determine and might be important since in determining the issue of inducement it is necessary to assume a fair presentation and ask what would have happened in the light of it, whereas in relation to additional matters which would have been raised by the brokers there would, as it seems to me, be at least an evidential burden on the insured to show that they would probably have been raised on the hypothetical broke. This conclusion seems to me consistent with the decision of this court in Drake Insurance v Provident Insurance [2002] EWCA Civ 1834. [62]; [74] [135].

What seems to me undesirable is that the insurer /reinsurer should be faced at trial, without prior notice either in a pleading or in any witness statement, of what the insured/reinsured, if wrong in contending that there was no non-disclosure, says, in the alternative, would have been the content of a fair presentation or would have been raised by the insured/reinsured or his brokers in any broke in which the matter which constituted the material non-disclosure relied on was in fact disclosed. If the matter is raised for the first time in cross examination (“If this statistic had been revealed and you had been told this, you would have written the risk, wouldn’t you?)” it may provide a good example of cross examination as an art form. But it involves the insurer/reinsurer coming to trial without notice of the hypothetical factual case that he has to meet and being required to answer on the hoof a question which on a presentation in the real world would not require so instant a response.

I agree.

I agree.

I don’t think I would have entered into a relationship with ARIG on that contract, simply because I did not, would not feel it necessary to give them a second chance on – because the company has demonstrated, not the underwriter, the company has demonstrated that they are not able to manage this business properly over time. If they allow an underwriter to do things like this, there is no controls, and just because it’s being said there is a new underwriter, and maybe I am even adding point 6 to that, I am being shown the new underwriting guidelines, I mean, we have an expression in Germany saying “paper is very patient”, I mean, you can put down the best intentions into an underwriting guideline… no I don’t think I would have supported it.