HLB Kidsons (A Firm) v Lloyd's Underwriters Subscribing To Lloyd's Policy No 621/PK1D00101 & Ors

This litigation is, as the judge, Mrs Justice Gloster, remarked, a contest about notification. It arises out of a professional indemnity insurance policy, made on a “claims made” basis, incepting on 1 May 2001 and terminating on 30 April 2002 (the “policy”), under which the claimant, here the primary appellant, a firm of chartered accountants, formerly known as HLB Kidsons, was the insured (“Kidsons”). As of 1 April 2002, shortly before the expiry of the policy, Kidsons merged with another firm of chartered accountants, Baker Tilly, and the merged entity took the name of Baker Tilly.

There are in fact three policies and five defendant underwriters, now respondents to this appeal (the “underwriters”). Nothing turns on any distinction between the policies. One is a Lloyd’s policy 621PK1D00101 subscribed to by lead syndicate 839, second lead syndicate 683, and a following market of other syndicates (collectively the first defendants and here the first respondents). The second is an IUA policy to which the Underwriter Insurance Company Limited and Royal and Sun Alliance Insurance plc subscribed, respectively the second and third defendants and here the second and third respondents. The third was a Companies Collective Policy to which International Insurance Company of Hannover Limited and Great Lakes Reinsurance (UK) plc subscribed, respectively the fourth and fifth defendants and here the fourth and fifth respondents. The two company policies incorporated the terms of the Lloyd’s policy.

Apart from Kidsons and the underwriters, there are two other parties to this appeal. One is Millers Professional Risks Limited, who were Kidsons’ insurance brokers and are now called Millers Insurance Services Limited (“Millers”). Millers are seventh defendants, but in this appeal appellants alongside Kidsons as third appellants. The other is CMS Cameron McKenna, a firm of solicitors which provided Kidsons with a claims handling service (“Camerons”). Camerons are sued as sixth defendants, but here appear, like Millers, as (second) appellants alongside Kidsons.

The essence of a “claims made” policy is that it provides cover for claims first brought against the assured during the policy year. As is common with such policies, however, it extends cover to claims first brought against the assured after the policy year provided such claims arise out of circumstances previously notified to the insurers of which the assured became aware during the policy year. The rationale of such policies has been described in J Rothschild Assurance Plc v. Collyear [1999] 1 Lloyd’s Rep IR 6 at 22 and in Friends Provident Life and Pensions v. Sirius [2005] 1 Lloyd’s Rep IR 135 at 142.

In the present case the clause in question by which this extension of cover was granted was General Condition 4 (“GC4”) which provided as follows:

“The Assured shall give to the Underwriters notice in writing as soon as practicable of any circumstance of which they shall become aware during the period specified in the Schedule which may give rise to a loss or claim against them. Such notice having been given any loss or claim to which that circumstance has given rise which is subsequently made after the expiration of the period specified in the Schedule shall be deemed for the purpose of this Insurance to have been made during the subsistence hereof.”

GC4 is central to the issues in this litigation, in which Kidsons sue underwriters for a declaration that they are entitled to be indemnified in respect of claims which have been made against them, after the expiry of the policy period, by clients in relation to the activities of Solutions @ Fiscal Innovation Limited (“S@FI”), a company formerly owned and managed by Kidsons, which marketed tax avoidance schemes. S@FI’s work was known as “fiscal engineering”. The essence of these claims, or many of them, is said to be that the clients had negligently been advised to enter into flawed schemes (“mis-selling” claims). The essential issue in this appeal is whether Kidsons gave to the underwriters effective notification of circumstances for the purposes of GC4 such as might cover such subsequent claims or some of them. Kidsons submit that they did, and that such notification embraced the full extent of concerns relating to the whole of S@FI’s fiscal engineering work, both as regards the tax schemes marketed and as regards the implementation of such schemes, which had been voiced by a tax manager employed in Kidsons’ Edinburgh office by the name of Mr Iain Torrance.

In 2001 Mr Torrance persevered in forcing to the attention of the S@FI board and of Kidsons’ national executive committee (“NEC”) his concerns that the tax avoidance schemes or “products” marketed by S@FI were fundamentally flawed and that the administrative procedures by which they were implemented were also liable to fall foul of the Inland Revenue. His strongly expressed criticisms, summarised in his memoranda of 23 and 24 August 2001 and elsewhere (eg “recipe for disaster”…“DOS, SHEP Selling, the CRC Scheme and the Conditional Share Award scheme…all represent unacceptable tax avoidance…It represents an assault on the Treasury…”), were considered at a NEC meeting on 29 August and at a specially convened S@FI board meeting on 30 August. As a result, the NEC resolved to set up an independent review body (or “IRB”) to investigate and review Mr Torrance’s complaints, starting with a product known as the Discounted Option Scheme (or “DOS”), about which Mr Torrance had a particular concern, especially relating to its implementation. The S@FI board concurred with that decision and also resolved to suspend the marketing of DOS. The S@FI board also resolved to give notice of circumstances to its insurers. The board minutes recorded:

“In regard to professional indemnity, circumstances existed that might lead to a claim, particularly in relation to the discounted option schemes, and a notification should be made to PII underwriters”.

In the immediate run-up to these meetings Mr Colin Tyre QC, Scottish tax counsel, had been consulted on 20 August 2001 in relation to DOS and in particular its implementation: two such cases had been put before him. His advice broadly supported Mr Torrance’s views on the problematic residency of the IOM companies which were set up as part of the scheme and thus of the returns and disclosures to be made to the Inland Revenue if as a result their residency was to be regarded as being in the UK. On 27 August Mr Tyre prepared a written opinion regarding the advice he had given in consultation. His opinion was circulated to the NEC on 30 August. It caused great consternation. He said that if full disclosure was not made, there was a risk that the Inland Revenue would take criminal action. He also opined that the merits of the scheme were open to challenge by the Inland Revenue.

In the event a letter was drafted by Mr T J Patten, Kidsons’ national partnership secretary, in collaboration with Mr Flaxman, an independent insurance consultant who regularly advised Kidsons on professional indemnity matters. This letter, dated 31 August 2001, is at the heart of the litigation and of this appeal. It was addressed to Mr Marcus Elwes who was a placing broker at Millers. Kidsons’ claims broker at Millers was Mr Steve Salter, who will figure later.

The letter (the “31 August letter”) read as follows:

“S@FI Limited

You will recall that the fiscal engineering activity of the practice has been channelled through S@FI, which is manned in entirety by partners and staff of HLB Kidsons. Fiscal engineering work has been developed significantly over the last year or two and now forms about 7% of the turnover of the practice.

The products marketed by S@FI have all been validated by virtue of Counsel’s opinion (in some cases two opinions) but a tax manager in Edinburgh, Iain Torrance, has expressed the view that the Inland Revenue, if minded, could be critical of some procedures followed in certain cases.

The Board of S@FI and the National Executive Committee of HLBK intend to investigate this view fully and have approached Ray Armstrong, who I gather has been a senior Inland Revenue official and has retired as a partner in PWC, to invite him to carry out the investigation and submit a report.

The Board has taken the view that this might be regarded as material information for insurers. There is no sign of a claim arising at the present time but the Board feels that it is appropriate in the circumstances to advise what is happening and to take your instructions.”

It was submitted at trial by Mr Gavin Kealey QC on behalf of the underwriters, and accepted by the judge, that there is no express mention in that letter –

“(a)

of any possible issue about the validity or essential technical efficiency of [S@FI’s] products; or (b) of any possible issue about the fundamental viability of the whole range of S@FI products; or (c) of any possible issue about the integrity or competence of S@FI senior personnel; or (d) of any possible issue about the viability of S@FI going forward; or (e) of any possible issue about mis-selling, or false representations to clients; or (g) of any possible issue falling outside the scope of procedures of which the Inland Revenue might, if minded, be critical” (judgment at para 201(iii)).

On the other hand, as is inherent in item (g) of the above submission, Mr Kealey made it clear at trial that he accepted that the letter did address Mr Torrance’s concern about implementation procedures: in the language of the letter, “but a tax manager in Edinburgh, Iain Torrance, has expressed the view that the Inland Revenue, if minded, could be critical of some procedures followed in certain cases”. Mr Kealey accepted that that language covered implementation procedures not only in the case of DOS but also in the case of other products. The judge, however, in accepting Mr Kealey’s submission in item (g) did so in the following terms (judgment at para 205(viii)):

“Thus the recipient was not being told…(g) that there was any possible issue falling outside the scope of some wholly unidentified procedures, relating to wholly unidentified products, of which the Inland Revenue might, if minded, be critical, in some unidentified cases.”

The 31 August letter was not submitted by Millers to the underwriters at that time. It is not quite clear why that was so, as Mr Elwes’s reply to Mr Patten dated 4 September had said (responding to the invitation in Mr Patten’s last paragraph “to take your instructions”) that “ I suggest that we get this noted by insurers for information purposes…I will confirm back once we have done this.” The letter was not however presented until 27 September 2001, when the first of what it will be convenient to describe as four presentations occurred.

I will briefly describe the four presentations below, but first it is convenient to refer to a further letter written by Mr Patten, this time to Camerons, dated 28 March 2002 (the “28 March letter”). This letter was used in the third and fourth presentations, which took place in April and July 2002 respectively. The letter read as follows:

“S@FI Limited

Some months have passed since we last corresponded. We have put a lot of effort into a technical investigation of the sale of products with the intention of having a report prepared by the Independent Review Body under Ray Armstrong referred to in my letter of 31 August 2001.

This work has been slowed to a certain extent because of health problems suffered by Ray Armstrong and his ability to carry on leading the investigations is now in question. There is likely to be a further delay in the production of the report.

A meeting was held on Tuesday 26 March 2002 with Colin Tyre QC who had raised observations on two transactions concerning Discounted Option Schemes. The result of the meeting was a general view that the technical efficiency of the products was accepted but in some instances there might be procedural difficulties involving the Trustees for each scheme affecting the implementation of the scheme and this might lead to the possibility of criticism in the future.

Mr Garner-Jones, a Tax Partner in our Chester office and one of the Managing Partners for S@FI, was present at this meeting and will very shortly be producing a report to summarise the results of the meeting and the general activity in this area over the last few months.”

It will be observed that at the centre of this letter is the observation that there might be problems about the implementation of DOS.

The four presentations to which I have referred were as follows:

The first presentation or September 2001 presentation was solely of the 31 August letter by itself. It was made on 27 September by Mr Stephen May, a broking director at Millers, in person to Mr Christopher Day, a placing underwriter for syndicate 839, the lead underwriter. As the judge found, this presentation was made “in a placing context”. Mr Day scratched the letter: “WP – Please keep uwrs fully advised. Noted for information only.” Mr Day did not understand the letter to be a notification of circumstances. He thought it was providing information which Kidsons thought might be material to the risk. Had he thought the letter to be attempting to notify a circumstance, he would have made it clear that the broker needed to advise the syndicate’s claims department. The judge regarded his reaction as that of an objectively reasonable underwriter in his position. The “wp” scratch was a standard means by which underwriters reserved their position, and it was so understood by Mr Elwes who reported back to Mr Patten at Kidsons in a letter of 5 October as follows:

“I refer to your letter of 31^st August 2001 and confirm that we have seen insurers on your behalf.

They have noted the information on a strictly without prejudice basis and have asked that they be kept fully advised. The details provided have been noted for information purposes only.

I would suggest that once Ray Armstrong’s investigation report has been completed that a copy is forwarded to us to show insurers further.”

The judge found that this was not an effective notification of a circumstance for the purposes of GC4. Although Kidsons had relied on it as such, Millers had accepted in their pleadings that it was not. The placing context was wholly inconsistent with the notion that Kidsons were attempting to notify circumstances which might give rise to a claim, and the anaemic language of the letter did nothing to alert underwriters otherwise. In different circumstances, the presentation of a notice of circumstances on the placing side might not necessarily be fatal to the effectiveness of such a notice: see Friends Provident [2005] 1 Lloyd’s Rep IR 45, 135 (CA). But there is no appeal from the negative finding of the judge as to this first presentation.

By the time of the first presentation on 27 September Mr Armstrong’s IRB initial report had already been presented in draft, albeit not yet formally considered by the NEC. As it transpired, the IRB’s work did not proceed much beyond the initial draft report, and that draft was never itself formally finalised. It appears to have satisfied Kidsons (more or less, there is some uncertainty at the margin) that, apart from concern about the way in which DOS might have been implemented in the past, there was nothing to worry about. This is although the draft initial report was itself essentially confined to DOS. The briefing memorandum which the NEC had prepared for the IRB had said that the IRB was to conduct an “in depth and immediate review…of all the tax planning business undertaken…since 1 May 1999” and was to identify “issues and potential issues which could prejudice [S@FI’s] or the partnership’s operations of tax planning…”. However, the memorandum also distinguished between an “Initial Review”, defined as an “early report covering particularly Discounted Option Schemes, disclosures, actions and any other critical recommendations” and a “Further Review”, defined as a “more comprehensive review of product range, disclosures, control procedures and management”. The initial review was to be completed by 17 September, but if no serious areas of concern then remained, the IRB was to advise the NEC so that closure of the IRB’s duties could then be determined.

The initial review was submitted in draft on 14 September 2001. It considered DOS in detail and concluded that the scheme was not fundamentally flawed, that there was no need to make blanket disclosure to the Inland Revenue, that an issue of the residency of Isle of Man companies set up for the purposes of the scheme should be resolved urgently given that if they were in fact resident in the UK returns would have to be filed, and that a case by case review of individual DOS transactions should be carried out to ensure that they had been implemented correctly. This would involve examining the facts of each case to determine the residency of the IOM companies and the returns and disclosure for each case. On the key issue of criminality which had been raised by Mr Torrance (and by Mr Tyre), the IRB concluded that there was no evidence that any criminal activity had taken place. Outside DOS, the report noted that in due course it might be necessary to review the entire portfolio of S@FI products and warned that “there are several operational issues emerging which, if not addressed, could have an adverse effect on other products promoted by S@FI either now or in the future”.

The draft report was treated by the NEC and S@FI as a final report. It was considered by the S@FI board on 20 September and by the NEC on 4 October. Both accepted it and regarded it as substantially supportive. Mr Torrance came in for criticism. The NEC minutes recorded:

“Products: The only product with which there was now a problem was the Discounted Option Scheme, given that the Capital Redemption Scheme had ceased in the previous November…”

The case by case review of DOS was in due course carried out over the next two years, culminating in a report dated 10 September 2003. That review was not assigned to the IRB but was carried out under the supervision of Mr Garner-Jones, a director of S@FI and a member of the NEC. The only further instruction given to IRB was a request made in February 2002 to consider the Second Hand Endowment Policies Scheme (or “SHEPS”). However, that review never went ahead because Mr Armstrong became ill. No other product was reviewed.

I have summarised findings of the judge made at paras 137 and 141/2 of her judgment. The judge analysed and put the matter in this way:

“138.

Although the briefing memorandum for the IRB expressly referred to all S@FI schemes, and the possibility of a further review into such schemes, the reality was that, at that time, the mandate given to the IRB was merely to investigate Discounted Option Schemes…Thus whether or not further schemes were to be reviewed, depended on what, if anything, came out of the Initial Review, and whether it became necessary to conduct a further review. This was consistent with Kidson’s low level of concern about Mr Torrance’s wider allegations. However, one of the reasons for leaving open the stated scope of IRB’s work was no doubt the need for Kidsons to be able to show the Inland Revenue, if necessary, that they had taken all reasonable steps to put its house in order, which was a point that Mr Torrance had raised…However, in my judgment, that fact that the scope of the investigation was left open, does not demonstrate that, contrary to my conclusions, Kidsons had any real or pressing concerns at this stage about the validity or implementation of the wider range of S@FI products, other than Discounted Option Schemes. Indeed, the evidence showed that, apart from an aborted attempt to review SHEPS in February 2002, no instructions were given to the IRB to undertake any Further Review into any S@FI products and the IRB did not in fact get beyond looking at Discounted Option Schemes…

143.

I agree with Mr Kealey’s analysis of the evidence that, after 14 September 2001, Kidsons regarded the IRB’s Draft Report as effectively endorsing its methodology and approach and that, subject to the outstanding review of individual Discounted Option Schemes cases, its house was in order. The IRB was effectively disbanded in about February or March 2002, when Mr Armstrong indicated that he was too ill to proceed with the review of SHEPS. This was confirmed by the subsequent Project Island Report prepared by Baker Tilly in March 2004 which stated “It should be noted that the IRB having concluded its review and prepared the draft report in September 2001, was disbanded”.”

These facts set the scene, at any rate internally to Kidsons, at the time of the second presentation, which was made on 17 and 18 October 2001. This is the presentation which is at the heart of the appeal. It arose out of a letter dated 5 October (the day after the NEC had considered and approved the IRB’s draft report) which Mr Patten of Kidsons wrote to Ms Sarah Turnbull of Camerons. In previous years, but not in the policy year with which we are concerned, Camerons had operated under a claims handling agreement whereby presentation of a claim to Camerons operated as a presentation to all insurers on risk, a matter built into previous policies under General Condition 12, which, however, did not survive into the policy in question. Mr Patten, however, appears to have overlooked this change, and indeed he did not see a copy of that policy until September 2002, after the policy year had expired. In previous years Kidsons generally addressed and sent notifications of claims and circumstances directly to Camerons not to Millers. The letter read in material part as follows:

“Claims Handling with effect 1 May 2001

…there are five additional cases, some of which pre-date the finalisation of the claims handling agreement and a couple are recent. I am enclosing a set of papers for each of the additional five. The details are as follows:

…

K101/009 S@FI Limited”.

The “set of papers” in relation to S@FI included the unscratched version of the 31 August letter and Millers’ two responses to it, ie Mr Elwes’ letters of 4 September and 5 October 2001 (see paras 13 and 17 above). This last letter was stamped as received by Kidsons on 8 October, which suggests that Mr Patten’s letter of 5 October to Ms Turnbull was misdated. Athough the letter was addressed to Ms Turnbull at Camerons, it appears to have been sent instead, as Mr Patten would accept by mistake, to Mr Salter, the head of the claims team at Millers. Mr Salter forwarded the letter and its contents to Ms Turnbull on 10 October, after first making copies for himself. He considered that they constituted notification of claims and/or circumstances in relation to each of the five matters and therefore proceeded to make up individual claims files for each of them and included them in the 2001 bordereaux file.

On 17 October a colleague of Mr Salter (Mr Robin Cundall) presented this material to Mr Matthew Howes, a claims examiner for syndicate 839. The S@FI claims file then presented to Mr Howes consisted of a Millers’ cover sheet which recorded (inter alia) the date of loss as “31/08/2001 notified” and loss details as “S@FI Ltd. K101-009”; the 31 August letter; Millers’ two responses to Kidsons dated 4 September and 5 October; and Kidsons’ 5 October letter to Camerons. The 2001 bordereaux file contained a bordereau sheet which, after identifying the relevant policy, read in material part as follows:

It is uncertain whether the copy of the 31 August letter then presented to Mr Howes was in its original unscratched state or was in the version scratched by Mr Day. Since, however, Mr Elwes’ letter of 5 October, reporting on Mr Day’s reaction, was in the claims file, the judge considered that it was unnecessary to resolve that uncertainty.

Mr Howes scratched Kidsons’ letter dated 5 October 2001 and the bordereau without comment: he saw no need to take any specific action, since Mr Day had already reserved the underwriters’ position. In cross-examination he said: “What I agree is that the insured were purporting to notify a circumstance irrespective of whether it had been shown to the placing side first or a claims examiner first” (F2.626); and on another occasion he answered “Agreed, yes” to the question “you must have thought that the letter of 31^st August amounted to a circumstance because otherwise you would have said: we do not accept this as a circumstance?” (F2.628). The judge’s comment on this evidence was this (at para 195):

“All he was actually accepting was that he appreciated that an attempt was being made to notify a circumstance, but the rights and wrongs of that were being left for another day. In any event, as all counsel agreed, his views as to whether a circumstance had been notified are not relevant.”

I believe that it would have been more accurate to say that counsel agreed that his views could not be determinative: but Kidsons submitted that in this case, as in other cases (such as Mr Elwes’s original reaction to the 31 August letter and Mr Salter’s different reaction to the material sent to him under cover of Mr Patten’s letter of 5 October addressed to Ms Turnbull, as well as Ms Turnbull’s reaction to that material, to which I will come below), the reaction of professional people in the industry at the time was relevant to the objective question of construction of the documents in issue albeit that was ultimately for the judge to decide. Thus the judge had previously commented that the reaction of Mr Elwes to receipt of the 31 August letter “was that of an objectively reasonable broker in his position”: his evidence had been that the letter was not understood by him to be notification of a circumstance, but that had the letter been addressed to Mr Salter, he would have seen it in a claims context and might have viewed it differently (at para 186). Similarly, the judge had referred to the fact that Mr Day, who had originally received the 31 August letter at Millers, “did not understand the letter to be notification of a circumstance…his reaction was that of an objectively reasonable underwriter in his position” (at para 188). Similarly, the judge had found that Mr Salter, on receiving Mr Patten’s letter to Ms Turnbull dated 5 October, “concluded that it and its enclosures constituted a notification of the circumstances referred to therein” (at para 193, although the judge did not there go on to observe that that was also the reaction of a reasonable claims director in his position). In this connection Kidsons relied on the terms in which the information in question had been summarised in the bordereau presented to Mr Howes on this occasion, namely “Possible tax errors in fiscal engineering work”: not only as being part of the presentation in itself, but also as reflecting Millers’ claims department’s understanding of the information contained in the 31 August letter. On behalf of Kidsons, Mr Nicholas Davidson QC submitted that this phrase was at the very least an apt summary of the notification and encapsulated Kidsons’ case, namely that the 31 August letter was general notification of the full range of Mr Torrance’s concerns and covered the whole of S@FI’s work. It was also submitted that, if necessary, this was material which could supplement the 31 August letter for the purpose of notification.

In this connection, it is convenient to mention here another document which, although not brought into existence until later, reflected another professional view of the material in question. This was a document made up by Ms Turnbull of Camerons on 29 November 2001 (and printed on 3 December 2001) in relation to “Miller Claim No. K101-009”, ie the S@FI claim file reference. This document has been described as a “data sheet” and appears to have been a standard document produced by Camerons in connection with their claims handling service. It was sent forward by Camerons to Millers on 28 January 2002 and subsequently was used in the April and July 2002 presentations as part of the relevant bordereau. It referred to “date of notification” as “31/08/01” and also contained the following:

“(b)

Claim situation: - Concern that tax products marketed by S@FI Ltd (manned entirely by staff and partners of Insured) could be criticised. Awaiting results of investigation being conducted by independent expert (retained by S@FI) to see if concerns founded.

Await update once investigations complete.

(c)

Liability:- Possible”

The reference to “tax products marketed” was relied upon as again supporting a wider view of the information contained in the 31 August letter. It was submitted that although Ms Turnbull was not an underwriter, her role in compiling a summary of the information provided (as part of the claims handling service performed by Camerons) was to all intents and purposes similar to the exercise which a reasonable claims underwriter would himself have undertaken.

The second half of the October 2001 presentation occurred on 18 October when Millers presented the same material to Ms Jean Hackworthy, a claims assessor for syndicate 683, the second lead Lloyd’s underwriter. Ms Hackworthy was under the mistaken impression that the previous years’ claims handling agreement was in place as a term of the policy, in which case she had no role to play in receiving notifications or handling claims. She therefore scratched the bordereau as follows:

“Seen nil o/s. Please advise how claims reserves are going to be managed at LCO. As per C/H [claims handling] agreement we do not need to see files routinely.”

It was for that reason that, come the third presentation in April 2002, when Millers saw only syndicate 839 (not syndicate 683), the judge held that presentation to the lead syndicate was also presentation to the second lead syndicate.

At this time, however, Millers did not make any presentation to the following Lloyd’s or company markets. The judge said that Mr Salter agreed that Millers should have done so and that the matter appears simply to have been overlooked.

I have already referred to the concession which Mr Kealey made at trial that the 31 August letter did contain a reference to Mr Torrance’s concerns but limited to the implementation of S@FI’s tax schemes. Despite this concession the judge had held, in support of Mr Kealey’s wider submissions, that the first presentation had not been an effective notification of a circumstance at all since it could not in context objectively have been regarded as such. However, Mr Kealey did not make, or at any rate did not maintain throughout the trial, a similarly wide submission in relation to the second presentation, that made in October 2001. In relation to that presentation Mr Kealey accepted that it did amount to an effective notification, that is to say that it was objectively to be regarded as intended as a notification of circumstances, and he also continued to accept as a matter of construction of the 31 August letter itself that it contained reference to Mr Torrance’s views relating to the implementation of S@FI’s tax schemes, albeit his concession went no wider than that. However, Mr Kealey also had a further point to limit the effective scope of the second presentation’s notification of circumstances, a point which depended not on the construction of the documents presented but on the internal state of Kidsons’ awareness.

It will be recalled that GC4 is written in terms of the assured’s awareness, viz “The Assured shall give…notice…of any circumstance of which they shall become aware during the period specified in the Schedule which may give rise to a loss or claim against them”. Mr Kealey submitted that by the time of the October 2001 presentation, in the light of the IRB report and its acceptance by the NEC, Kidsons were no longer aware of any circumstance (which might give rise to a loss or claim against them) beyond that arising from defects in the implementation of DOS and no other scheme. Therefore Mr Kealey’s position with regard to the October 2001 presentation was a composite one: Yes, it was an effective notification of circumstances; Yes, as a notification it was potentially broad enough to cover claims arising out of criticisms of the implementation of S@FI’s tax schemes as a whole; No, it went no further than implementation and did not extend to concerns or criticisms about the tax schemes (“products”) themselves or their mis-selling; but finally, No, there was no awareness on the part of Kidsons at the relevant time beyond that of potential problems with the implementation of DOS. I shall refer to this last point as the “awareness point”.

The net result was that Mr Kealey’s overall concession with regard to the October 2001 presentation was that there was an effective notification of circumstances limited to the implementation of DOS. This concession was set out in the underwriters’ closing written submissions and explored in oral argument with the judge, wherein Mr Kealey made his concession absolutely clear. The judge referred to it in para 201 of her judgment, and in particular para 201(v):

“The only circumstance of which notice was effectively given on 17 and 18 October 2001 to the two leading Lloyd’s syndicates was in respect of procedures followed in certain cases relating to Discounted Option Schemes: this was because by the time when purported notification was made to these two syndicates of the circumstances set out in the 31 August 2001 letter, the only circumstance of which Kidsons was aware falling within the description of “some procedures followed in certain cases” related to Discounted Option Schemes.”

The judge rejected Mr Kealey’s awareness point, essentially on the ground that the test “which may give rise to a loss or claim” was an entirely objective one, Kidsons were aware of the whole gamut of Mr Torrance’s concerns, and the fact that at a later stage they may have become more or even entirely sanguine about those concerns did not remove their awareness. Implicit in that ground was her view that awareness of Mr Torrance’s concerns was awareness of circumstances which, objectively speaking, might give rise to a loss or claim. On that basis it might be said, and the appellants do submit, that the judge should have accepted Mr Kealey’s concession in the wider form which his failure on the awareness point would require: viz that there was an effective notification of circumstances in October 2001 relating to (but also limited to) the implementation of all S@FI schemes. However, the judge did not accept Mr Kealey’s concession at all, but concluded that the October 2001 presentation was as ineffective in total as a notification of circumstances as the September 2001 presentation had been. As she said (at para 205):

“However, for reasons different from those put forward by Mr Kealey, and indeed more fundamental than his, I conclude that there was no effective notification…”

On this appeal, the appellants say that the judge was wrong not to have accepted Mr Kealey’s concession, as far as it went: but, of course, their primary case goes further and extends to the submission that the true scope of the effective notification given in October 2001 embraced, as a matter of construction of the 31 August letter together with the other documents then presented, the whole range of Mr Torrance’s concerns. That primary case is assisted (but not of course made) by Mr Kealey’s concession, in as much as the October 2001 presentation is accepted as a valid and effective notification of circumstances, and also in as much as Mr Kealey accepted that it is not limited to DOS itself (subject to his awareness point). The appellants do not say that it was illegitimate or procedurally unfair for the judge not to have accepted Mr Kealey’s concessions, but they say that she was wrong not to have done so, and in any event wrong to have rejected the October 2001 presentation as totally ineffective, and that her reasons for doing so do not justify her conclusion.

For his part, Mr Kealey on behalf of the underwriters submits as his primary case that the judge was right in the conclusion to which she came, even though it went beyond his position at trial; and his secondary alternative case is that, if there was an effective notification of circumstances and/or if his trial concession is given effect on appeal, he nevertheless was right to limit it to notification concerning implementation of DOS, partly as a matter of construction of the documents concerned and partly because of his awareness point, which he raises by means of a respondent’s notice.

I shall revert to the detailed arguments on such matters below.

In between the second and third presentations, Kidsons were primarily concerned with the merger with Baker Tilly which took effect on 1 April 2002. The meeting of the NEC at which Kidsons resolved to pursue merger discussions with Baker Tilly took place on 18 October 2001. On 23 and 24 October management teams of both firms met. There was a private conversation during this meeting on a managing partner to managing partner basis in which Mr Greatorex (the managing partner of Kidsons) told Mr Longe (the managing partner of Baker Tilly) that the IRB had been set up to look into allegations made by a member of staff about S@FI, but this did not strike Mr Longe as significant at the time and he did not report it to the Baker Tilly due diligence team. Due diligence was conducted in November/December 2001, concluding on 13 December. In the course of this process Kidsons made formal written disclosure about potential procedural difficulties concerning the tax residency of the IOM companies set up as part of the implementation of DOS, but of no other concerns about S@FI or its operations. The disclosure letter said that the situation about DOS had been “fully disclosed to [Kidsons’] professional indemnity insurers”. The judge observed (at para 149):

“I conclude that if Kidsons had at this stage been genuinely concerned about wide-spread and deep-seated problems across the entire range of S@FI products, it is highly likely that these concerns would have been reflected in the Disclosure letter or uncovered by Baker Tilly in the course of its due diligence enquiries.”

The partners of both firms voted to proceed with the merger on 19 December 2001.

On 28 February 2002 Kidsons’ NEC discussed the on-going review of DOS cases under Mr Garner-Jones and suggested that “a summary of the situation should be prepared to put the PII underwriters in the picture”. On 26 March 2002 Mr Garner-Jones attended a further consultation with Mr Tyre QC concerning DOS. It will be recalled that Mr Tyre had been previously consulted in August 2001. He was guarded but generally more reassuring in his advice. On 27 March 2002 Mr Garner-Jones reported on the consultation to the NEC. The minutes report:

“b.

IRB Report

GGJ described his recent meeting with Colin Tyre QC, the Counsel who had originally doubted the efficacy of certain of the DOSs. With appropriate proofs available, he was now of the view that if firm supporting arguments were put forward there was a reasonable chance of the schemes succeeding…

GGJ suggested that DOS cases should be dealt with on an individual basis…There was no need now to make any pre-emptive reports to the Inland Revenue. Much work needed to be done, but GGJ believed that all difficulties could be handled in the normal course, generally by correspondence. All this should be presented to the IRB who, hopefully, would be able to approve the situation…

DG [Mr Gwilliam, another member of the NEC]…believes that there was now less risk from the Inland Revenue, but there was still a risk of claims from clients where schemes had failed because of procedural mistakes…

c.

Report to PII Underwriters

As mentioned in the previous NEC meeting, it would now be appropriate for a report on the situation to be prepared for PII underwriters.”

This was the immediate background and impetus to the third presentation. On 28 March 2002, the day following the NEC meeting, the 28 March letter (see para 14 above) was written by Mr Patten to Ms Turnbull at Camerons. On 4 April 2002 Ms Turnbull replied to Mr Patten as follows:

“I note that investigations are continuing but that there is likely to be a delay in the production of Ray Armstrong’s report. I look forward to hearing from you once you are in a position to provide me with further information, when presumably, it will be possible to tell whether this matter is likely to develop. Until such time it, of course, remains unclear as to whether this constitutes a circumstance/claim within the terms of the policy and, accordingly, we are not in a position to confirm whether cover will apply.”

The judge referred to this response (at para 207) as interesting but not relevant. I respectfully agree that it is not relevant in the sense that, unlike Ms Turnbull’s data sheet dated 3 December 2001, it was not a document presented to underwriters. As for any relevance that it might have as the expressed view of a professional person concerned with claims handling, it is not inconsistent with her earlier summary. The reason for her natural caution is not, however, clear. As expressed, it appears to be because she believed that matters would depend on how they developed. It is possible that she was influenced by Mr Howes’ original “wp” scratch. As it is, the judge found that the 28 March letter was an effective notification of circumstances.

On 19 April 2002 Millers presented the S@FI claims file to Ms Constable on behalf of lead syndicate 839. The claims file now included the 28 March letter. The bordereaux file now included Ms Turnbull’s data sheet. Ms Constable scratched the 28 March 2002 letter “Await any relevant CMS [Camerons] comments”.

The judge found (at paras 209/210):

“In my judgment, the presentation of the 28 March 2002 letter, together with the other documentation, did at this stage amount to a valid notification of circumstances...On a fair reading of the 28 March 2002 letter, and of the summary contained in the S@FI data sheet prepared by Camerons of the earlier 31 August 2001 letter, and in the context of the investigation into S@FI products, in my judgment the reference in the 28 March 2002 letter to the fact that

“in some instances there might be procedural difficulties involving the Trustees for each scheme affecting the implementation of the scheme and this might lead to the possibility of criticism in the future”

cannot sensibly be read as confined to the two specific named examples of Discounted Option Schemes that Mr Tyre was considering…The bordereau adverted to the possibility of claims, and liability, and although the information provided was still exiguous, the reasonable recipient would at this stage, in my judgment, have appreciated…the possibility of claims…”

The judge rejected, however, any more extensive notification, holding that the data sheet and bordereau could not extend the notification contained in the letter. Moreover, on the basis that for these purposes time only started with Mr Tyre’s March advice (“it was reasonable for Kidsons to wait until it had received the advice of Mr Tyre in relation to these schemes before making its notification to Underwriters” at para 212), there was no difficulty about Kidsons having acted “as soon as practicable”. For these purposes the judge took a start date of 27 March 2002, the date when the NEC considered Mr Tyre’s advice.

On the previous day, 18 April 2002, Millers had entered the S@FI claim details on the LPC CLASS electronic claims system to which the second and third respondent underwriters subscribed. The system is operated by the broker inputting a sequence for a particular risk. The sequence comprises a number of screens, containing information in respect of the policy to which the sequence relates. The sequence is intended to be read in conjunction with the underlying claims files. The purpose of the information in the sequence is to provide a link to the relevant paper bordereau and from there to the individual claims files. The CLASS system does not dispense with the need for presentation of paper claims files. The practice is for these to be presented to the IUA lead insurer (the second respondent) once the CLASS sequence has been entered. On 19 April the bordereaux file was presented to the second respondent, who scratched it and returned it to Millers on 22 April. On the same day the S@FI claims file was left with, but not personally presented to the second respondent’s claims manager, Mr Gage. No indication was given by Millers that there was any urgency in the matter. Mr Gage got round to reviewing the claims file (the judge said “in due course”) on 10 May 2002. That was ten days after the expiry of the policy period. He considered the information available was insufficient to constitute proper notification of circumstances and scratched the 28 March letter accordingly –

“We do not accept this as a notification. It may be as the insured say material information that should be presented to uw’rs.”

In cross-examination Mr Gage was in some difficulties in explaining how the information presented was material without being a notification of circumstances. Be that as it may, Millers did not react to Mr Gage’s rejection of the notification: it appears to have been completely overlooked.

The second respondent returned the claims file to Millers on 14 May 2002. The third respondent essentially followed the second respondent’s lead.

In the meantime, presentation had also been made to the fourth and fifth respondents, the other company underwriters who had not subscribed to IUA and the CLASS system. They were represented by a company called Admiral, which was provided with Camerons’ bordereau and data sheets on 17 April 2002, and with Millers’ bordereaux file and the underlying claims file on 21 May 2002 (ie one week after the latter was returned to Millers by the third respondent). Admiral returned the bordereaux and claims file to Millers, scratched without further comment, on 11 June 2002. I shall refer to this date again with reference to the timing of the fourth presentation to the following Lloyd’s market.

Thus the third presentation embraced the two leading Lloyd’s syndicates and the four companies making up the company market underwriters, but not the following Lloyd’s market. This presentation was left in limbo by Ms Constable, rejected by Mr Gage, and scratched without comment by Admiral. The judge held, however that it was an effective notification of circumstances based on the 28 March 2002 letter supported by the Camerons data-sheet; and jumping off, as it were, from Mr Tyre’s advice of 26 March 2002. The judge rejected the relevant underwriters’ submission that the notification was too late (not “as soon as practicable”) for the same reason. Indeed, the underwriters were willing to accept that, on such a rationale of the timing involved, the judge was entitled to find that the April 2002 presentation was not too late.

I have described this third presentation as the April 2002 presentation but, as will have been observed, it is a series of presentations to a number of different underwriters, all of whom needed to be notified, save to the extent that any of them was represented for such purposes by any of the others (as in the event the second lead syndicate was represented by the first, and as the fourth and fifth respondents were both represented by Admiral). As will also have been observed, the presentation to the second and third respondents extended (in a sense) into May, and the presentation to the fourth and fifth underwriters extended at least into late May (and possibly even into June). It is not entirely clear from the judge’s findings when she considered those presentations to have been made and/or completed. At any rate no point in any event arises in connection with these presentations on this appeal: on behalf of all the underwriters except for the following Lloyd’s market Mr Kealey accepts (in, as he puts it, the limited sense of “takes no point”) that these presentations were in time. In the circumstances it is unnecessary to delve further into the timing implications of the history found by the judge.

The judge’s finding in relation to this presentation were as follows:

“220.

No information was presented to XCS, the successor to the Lloyd’s Claims Office representing the following Lloyd’s market, until July 2002 (over 2 months from the expiry of the Policy). No presentation at all was made to XCS during the Policy period. The first presentation to XCS in respect of the policy year was on 24 July 2002…

221.

In my judgment, the notification given to the following Lloyd’s market was not given as soon as practicable and therefore was not compliant with the time requirements of GC4. Notice which was not given until 24 July 2002, almost three months after the expiry of the policy period, cannot, on any realistic basis, be regarded as given “as soon as practicable”, if the start date is taken as 27 March 2002. There was no impediment to the following market being notified within the Policy Period, as indeed the Lead Underwriters and for the most part, the Company market were. Although the experts agreed that it was not unusual for there to be delay in presentation to the following market, and Mr Ellis opined that “in practice” no point would be taken, Underwriters are nonetheless entitled to take the point that strict compliance with the requirements of the Policy was necessary and that, even allowing for latitude in presentation to the following market, notification at the end of July was not on any basis “as soon as practicable”.

222.

Even if I were wrong in this conclusion, for similar reasons to those stated above, any notification was confined to procedural difficulties affecting the implementation of Discounted Option Schemes as referred to in the letter dated 28 March 2002.”

In this connection I observe that the judge had in mind the extent of the latitude allowed for the purpose of presentations. She recorded this about the expert evidence (at para 174(xxvi)):

“It is recognised in the market that it takes time for notification of a claim or circumstance to be made to following insurers, with the result that there may be a delay between the time of first notification of the matter by the assured to his broker and notification by the broker to the following market. The kind of delay that is acceptable would normally be measured in weeks or at most “some months”. More than this would not be acceptable, especially since the market recognises the importance of insurers receiving information about claims and circumstances promptly.”

It will be recalled that the last date in the sequence of the history of the April 2002 presentation was 11 June 2002 when Admiral returned the bordereaux and claims files to Millers. This date has, on this appeal, become the basis for Millers’ submission that the timing of the fourth presentation to the following Lloyd’s market should, for the purposes of GC4’s “as soon as practicable”, be judged only by reference to it, viz 11 June 2002, and not some earlier date such as the judge’s 27 March 2002 (the day after Kidsons learned of Mr Tyre’s further views).

Thus Mr Graeme McPherson QC submits on behalf of Millers that the various presentations had to be done severally and in sequence and with the use of the original bordereaux and claims files: so that if the whole of the April 2002 presentation was in time, then it ought to follow that the July presentation was in time as well.

The difficulties with this submission, however, are twofold. The negative aspect is that such a submission was not made to the judge. The groundwork for some elements of the basis for it was perhaps prepared in the expert evidence (see for instance para 174(xxvi) of the trial judgment cited above), but Mr McPherson has been unable to show us any sign in the material before us including the transcript of the trial that the argument made to us by reference to the date of 11 June 2002 was addressed to the judge. The positive aspect of Mr McPherson’s difficulties is that not only were the experts agreed that the July presentation was not made as soon as practicable but counsel for Millers at trial (Mr Roger Stewart QC) himself conceded in closing submissions that the July 2002 presentation was not made as soon as practicable.

I am therefore satisfied that the way in which Mr McPherson has advanced this point based on the date of 11 June 2002 amounts to a new point on appeal and thus one which, on a question of fact of this kind, this court should not accommodate. In any event, for reasons that will I hope become clear in my discussion of the second presentation below, unlike the judge I do not regard the third presentation as beginning as it were a new chronology with the date of 27 March 2002. The history really goes back at least to August 2001, not only because of Mr Torrance’s concerns but also because, even in the limited context of DOS, Mr Tyre’s original advice in that month about DOS very much added to Kidsons’ consternation about at any rate that scheme. In my judgment therefore the real starting point for the July presentation goes back to August 2001 and in that context even Mr McPherson would find it impossible to suggest that the July 2002 presentation was as soon as practicable. In truth, his submission was only able to present some initial plausibility based on the judge’s date of 27 March 2002 and the other underwriters’ complaisance about the timing of the April 2002 presentation. Having said that, and bearing in mind that complaisance, there is a certain unattractiveness in the overall scheme of things about the Lloyd’s following market insisting on the delay point. But there it is: they are entitled to do so, and the consequence is that a long and detailed argument ensued about the legal effect of GC4’s “as soon as practicable” requirement in the construction of the policy terms as a whole, which it will be necessary to turn to in due course.

In the meantime I revert to the 31 August letter which is the basis of Kidsons’ reliance on the second (October 2001) presentation. It was in the context of their original reliance at trial on that letter for the purpose of the first (September 2001) presentation that the judge made detailed fact findings about how it was that that letter (on at any rate the underwriters’ argument) was written in such limited and anaemic terms compared to the width and depth of Mr Torrance’s concerns.

Thus Mr Patten and Mr Flaxman who drafted that letter in concert were worried about saying too much to excite the underwriters’ concerns at a stage when Kidsons were still uncertain about the validity of Mr Torrance’s criticisms. This was in part because “[t]he contemporaneous documents and the evidence of the Kidsons partners showed that there was…a marked contrast between what Kidsons thought about Mr Torrance’s wider allegations and what it thought about his specific allegations about Discounted Option Schemes” (at para 123); in part because it was unclear whether there was real reason to believe that claims might be made against the firm and the position might be clearer after the IRB had reported; in part because they were concerned about Kidsons’ poor claims record and the prospect that that held out of higher insurance premiums at next renewal which gave further reason to be cautious (at para 179); and in part because there was a sensitivity that S@FI had not even been covered as an assured at last renewal (at para 180). It was therefore deliberately chosen to test the water by mentioning the Torrance affair to the placing rather than the claims side of underwriters and by doing so in a low key. “If the mention of S@FI and its cover caused no ripples, this would clearly smooth the path for any notification that might ultimately have to be made” (at para 181). Thus Mr Patten’s and Mr Flaxman’s contemporaneous notes speak in terms of “less said best – not seeing it as a claim” (at para 182).

The judge concluded (at para 183):

“I agree with Mr Kealey’s analysis of the evidence, namely that the notes of the conversation between Mr Patten and Mr Flaxman show that they intentionally put together a letter which was studiedly non-committal (“not to be alarmist”), gave away as little as possible (“less said best”) and was deliberately sent to the placing, and not the claims, side of Millers. Their clear objective was to mention, almost as it were in passing, the possibility of a matter in the briefest language, and to avoid notifying a circumstance until and unless it was established that it was necessary, following receipt of the IRB’s report, which was expected shortly.”

What is the relevance of these detailed findings, all of which lie internally to Kidsons, as it were on the Kidsons’ side of the curtain? Earlier in her judgment the judge had said this (at para 72):

“Underwriters accepted that the fact that a document was not intended by an assured to constitute notice under GC4, did not preclude it from qualifying in fact as such a notice…I agree that, whilst Kidsons’ intention as to what it was intending to notify, is not relevant to the objective interpretation of a purported notice from the perspective of the reasonable recipient, Kidsons’ state of mind is nevertheless relevant to determine the extent to which it was aware, and hence capable of notifying, circumstances which might give rise to a loss or claim under GC4. I also accept that such evidence informs the Court why the 31 August 2001 and 28 March 2002 letters came to be written in what I regard as the coy and restricted terms in which they were written. Moreover, and more generally, the evidence relating to what Kidsons intended to notify to Underwriters necessarily informs the Court in relation to the underlying merits of the case, which even in a case of construction it cannot disregard.”

On this appeal, Mr Kealey continued, as below, to accept that such internal and subjective material of intention was irrelevant to the question of construction of the letters in their objective context; but at the same time he continued also to promote his submission of the relevance of Kidsons’ knowledge to his awareness point; and to encourage the court’s attention to such material for the purpose of the general “underlying merits of the case”, which he sometimes addressed by reference to the need for Kidsons to comply with a pervasive duty of good faith. In this last context he referred, as did the judge, to the experts’ evidence which satisfied her that “both the broker and the insurer are expected to act in good faith and be frank and open with each other” and that the insurer is dependent upon his assured to tell him what the assured knows “in fair, comprehensive and comprehensible terms” (at para 174 (ii) and (ix)).

Although it is not a matter in issue on this appeal, I should nevertheless state my diffident and untutored opinion that in certain exceptional circumstances intention may not be irrelevant to a unilateral question such as the giving of a notice. In general, questions of subjective intention are irrelevant to an issue of objective construction: but, as Chitty on Contracts, 29^th ed, 2004, Vol 1, at para 2-156puts it – “the objective test is…subject to the limitation that it does not apply in favour of a party who knows the truth”. Thus a person who does not intend to contract will be bound by the objective appearances of contract, but may not himself be entitled to invoke the objective test so as to hold another party to an alleged contract.

Whether that is so or not, however, has not been explored on this appeal and I do not think it would be right, particularly in an area of some difficulty and possible controversy, to go behind the underwriters’ acceptance that, save for the purposes of their separate awareness point, subjective factors are irrelevant. In such circumstances, I am rather sceptical about the judge’s comment that such subjective factors might somehow be relevant as part of the “underlying merits of the case” to the question of construction. Indeed, as will become apparent when I deal below with the competing submissions concerning the effectiveness of the second presentation, I suspect that the judge’s views about those underlying merits, at any rate as they may have been found to exist at the time when the 31 August letter was first composed and/or first presented, may have misled her when dealing with an objective appraisal of the second presentation.

It is important in this context to make clear that, despite Mr Kealey’s making play at times with the concept of good faith and the absence, as he submitted, of frank disclosure of the full width of Mr Torrance’s concerns at the time of writing the 31 August letter, ultimately he had no submission, as he himself volunteered, that there had been any breach of any duty of good faith in connection with at any rate the second and subsequent presentations. Indeed, it is important to appreciate that, come 17 October 2001 and the time of the second presentation, the underwriters’ submission is not that that presentation was in bad faith or was inadequate or in any sense coy, but that on the contrary Kidsons lacked awareness of any circumstances requiring notification beyond problems with the implementation of DOS.

It will be recalled that Mr Kealey submitted that at the time of the second presentation the underwriters had ceased to be aware of any circumstances which might give rise to a loss or claim save for those concerned with the implementation of procedures with respect to DOS. His submission to the judge below and again on appeal was that, although he accepted that the relative clause “which may give rise to a loss or claim against them” raised an entirely objective question and one that did not depend upon the assured’s subjective assessment of circumstances, nevertheless the concept of awareness necessitated some subjective element of belief in the possibility of a loss or claim in relation to relevant circumstances. The judge recorded his submission (which was premised on Mance J’s observations about “knowledge” in Insurance Corporation of the Channel Islands v. The Royal Hotel Ltd [1998] Lloyd’s Rep IR 151 at 162 (ICCI), as follows:

“99.

Mr Kealey submitted that awareness, for the purposes of GC4, was tantamount to knowledge, and that that required Kidsons to have a genuine belief, at the actual time of notification, that there was a real possibility that claims would be made in relation to such wide-ranging matters. In the absence of such a belief at the time of notification, Mr Kealey submitted that there was no relevant “awareness” for the purposes of GC4.”

However the judge rejected this submission in these terms:

“100…In my judgment, an assured can envisage the possibility that a known circumstance or circumstances may give rise to a loss or claim against him, and thus have the requisite “awareness” for the purposes of GC4 without having, as Mr Kealey requires, “a genuine belief at the actual time of notification that there was a real possibility that claims would be made in relation to such wide-ranging matters”…

Nevertheless the judge went on to sum up her findings of fact on the hypothesis that she was wrong in that conclusion, in these terms:

“101…Thus if, contrary to my conclusion, Mr Kealey’s submissions as to what was required for an assured to be “aware” were correct, and it were relevant to examine the strength of Kidsons’ “belief”, I would have concluded that Kidsons did not have “a genuine belief at the actual time of notification that there was a real possibility that claims would be made in relation to such wide-ranging matters”. The only exception was in relation to concerns over the implementation of the Discounted Option Schemes, where Kidsons did take the view that such concerns might indeed be well founded…

102.

Accordingly, I conclude that Kidsons were “aware” for the purposes of GC4, during the Policy Period, and at the time of notification, of the wide ranging circumstances adverted to by Mr Torrance in his various communications and the possibility that, if his concerns were well-founded, claims might arise as a result. However, the reality of the position was that, at the time when Kidsons wrote the 31 August 2001 letter, and the 28 March 2002 letter, and at the time those letters came to be presented to Underwriters, the firm as such did not believe that such wider concerns were well founded, or would in practice give rise to claims. The only area where the firm had real concerns that claims might arise was in relation to Discounted Option Schemes.”

Both appellants and respondents rely on these passages. The appellants rely on the judge’s finding that Kidsons were aware of relevant circumstances. Therefore on any view, they say, there is no limitation by reason of the concept of awareness on the width of the notifications to be found objectively in the presentations concerned. The underwriters, however, point to the fact that the judge found that there was no genuine belief in the existence of any circumstances other than in connection with the implementation of DOS.

The importance of this point is that if the underwriters’ trial concession were to be adopted, that is to say the concession that the second presentation was effective and extended, as a matter of construction even if not as a matter of Kidsons’ awareness, to the implementation of Kidsons’ schemes generally, then, subject to the awareness point, the second presentation could not be limited to DOS, unless the underwriters’ awareness point, raised by their respondent’s notice, prevailed.

In my judgment, any difficulty in this point rests in the ambiguity of identifying the relevant “circumstances”. Normally such circumstances arise outside the insured itself, for instance in the intimation of a possible complaint. In such a case which I would regard as typical, the two questions will be: (i) Have such circumstances come to the attention of the insured (during the policy period) so that he can be said to be aware of them? (ii) Are such circumstances such that they “may give rise to a loss or claim against them”? The latter question is an objective one; the insured may have his own views about the complaint, but the question has to be looked at objectively. In the present case, however, the problem which arose was internal, generated by the views of Mr Torrance. Normally the subjective personal views of an insured about the nature of a risk which he presents to underwriters for cover are irrelevant: provided, of course, that all material information is fairly presented, it is for the insurer to rate the risk, not for the assured. Mr Torrance, moreover, was only an employee: he was not a member of the firm. What then in such a case are the “circumstances” for the purposes of GC4?

There has been no clarity about this. One possibility is that they were the expression of Mr Torrance’s concerns themselves, as illustrated by the memoranda presented to the S@FI board and to the NEC. Another possibility, however, is that they were the underlying substance of those concerns, as illustrated for instance by Mr Tyre’s opinion, or by the IRB’s assessment of them. In the ordinary way, I am doubtful that an insured’s own concern, without more, that he may have made a mistake is a relevant circumstance which can entitle him to give notice of circumstances, thus extending his claims made policy into future years: otherwise, every insured could extend his policy indefinitely simply by a notification based on his own lack of confidence.

In this connection, we had cited to us the example given by Sir John Donaldson MR in Thorman v. New Hampshire Insurance Co (UK) Ltd [1988] 1 Lloyd’s Rep 7 at 10, where, in a case concerning an insured who was an architect, he said:

“A typical example would be a belated realization, based upon a study of professional journals, that perhaps he had specified inadequate foundations for a building which he had designed and which had already been erected.”

The significant qualification there, however, is “based upon a study of professional journals”. So here, the consternation caused by the report of the first consultation with Mr Tyre, in August 2001, represents a substratum of underlying external fact, over and above Mr Torrance’s mere concerns, which, if the point had been taken at trial, could stand as a “circumstance” for the purposes of GC4.

This ambivalence is present in the judge’s judgment, at para 102, where she says that Kidsons were aware “of the wide ranging circumstances…adverted to by Mr Torrance in his various communications” (one possibility) and that “if his concerns were well-founded” (the other possibility), claims might arise as a result. Ultimately, however, it seems to me that Kidsons were relying on (and the underwriters were responding to) the former as the relevant circumstances: this is illustrated by Mr Kealey’s concession at trial that the second presentation gave effective notice of Mr Torrance’s “view that the Inland Revenue, if minded, could be critical of some procedures followed in certain cases” (to quote the 31 August letter). There was no dispute on the part of the underwriters that claims might arise as a result, only as to Kidsons’ subjective appreciation of such a possibility. However, Kidsons were plainly aware of Mr Torrance’s concerns. The question was whether, objectively speaking, they might give rise to claims, and this never appears to have been challenged. Indeed, the absence of challenge is duplicated by the oddity that there is no express finding in this or any other passage of the judgment below that Mr Torrance’s concerns were in themselves circumstances which “may give rise to a loss or claim against them”. This appears to have been taken for granted.

In other parts of her judgment, however, the judge appears to have considered that the underlying facts, as distinct from the mere expression of Mr Torrance’s opinion, are what constitute “circumstances” and might therefore have led to a valid notification. Thus, in dealing with the third presentation (that of April 2002) and the 28 March letter, the judge seems to have regarded the circumstances there notified as being not so much a limited form of Mr Torrance’s concerns as “in the light of the adverse observations of Mr Tyre,…the possibility of claims arising from procedural defects (“difficulties”) involving the Trustees in relation to the implementation of Discounted Option Schemes” (at para 209).

It seems to me, in other words, that Mr Kealey’s awareness point proceeds (implicitly) from an alternative analysis of the situation. If Mr Torrance’s concerns were in themselves not “circumstances”, because they were too internalised, then they were not relevant for notification. If, however, they were “circumstances”, then plainly Kidsons were aware of them. If, however, internal concerns of such a kind only become “circumstances” when once they have been given some objective support, such as by Mr Tyre’s opinion or the IRB investigation, then of course there is need to consider the nature of such opinion or investigation before it is possible to speak in terms of circumstances of which an assured is aware. This is not so much a point about awareness as about the nature of the relevant circumstances.

As it is, however, both Kidsons and the underwriters were prepared to approach the question on the basis that the relevant circumstances were Mr Torrance’s concerns. Since Kidsons were plainly aware of these, it seems to me irrelevant that Kidsons may themselves, at some or other time, have lacked “a genuine belief” that claims might ensue (para 101). In any event, Mr Kealey’s test for awareness of “a genuine belief” is taken from Mance J’s test for knowledge, not awareness. We must surely be aware of allegations about many things in which we have no genuine belief.

In my judgment, therefore, the underwriters’ respondent’s notice, to the effect that Kidsons were not aware of the circumstances relied on by them, fails. If there is a valid point about the ineffectiveness of Mr Torrance’s concerns, by themselves, as “circumstances” for the purposes of GC4, it was not taken.

I am now in a position to deal with the question of the effectiveness and construction of the second, October 2001, presentation. I have set out the facts about it and Mr Kealey’s concession concerning it, at paras 24/39 above.

The judge considered that this presentation was no more effective than the first presentation, even though (unlike the first presentation) it was made by a member of the claims team at Millers to a member of the claims department at syndicate 839, consisted not only of the 31 August letter, but also of a claims bordereau which was headed “Claim circumstance notification bordereaux” and referred to “Possible tax errors in fiscal engineering work”, and was regarded by the syndicate’s claims examiner, Mr Howes, as an attempt to notify a circumstance.

The judge rejected this attempt, however, on the following grounds (see her paras 74/76, 171, 174, and 192/205). The test which she applied was ultimately that of Lord Steyn in Mannai Investment Co Ltd v. Eagle Star Life Assurance Co Ltd [1997] AC 749 at 767/8, namely that “the contextual scene is always relevant…the inquiry is objective: the question is what reasonable persons, circumstanced as the actual parties were, would have had in mind”. In the particular context of “unilateral notices served under contractual rights reserved”, Lord Steyn went on to adopt the test discussed in Delta Vale Properties Ltd v. Mills [1990] 1 WLR 445 at 454 in these terms:

“Even if such notices under contractual rights reserved contain errors they may be valid if they are “sufficiently clear and unambiguous to leave a reasonable recipient in no reasonable doubt as to how and when they are intended to operate:” the Delta case, at p. 454E-G, per Slade L.J and adopted by Stocker and Bingham L.JJ.; see also Carradine Properties Ltd. v. Aslam [1976] 1 W.L.R. 442, 444. That test postulates that the reasonable recipient is left in no doubt that the right reserved is being exercised. It acknowledges the importance of such notices. The application of that test is principled and cannot cause any injustice to a recipient of the notice. I would gratefully adopt it.”

The judge also relied on the expert evidence before her to the effect that an insurer is dependent upon his assured to tell him what he knows “in fair, comprehensive and comprehensible terms”.

On this basis, the judge gave her reasons, as I have already remarked “more fundamental than” Mr Kealey’s, for totally rejecting the effectiveness of the October 2001 presentation, as follows. The reasonable recipient would have thought that the documents presented were “insufficiently clear and unambiguous to constitute notice of a circumstance giving rise to a claim under GC4”; they were “far too vague and nebulous”; there was no identification of any error, act or omission or possibility of any claim; there was no identification of any products or procedures. She continued:

“He would not be any further informed by the description in the October bordereau that the nature of the claim was “possible tax errors in fiscal engineering work”, not merely because the description is so vague as to be useless, but also because it does not even reflect the view of Mr Torrance, as stated in the 31 August 2001 letter, that the Inland Revenue might be critical of “procedures”.

Finally, the reasonable recipient, circumstanced as the parties were, would necessarily have appreciated that the context in which the 31 August 2001 letter had been originally written and presented to Underwriters was ostensibly to inform the placing side of Millers and Underwriters about the nature of S@FI’s business. This again would have militated against any interpretation of the letter as a notification of circumstances.”

With respect to the judge, I do not think that this is correct. The letter may not have been a very satisfactory letter, especially if, as Kidsons wished to submit (but which may be doubted), it was supposed to notify the underwriters of the full width of Mr Torrance’s concerns. But it was on all objective criteria intended to be a notification of circumstances, it was presented by the claims side of Millers to the claims side of syndicate 839, with an accompanying bordereau whose terms reflected that fact, and the question therefore was rather what the presentation reasonably conveyed to its recipient. Although the question of construction being dealt with by the judge was ultimately one of law and the judge was therefore not strictly bound to accept Mr Kealey’s concession with respect to this second presentation, I do not think that the judge ought to have rejected it. After all, the question of construction on a one-off presentation concerned only the parties to it, and Mr Kealey’s concession had obviously been carefully considered and approved by his clients, who are clearly as experienced in reading and evaluating such documents as anyone. A judge would have to be very sure of his or her ground before rejecting such a concession.

In my judgment, the concession was properly made (I leave open for discussion below whether the appellants are right to submit that the presentation went wider than that which Mr Kealey was prepared to concede). The question for present purposes is what the letter said, not what the letter did not say. It was presented as a matter of the claims side of things. The letter did not say that any claim had been made, indeed it said that no claim had been made, and therefore the essence of it must, at any rate in theory, have been to provide information of a circumstance which might give rise to a claim. That is what the bordereau confirmed, when it was headed “Claim circumstance notification bordereaux”. I accept the judge’s approach, based on the expert evidence before her, that a bordereau cannot go further than the essential letter of notification: but it has to be read together with that letter, and it can clarify, even if it cannot extend. The letter itself had spoken of Mr Torrance’s view that “the Inland Revenue, if minded, could be critical of some procedures followed in certain cases”: that, in context, is the procedures by which certain tax products had been implemented.

GC4 says nothing about how a notification is to be made, other than that it must be in writing and given as soon as practicable after awareness of circumstances which may give rise to a claim. That is, on the face of it, a fairly loose and undemanding test. It is quite unlike other possible notice requirements which might specify more precisely what a notice must contain and/or when it must be given (eg as in Mannai itself, where the notice to determine the lease had to expire “on the third anniversary of the term commencement date”). Both the requirement of awareness, and the test that a claim “may” arise, are open-textured. Moreover, if circumstances arise where notification should be given, the assured is required to give notice (“shall give to the Underwriters notice”): it is not in his option, as many contractual notices might be. The assured is thus put in danger of either being required to give notice at a time when the circumstances of which he is aware require investigation before he can speak precisely about them, or of being told that he has failed to give notice at all as soon as practicable. Moreover, the authorities on such clauses do not seem to demand great specificity: see Rothschild Assurance at 22/23, Friends Provident at 147 (Moore-Bick J) and 55 (CA).

In this context I am somewhat sceptical that the Delta Vale test (“sufficiently clear and unambiguous to leave a reasonable recipient in no reasonable doubt as to how and when they are intended to operate”) is of great assistance. I would be unhappy to accept any suggestion that a purported notification of circumstances to an insurer becomes ineffective whenever there might be a real point of argument as to its proper width. I suspect few such notifications could survive such a test, if rigorously applied. The “no reasonable doubt” test is a creature of the criminal law, and even if transformed into the civil law’s more familiar test of the unequivocal representation, the familiar context of the latter is where an estoppel prevents a party setting up his contractual right. In the present context, however, the notification is a means of working out existing contractual rights, not of departing from them. In my judgment, it has to be remembered that the Delta Vale test operates typically where the notice is subject to a mistake and the question remains whether it can nevertheless operate successfully despite that error. Thus in Carradine Properties v. Aslam the notice to quit gave an incorrect date for the determination of the lease, 1973 instead of 1975. In Delta Vale itself the notice to complete specified a more generous 28 days instead of the contract’s provision for 15 days. In Mannai, the notice to determine the lease had to take effect on 13 January, but wrongly specified 12 January. In such circumstances the courts had to determine whether the notices were effective at all. In Mannai, the minority (Lord Goff of Chieveley and Lord Jauncey of Tullichettle) would allow of no departure at all from the contractual requirements. The majority, however, were willing to accept that even a notice containing an error might be reasonably recognisable as an effective notice and employed the Delta Vale test to police the latitude allowed for an erroneous notice to take effect according to contract. In my judgment, such a test has little to bring to the different problem faced in the present case. (It may be otherwise if there is doubt about whether a notice is a notice of circumstances at all, but as I have observed above, it clearly was objectively intended to be such a notice, and Mr Kealey’s concession accepted that it was, and the real question was as to its width.)

The judge to some extent appears to have recognised this herself, citing different and more general formulations (“sufficiently clear” Nunes v. Davies Laing & Dick (1985) 51 P&CR 310 at 314 per Sir Nicolas Browne-Wilkinson V-C, and “reasonably clear to a reasonable person” Barclays Bank v. Bee [2001] EWCA Civ 1126, [2002] 1 WLR 332 at para 45 per Arden LJ), and concluding (at para 76 of her judgment):

“I am not therefore convinced that semantic cavilling over the precise formulation of the test assists the ultimate resolution of the problem. There may well be uncertainty at the time of notification as to what the precise problems or potential problems are…”

To some extent, I think the judge may have been misled into rejecting Mr Kealey’s concession by his own submissions elsewhere as to the deliberate coyness with which the 31 August letter had been written; and by her own views, thus engendered, of what she called “the underlying merits of the case” (see at paras 62/66 above). However, whatever one might think about the opportunism of this litigation, I do not think that the 31 August letter, when presented, was other than a fair and honest, if cautious, representation of the problem viewed by Kidsons at that time. Certainly in the aftermath of the IRB report, Kidsons’ concerns had been essentially laid to rest, otherwise than in matters of implementation of DOS. Therefore, as for the experts’ view that a notification should be “fair, comprehensive and comprehensible”: there was ultimately no submission that the presentation was unfair, in the sense of being in breach of the duty of good faith; if it was not comprehensive, that was because Kidsons did not credit Mr Torrance’s views in the light of the IRB report, and in any event, a non-comprehensive notice of circumstances of which an assured is aware carries its own penalty in that it is unable to give notice of that which is omitted; and as for its comprehensibility, that is a matter of construction on which Mr Kealey’s concession was entitled to the fullest credit.

The remaining question, therefore, is whether the second presentation went further than Mr Kealey’s concession and extended, as the appellants submitted, to the whole gamut of Mr Torrance’s views. In support of that submission it was said: that what was being notified was the existence of unresolved concerns, possibly inchoately mentioned in the letter, but easily ascertainable by reference to Kidsons’ documents recording Mr Torrance’s views; the fact that there was to be an external investigation demonstrated the seriousness with which the problems were being handled and the importance of the situation; the letter’s very first paragraph mentioned “fiscal engineering” as the subject matter of S@FI’s business, and this was again emphasised in the bordereau’s reference to “Possible tax errors in fiscal engineering work”.

In my judgment, however, this submission fails. While the letter’s first paragraph introduces fiscal engineering as the business of S@FI, the second paragraph in that context distinguishes between “products” which have all been validated by counsel and “procedures” where the Revenue, in the view of Mr Torrance, could be critical in some cases. The third paragraph then states that it is “this view” (and none other) that the NEC intended to investigate. A reasonable reader would take from the letter that the only concern which was being notified to underwriters was “this view”. That, however, is the extent and subject-matter of Mr Kealey’s concession. “[T]his view” relates to procedures in way of implementation of some of S@FI’s products. “Possible tax errors in fiscal engineering work” ex the bordereau in context goes no further than that. The implementation of procedures is part of S@FI’s fiscal engineering work, and it is the possible tax errors in the course of such procedures that Mr Torrance has identified as being potentially subject to criticism by the Revenue.

In my judgment, therefore, the second presentation gave effective notification to syndicates 839 and 683 of Mr Torrance’s view that the implementation of certain of S@FI’s products might be criticised, and thus might give rise to possible claims or losses.

The third (April 2002) presentation (see paras 41/52 above) updates the position by means of Kidsons’ 28 March letter (para 14 above) and extends the notification to the company market.

The 28 March letter, although accepted by the judge as an effective notification of circumstances albeit limited to “procedural difficulties…affecting the implementation” of DOS, is in its way even more unsatisfactory than the 31 August letter. Its opening words (“Some months have passed since we last corresponded”) refer to the 31 August letter, which is mentioned expressly at the end of the first paragraph. That earlier letter had spoken only of an investigation into Mr Torrance’s view (“this view”) in relation to procedures. The first paragraph of the 28 March letter, however, states, incoherently and inaccurately, that “We have put a lot of effort into a technical investigation of the sale of products”. The previous letter had never spoken of any investigation of product selling, nor had the initial IRB report covered any such investigation. That report had been produced back in September/October 2001 but never made the subject of a follow up letter to underwriters. The first and second paragraphs of this second letter wrongly suggest that the IRB had not as yet reported. The third paragraph then deals with the March 2002 consultation with Mr Tyre, which is the real catalyst of this letter. Again, as in the 31 August letter, a distinction is made between products, whose “technical efficiency was accepted”, and implementation, where “procedural difficulties” in relation to DOS are referred to. In the event, this letter is narrower than the 31 August letter, which was not confined to DOS. It certainly does not constitute notification of Mr Torrance’s concerns across the board. Indeed, those concerns are no where addressed. There was only a weak submission on the part of Kidsons to the contrary. Nothing that accompanied the presentation made up for these deficiencies.

The judge regarded this letter as constituting the first effective notification, albeit confined to DOS implementation. I would regard it as merely updating, however unsatisfactorily, the second presentation’s concern with implementation. It follows that I would hold that the somewhat wider terms of the 31 August letter included in the third presentation means that the company market as well as the two lead Lloyd’s syndicates have been given effective notification of possible problems arising out of the implementation of S@FI products generally, not limited to DOS. Moreover, as I understood Mr Kealey’s position, he accepted that that was the case.

Because I have concluded above that the judge’s decision that the fourth presentation in July 2002 to the Lloyd’s following market was not made “as soon as practicable” was correct it is necessary to consider the policy terms for the purpose of evaluating the parties’ conflicting submissions as to whether such delay was destructive of the complete efficacy of that presentation. These submissions of course centre on GC4 itself, but go much wider than that term alone. I have set out GC4 towards the beginning of this judgment. I will set it out again below, in its wider context.

The policy begins with a preamble which makes Kidsons’ written proposal the basis of the insurance. The preamble is immediately followed by the insuring clauses, as follows:

“Whereas the Assured have made the Underwriters a written proposal bearing the date stated in the Schedule containing particulars and statements which together with any other information which may have been supplied it is hereby agreed are the basis of this Insurance and are to be incorporated herein.

The next relevant part of the policy terms is headed “General Institute Conditions”. These conditions (“GIC”) and especially the last paragraph of them (“GIC(b)”) are important to the structure of the parties’ submissions because the appellants found their essential argument that GC4’s “as soon as practicable” is either not a condition precedent at all or is excused its status as such by reason of the terms and function of these provisions. Thus –

“GENERAL INSTITUTE CONDITIONS

The Underwriters will not exercise their right to avoid this Insurance where it is alleged that there have been untrue statements or non-disclosure or misrepresentations of facts in the Proposal Form or in any other information which may have been supplied provided always that the Assured shall establish to Underwriters’ satisfaction that such alleged untrue statements or non-disclosure or misrepresentation of facts was free of any fraudulent conduct or intent to deceive. However,

a)

In any case where the Assured should have notified under any preceding insurance a loss or a claim made against them or circumstances which could give rise to a a loss by or a claim against them and the indemnity or cover available hereunder is greater or wider in scope that [sic, sc than] the indemnity to which the assured would have been entitled under such preceding insurance (whether with other Insurers or not) then the Underwriters shall only be liable to indemnify the Assured in respect of the loss or claim to the extent of the indemnity which would have been afforded by such preceding insurance.

b)

Where the Assured’s breach of or non-compliance with any conditions of this Insurance has resulted in prejudice to the handling or settlement of any loss or claim the indemnity afforded by this Insurance in respect of such loss or claim (including costs and expenses) shall be reduced to such sum as in the Underwriters’ opinion would have been payable by them in the absence of such prejudice.”

There follow “General Exclusions” of which it is only necessary to refer to the fourth and fifth (“GE4” and “GE5”):

“This Insurance shall not indemnify the Assured against any claim or for any loss…

4)

Arising out of any claim or circumstance which has been notified under any other policy or certificate of insurance attaching prior to the inception of this Insurance.

5)

In respect of dishonest or fraudulent acts or omissions committed by any person after discovery by the Assured of reasonable cause for suspicion of fraud or dishonesty on the part of that person.”

The “General Conditions” come last of all, of which the following are relevant and in particular of course GC4, of which it will be convenient to speak with reference to its two sentences:

“3.

The Assured shall as a condition precedent to their right to be indemnified under this Insurance give to the Underwriters notice in writing as soon as practicable

a)

Of any claim made against them or any of them

b)

Of the receipt of notice from any party of an intention to make a claim against them

c)

Of any loss suffered by them or any of them

d)

Of the discovery of reasonable cause for suspicion of dishonesty or fraud on the part of any former or present partner consultant sub-contractor director or employee of the Firm(s) whether giving rise to a loss or claim under this Insurance or not

4.

The Assured shall give to the Underwriters notice in writing as soon as practicable of any circumstance of which they shall become aware during the period specified in the Schedule which may give rise to a loss or claim against them. Such notice having been given any loss or claim to which that circumstance has given rise which is subsequently made after the expiry of the period specified in the Schedule shall be deemed for the purpose of this Insurance to have been made during the subsistence hereof.

6.

Any claim first notified to the Assured prior to the expiry date of this policy will be deemed to be dealt with under this policy provided it is properly notified to Underwriters within 15 calendar days of the expiry day.

11.

In any dispute in connection with the terms conditions exclusions or limitations of this Insurance it is specifically understood and agreed that the terms conditions exclusions and limitations of the Approved Wording contained in the Prospectus for Approved Insurers to the Institute of Chartered Accountants (in England and Wales/of Scotland/ in Ireland) shall take precedence over any terms conditions exclusions or limitations contained herein which are less favourable to the Assured.”

GC3 and together with it GC6 are relevant to the question of whether the second sentence of GC4 is a condition precedent extending to the timing (“as soon as practicable”) of the required notice. GC11 incorporates as clauses paramount the applicable Institute (“ICA”) clauses. The ICA clauses in question, to which the court has been referred as throwing light on the construction of the policy terms and/or as clauses paramount in their own right, read as follows:

“B.6. The Insurers will not exercise their right to avoid this Certificate or any contract of insurance therein, or claim to be discharged from any liability hereunder, on the grounds of any alleged non-disclosure or misrepresentation of facts or alleged untrue statements in the Proposal Form, provided always that the Insured shall establish to Insurers’ reasonable satisfaction that such alleged non-disclosure, misrepresentation or untrue statement was free of any fraudulent conduct or intent to deceive.

However, if such alleged non-disclosure, misrepresentation or untrue statement consists of or includes a failure to inform the Insurers of any circumstance of which the Insured was aware which might give rise to a loss or claim against the Insured the indemnity hereunder in respect of that loss or claim shall be limited as follows:

(a)

Where the Insured should have notified such circumstance under any previous insurance (whether with other Insurers or not) the indemnity hereunder shall be limited to the indemnity which would have been available under the earliest such previous insurance if such circumstance had been properly notified.

(b)

Where the Insured should have notified such circumstance under this Certificate prior to obtaining any increase in the limit of indemnity hereunder or other variations of the terms hereof the indemnity hereunder shall be limited to the indemnity which would have been available prior to such increase in cover or variation of the terms hereof.

B.7. Where the Insured’s breach of or non-compliance with any condition of this Certificate has resulted in prejudice to the Insurers:

(a)

in the handling or settlement of any claim against the Insured;

(b)

in the amount of any loss sustained by the Insured;

(c)

in the obtaining of reimbursement from any dishonest or fraudulent person as referred to in condition B.11,

the indemnity afforded hereunder (including liability for claimants costs) shall be reduced to such sum as in the Insurers’ reasonable opinion would have been payable by them in the absence of such prejudice.

B.9. The Insured shall as a condition precedent to their right to be indemnified hereunder give to the Insurers notice in writing as soon as practicable:

(a)

Of any claim made against them or any of them.

(b)

Of receipt of notice from any party of an intention to make a claim against them.

(c)

Of any loss suffered by them or any of them.

(d)

Of the discovery of any reasonable cause for suspicion of dishonesty or fraud on the part of any former or present partner, director, employee, consultant, sub-contractor or alternate of the Firm(s) whether giving rise to a loss or claim hereunder or not.

B.10. If the Insured shall become aware during the period of any insurance of any circumstance which may give rise to a loss or claim the Insured shall give notice in writing to the Insurers as soon as possible. Such notice having been given:

(a)

any claim which may subsequently be made against the Insured arising out of that circumstance shall be deemed to have been first made against the Insured during the Period of Insurance;

(b)

any loss which the Insured may subsequently discover they have sustained, being a loss arising out of that circumstance, shall be deemed to have been first discovered by the Insured during the Period of Insurance”

It will be observed that, broadly speaking, ICA clause B6 coincides with GIC other than GIC(b); ICA clause B7 coincides with GIC(b), ICA clause B9 coincides with GC3, ICA clause B10 coincides with GC4, and there is no equivalent of GC6.

ICA Guidance concerning the ICA terms provided as follows:

“Claims handling

6.5

All members, whether sole practitioners, partners or directors, together with their employees, should be aware of the importance of notifying insurers promptly of claims or circumstances which may give rise to a claim. Everyone in the firm should know that failure to comply with underwriters’ requirements in this regard could seriously prejudice the firm’s rights and entitlement to indemnity under the policy.

6.6

One person, at the level of principal, should be given the task of recording and coordinating information about claims or circumstances and of notifying brokers/underwriters accordingly. That person should regard the prompt notification to brokers/underwriters as a first priority and should not wait until there have been developments or until a detailed report of the matter has been prepared.

6.7

All staff should be encouraged to report promptly to the individual designated in the above paragraph any matter of which they become aware.

6.8

Claims or circumstances should be regarded objectively. If there are circumstances which might reasonably give rise to a claim then insurers should be notified immediately. This is regardless of the fact that currently allegations may be vague or not specified and regardless of whether the member personally thinks liability is unlikely. (In this latter regard the question of liability is a legal one which only lawyers and, ultimately, the courts are competent to decide.)”

A number of matters may be observed about the above terms. (i) GC3 (and its ICA equivalent clause B9) is expressly stated to be a condition precedent. (ii) The first sentence of GC4 (and of its ICA equivalent B10) is not expressly stated to be a condition precedent. (iii) The second sentence of of GC4 does not in terms state that the fulfilment of the obligation contained in its first sentence is a condition precedent, but it is inherent in the opening words of that second sentence (“Such notice having been given”) that the giving of “Such notice” is a condition of the deeming provision thereafter set out. It would seem to follow that if “Such notice” has not been given, then a loss or claim subsequently arising out of relevant circumstances about which notice should have but has not been given as required by the first sentence will not be deemed to have been made during the policy period. It is disputed by the appellants, however, that “Such notice” embraces the timing requirement of the first sentence. (iv) Whether or not the requirements of the first sentence have been fulfilled, if a loss is suffered or a claim is made during the policy period, then the deeming provision of GC4 is not required to render such a loss or claim prima facie within the insuring clause of the policy. (v) GIC(b) states, in general, that a breach of “any conditions” of the policy will sound in damages by way of a reduction in the “indemnity afforded” in a quantum which depends on the underwriters’ own assessment of the prejudice caused to them by such breach. There is a dispute between the parties whether the expression “any conditions” covers conditions precedent such as that to be found prima facie in GC3 and, on the underwriters’ case, in the words of GC4 “Such notice having been given”. The underwriters submit, and the judge agreed, that GIC(b) does not cover conditions precedent, otherwise the clause could not speak of “the indemnity afforded”. The appellants refer to the width of the words “any conditions” and to the oddity that the most basic failure to have given any notification of circumstances under a previous insurance would not (by reason of GIC(a)) prevent (although it might limit) recovery of an indemnity in the absence of fraud: whereas the mere failure to give a notice in time under GC3 or GC4 would, on the underwriters’ argument, be fatal.

The ICA Guidance, in the passage on claims handling cited above, expresses repeatedly and in strong terms the importance of prompt notification of claims or circumstances, together with the warning that failure to comply with underwriters’ requirements could seriously prejudice the member’s rights to indemnity.

The hammer and the anvil of the problem set before the court consisted principally of two conflicting considerations.

On the one hand, the appellants pointed to GIC, with its three aspects, as they submitted, for ameliorating the position of an insured who, in the absence of fraud or any intent to deceive, had failed to do what he should have done, either in the way of giving proper disclosure etc for the purpose of making the policy or in the way of breach of policy conditions. If (1) a total failure to give proper disclosure or a misrepresentation or an untrue statement in the proposal form (which is itself made the basis of the policy) are excused and do not entitle the underwriters to avoid; if (2) a total failure to give notice of a claim or of circumstances which might give rise to a claim under any preceding year’s insurance is excused, subject to a reduction in the amount of the indemnity afforded to that which would have been provided under such preceding insurance; and if (3) a breach of any conditions of the policy itself is merely to result in a set-off for any prejudice suffered by the underwriters: then, it is anomalous and inconceivable that a late notice of claim or of circumstances without more could prevent recovery. Even a late notice of circumstances within the policy year itself, and one that caused no prejudice at all, would, if timely notice were a condition precedent, be destructive of any indemnity for a connected claim which fell outside the policy year. For there could then be no recovery under the second sentence of GC4, and the notice already given would prevent cover for the risk under a subsequent policy: either because future years’ insurers would be unwilling to cover that risk or because any future insurance would itself contain a clause like GE4 excluding indemnity for any “circumstance which has been notified under any other policy…attaching prior to the inception of this insurance”. Alternatively, the premium charged for any cover under future insurance would become prohibitively expensive.

On the other hand, if a late notice of circumstances was not destructive of the right to an indemnity for any loss or claim arising out of such circumstances, even though such loss or claim should emerge after the end of the policy year, then a claims made policy would become entirely open-ended: however much delayed a notice of circumstances may be, and however long after the end of the policy period the claim which arises out of such circumstances comes to be made, the policy extends to cover that claim.

Thus expressed, the argument becomes one between on the one hand avoiding an anomaly which would otherwise arise on the detailed provisions of the policy and on the other hand turning a claims made policy on its head.

Neither consequence is of course attractive. To evaluate this clash of contentions it is necessary to descend to the detail of submissions about the construction of the various clauses implicated in this issue.

Two critical questions arise. One is whether the expression in GC4 “as soon as practicable” is a condition precedent at all. If it is, the second question is whether its force as a condition precedent is removed by GIC(b).

The appellants make the following points in support of their submission that GC4’s “as soon as practicable” is not a condition precedent. It is not expressed to be such in the first sentence of GC4, which merely states that the assured “shall” give notice as soon as practicable. That can be contrasted with GC3 (dealing with notice of a claim) whose opening words expressly state the requirement to give a notice as soon as practicable in the words “shall as a condition precedent”. The expression “as soon as practicable” taken by itself is not redolent of a condition precedent. It does not set a firm deadline and suggests an element of laxity, which the expert evidence confirms exists in practice. Neither does the expression “Such notice having been given” at the beginning of GC4’s second sentence implicitly involve a condition precedent, because the word “Such” speaks only to the nature and content of the notice, for instance that it must be in writing, not of its timing. That is supported by the somewhat different wording of ICA clause B10 (“If the Insured shall become aware…of any circumstance…the Insured shall give notice to the Insurers as soon as possible”) as contrasted with GC4’s “shall give to the Underwriters notice in writing as soon as practicable of any circumstance” etc: because in the former case the timing of the notice is given its own separate space as a merely adverbial phrase, whereas in the latter case the question of the timing of the notice is sandwiched between the references to it being in writing and to its content. Support is also to be found in GC6’s reference to a notice which may be “properly notified” even though given as late as 15 days after the expiry of the policy period, where “properly” is said to relate only to matters of form and content and not of timing.

Moreover, the absence of a condition precedent role for the timing of a notice of circumstances is supported by GIC(b) whose paradigm concern appears to be with providing for a special regime in cases where a notice (whether of claim or of circumstances) is late, since it is lateness which typically may cause “prejudice to the handling or settlement of any loss or claim”. The special regime is that breach of or non-compliance with any policy conditions is not to be visited by a failure of indemnity but by making allowance for the cost to the underwriters of any such prejudice. The special regime of GIC(b) is part of the GIC as a whole, whose essence is to prevent the underwriters from escaping all liability even where the assured has been guilty of the most serious or complete of failures of disclosure or notification, as long as there has been no fraud or intent to deceive. The significance of GIC(b) is underlined by the fact that it even applies, and typically would do so, to late notification of a claim which had been made or of a loss which had been discovered during the policy year, even though that is the subject matter of GC3, which does expressly refer to the giving of a notice in writing as soon as practicable as a “condition precedent”. In Diab v. Regent Insurance Co Ltd [2006] UKPC 29, [2007] 1 WLR 797 at para 14, Lord Scott of Foscote observed, with reference to an insurance clause which stated that “No claim under this policy shall be payable unless the terms of this condition have been complied with”, although the point did not arise on the matters debated before the Privy Council, that –

“It does not necessarily follow, however, that every element of condition 11 must be treated as a strict condition precedent with any failure to comply barring the claim. Their Lordships have particularly in mind the 15-day period after the loss or damage has been incurred within which a claim in writing accompanied by the requisite details is required to be delivered to the insurer.”

That, it is said, applies here to the requirement of “as soon as practicable”.

Some parts of this analysis are more cogent than others, but in my judgment as a whole it does not convince. I find it impossible to say that the expression “as soon as practicable” can be divorced from the rest of the first sentence of GC4 (or of B10) for the purposes of the phrase “Such notice having been given” (which is found in both GC4 and B10). Moreover, the function of that phrase is clearly that of a condition precedent. The policy is a “claims made” policy, that is to say it primarily insures only claims made against the assured within the policy year, but “Such notice having been given” it extends to cover claims made outside the policy year (“after the expiration of the period specified in the Schedule”) where such claims arise out of the circumstances of which notice has been given. The extension is achieved by deeming such post-expiry claims to have been made during the policy year. It is submitted by the appellants that this is not a true “extension” but part and parcel of the package bought by the assured. I agree that it is part of the package, and that a “claims made” policy could hardly work on any other basis (see Rothschild Assurance at 22, Friends Provident (Moore-Bick J) at 142). However, it is commonly and I consider accurately referred to as an extension of cover (ibid), because it changes the essential basis of the policy, which is limited to claims made within the policy year, into something different. The deeming provision, however, emphasises even as a matter of drafting technique that the basis of cover remains the same, as stated in the insuring clause in Section 1. But the price, or condition, of this extension, or of this deeming provision, is the giving of a proper notice. Whereas GC3 states expressly that the notification of a claim to the insurer is a condition precedent, and needs to do so since the insuring clause is written in terms of what the assured has received (“claims made” against him) and not in terms of claims passed on to underwriters within the policy year, the language of GC4 reflects the functional and purposive basis of the insurance as a whole by making the extension or deeming of cover depend on the giving of the notice defined in GC4’s first sentence. That is a paradigm example of a condition precedent. Without the notice, there is no extension of cover. If, on the other hand, the only remedy for a late notice of circumstances were to be a liability in damages (subject to the possibility of a repudiatory breach, in context a most unhelpful theoretical possibility), then the “claims made” policy is turned into an open-ended policy which covers claims made at any time. Since the essence of the extension granted is concerned with the timing of claims made to the assured, it is only to be expected that the timing aspect of the notice of circumstances to be given by the assured is an important part of the provisions concerned.

It was argued that it would make no sense for the first sentence of GC4 to be a condition precedent, since failure to give a notice of circumstances as soon as practicable might destroy cover even for a claim first made against and properly notified by the assured within the policy year itself. However, the first sentence of GC4 is not in itself a condition precedent to cover. It is not stated to be. Therefore the consequence held up to criticism does not arise. But the second sentence of GC4 is a condition precedent, for that is the means (“Such notice having been given”) by which a claim made or loss discovered outside the insuring clauses is brought within them.

Therefore the timing expression “as soon as practicable” is part of the condition precedent contained in the second sentence of GC4. The next question is whether GIC(b) unmakes that condition precedent, as it is submitted by the appellants that it does. In my judgment, however, it does not. Although GIC(b) refers to “any conditions of this Insurance”, which is of some assistance to the submission, it does not in terms refer to conditions precedent. There are at least three difficulties in the way of construing “conditions” to include effective conditions precedent. The first is that the whole clause is premised on “prejudice to the handling or settlement of any loss or claim”. Where, however, there has been a failure of a condition precedent, in the absence of waiver by the underwriters of such failure, no question arises of such prejudice. Secondly, the expression “the indemnity afforded by this Insurance” is inapposite to a failure of a condition precedent, because in such a case the insurance affords no indemnity. Therefore, that phrase has to be construed by the appellants as meaning “the indemnity afforded or which would in the absence of such breach have been afforded by this Insurance”. That reformulation is made all the more difficult in light of the fact that similar expressions are found in GIC(a) – “the indemnity to which the Assured would have been entitled under such preceding insurance” and “the indemnity which would have been afforded by such preceding insurance”. Therefore the draftsman had the concept and the words to hand but did not use them in GIC(b). The appellants submit that the subjunctive expression is appropriate in GIC(a) because that clause is dealing with a hypothetical situation. So be it: but on the basis that GIC(b) covers failures of conditions precedent, the same logic applies to GIC(b) as well. Thirdly, GIC(b) proceeds to speak of “such sum as in the Underwriters’ opinion would have been payable by them in the absence of prejudice”. However, if there has been a failure of a condition precedent and no waiver, then the policy would have provided no indemnity and nothing would have been payable, even in the absence of prejudice.

These points emphasise that the whole premise and assumption of GIC(b) are that any breach of a condition precedent has been waived, so that the breach or non-compliance being spoken of by the clause can only be remedied by damages and the breach concerned has been transformed into the breach of a merely innominate term. In other words, it is not so much that “conditions” does not in itself extend to “conditions precedent”, as that the situation being discussed is one in which any failure of a condition precedent is being treated as a mere breach of an innominate term. Moreover, the clause as a whole is not written in terms which are designed to undo the effect of conditions precedent and to create cover where none exists. If it were, one would expect the point to be addressed straight on: the clause would emphasise that the only remedy for breach of a condition precedent (or any breach) would be a reduction in such sum as the underwriters assessed would compensate them for any prejudice caused to them. However, if that were to be the intended effect of GIC(b), then it would be surprising that the provisions to be found elsewhere in the policy would bother to speak the language of condition precedent at all.

In this connection, there was discussion as to whether a breach of the express condition precedent found in GC3, relating to notification of claims, was within GIC(b). My reasoning above would suggest that it is not (unless waived). It was submitted nevertheless that such a breach would be the archetypal context of the rationale of GIC(b). And if GC3’s condition precedent was within GIC(b) then there was no reason why GC4’s condition precedent should not be within it as well. However late a GC3 notice was made, if it caused no prejudice, there was no reason why it should not be effective, and if it caused prejudice the financial consequences of that prejudice were met by the GIC(b) disposal – under which underwriters were entitled to assess their own loss.

There is force in this submission which may be said to gather further strength from the position of GIC(b) as part of a series of provisions favourable to the assured taken within GIC as a whole. If GIC(b) does not cover the unwaived breach of a condition precedent, then it does not give the assured anything at all: on the contrary, it permits underwriters to reduce the indemnity by a figure of their own assessment to represent prejudice suffered by them. That said, I am not persuaded that GIC(b) does cover the unwaived breach of GC3. It is not what GIC(b) says (see above). On the basis of this submission, it is surprising to find GC3 still expressly stated to be a condition precedent, and if anything even more surprising to find that ICA’s clause B9, the equivalent of GC3, also contains the express language of condition precedent. It is noticeable moreover that ICA’s clause B7 (the equivalent of GIC(b)) is not run together with B6 (the equivalent of the rest of GIC) but stands separately from it. Moreover, ICA Guidance (cited above) emphasises in the strongest terms that failure to give timely notice of claim or circumstance “could seriously prejudice the firm’s rights and entitlement to indemnity under the policy”. That goes very much further than a warning that a full indemnity might be reduced by the underwriters’ assessment of the cost of any prejudice. If in such circumstances the ICA terms were nevertheless intended to operate as the appellants suggest they were, then they need to be revisited.

Finally, the appellants contrast the favourable terms afforded to an assured under GIC generally (apart from GIC(b)) with underwriters’ construction of GC4 and GIC(b). If an assured who has failed to give any disclosure or notice at all, when he should have done, is nevertheless not thereby to be deprived of cover but to receive an indemnity (even if its scope or amount may be reduced under the provisions of GIC(a)), it is submitted to be anomalous for a late notice of circumstances to be entirely destructive of cover. Indeed, it may be said that an assured who has given a late notice of circumstances is in a worse position that one who has given no notice at all. The answer to this submission, however, in my judgment lies in the contrast between a claim or loss which falls within the current policy period and a claim or loss which might (or might not) arise in the future, after the expiry of the current policy period. If a claim or loss falls in the current policy period, then prima facie it is squarely covered by the insuring clauses under a “claims made” policy. Even so, that prima facie indemnity may (a) be lost where the right to avoid survives because there has been fraud or an intent to deceive; and (b) be reduced to earlier levels of indemnity where a proper notice under any preceding year’s insurance would have fixed the claim or loss within that earlier year. Where, however, the claim or loss in question has not been made or suffered within the policy year, then the absence of a proper notice is an altogether more problematical matter. The underwriters are then asked to accept an open-ended liability into the future. If in such circumstances there are the possibilities of some anomalies, I do not consider that that is a good enough reason for forcing contractual language to bear a meaning which it does not reasonably bear. I agree that an assured who makes a late notice of circumstances under a year 1 policy may therefore possibly be even worse off than an honest assured who in error gives no notice at all – either because the former cannot then obtain insurance for year 2 or cannot obtain it save at a prohibitive price, or because of the inclusion in his year 2 policy of a clause like GC4 – however those problems lie in the future and do not necessarily concern the underwriters of the year 1 policy.

It is not an easy matter to construct a policy on a “claims made” basis which allows for an extension to cover future claims to mitigate the difficulty for an assured who learns during the policy year of circumstances which may give rise to a claim in the future after that year’s expiry. Nevertheless, I do not consider that it would be right to depart from the language which the parties have chosen in order to mitigate all the difficulties which a careless though honest assured might find himself to be in because of his failure to give a proper and timely notice of circumstances of which he is aware which may give rise to a future claim or loss. Ultimately, the danger of possible anomalies in such circumstances must make way for the greater danger of turning a “claims made” policy on its head (see para 109 above).

I have sought to take account of all the appellants’ submissions at all stages of their argument, even though in working through them I have had to express what might properly have to be described as provisional conclusions. At the end of the day, I am not persuaded by those submissions and would uphold the judge’s own conclusions about them.

As something of a postscript, I would add some diffident observations about GC6, which it will be recalled provides as follows:

“6.

Any claim first notified to the Assured prior to the expiry date of this policy will be deemed to fall to be dealt with under this policy provided it is properly notified to Underwriters within 15 calendar days of the expiry day.”

The judge regarded this clause as providing for a long stop date of 15 days after the end of the policy year for the giving of any notice of claim to underwriters. If this is correct, the clause does not seem appropriately worded, for it gives the impression of granting something to the assured, while the judge’s interpretation amounts to a limitation on the meaning of “as soon as practicable” for claims requiring notification towards the end of the policy year. Moreover the expression “deemed to fall to be dealt with under this policy” suggests that a claim outside the cover is being brought within the cover by the fulfilment of the proviso. In my judgment a better interpretation is that any claim first made against the assured during the policy year, ie any claim prima facie within insuring clause 1, will be dealt with as a matter for indemnification under the policy, even if it is not notified to the underwriters as soon as practicable, as long as it is notified to underwriters within 15 days of the end of the year. Thus “properly notified” must in context mean “otherwise properly notified”; and so to this extent the appellants are right to suggest that “properly” here does not include timeously. There must however be something wrong with the notification to underwriters of a claim first made against the assured within the policy year if it has to be “deemed to fall to be dealt with under this policy”, and all that is left, if the notice is otherwise properly notified is the requirement of timeousness, “as soon as practicable”. However, while maintaining cover for a late notice to underwriters of a claim made against the assured, GC6 on this interpretation does not entirely excuse the breach of the GC3 term “as soon as practicable”: so that this breach will be dealt with under GIC(b). On this interpretation, a claim first made to the assured right at the end of the policy year will not need the assistance of GC6 if it is properly and timeously notified to underwriters “as soon as practicable”, even if that means after the 15 day deadline.

If this interpretation is correct, then there is additional support for the survival of the condition precedent in GC3, as well as a good explanation, outside GIC(b), of how it is that a late notification to underwriters of a claim first made against the assured within the policy year may yet survive for indemnity, subject to the terms of GIC(b), provided it is at least notified to underwriters within 15 days after the end of the policy period. It is understandable that this dispensation is not, however, granted to late notification of a loss first discovered during the policy period under insuring clause 2, because presumably it is still more important for such a loss, caused by dishonest or fraudulent acts or omissions of a partner etc, to be notified promptly. However, I have reached my conclusions about GC4 and GIC(b) independently of these additional considerations.

In this connection there was a submission, ultimately I think withdrawn, made by Michael Harvey QC on behalf of Camerons (alone of the appellants) that any notice of circumstances had in any event to be notified to underwriters within the policy year itself (or by analogy with GC6 within 15 days of its expiry). However, I do not think that is correct, although such a limitation to the policy year is sometimes made an express term of such an insurance. In the present case, however, GC4 does not so limit the giving of a notice of circumstances. On the contrary, the timing of such a notice is differently expressed, namely “as soon as practicable”: whereas the concept of the duration of the policy year is used in the clause itself for a different purpose, namely for the period during which the assured becomes aware of the circumstance in question: thus “any circumstance of which they shall become aware during the period specified in the Schedule”. In my judgment therefore I think that Mr Harvey’s quondam submission is an impossible one.

It follows from my conclusions in this section of my judgment that the judge was right to say that the fourth presentation (to the Lloyd’s following market) being out of time was thus ineffective.

In sum, I would dismiss this appeal, save to this extent: in addition to the effective notification relating to the implementation of DOS which the judge considered had been made to the two lead Lloyd’s syndicates and to the four parties in the company market by means of the third presentation of April 2002, I would hold that the second presentation of October 2001 to the two lead Lloyd’s syndicates was an effective notification relating to the implementation of S@FI products generally. It also follows from my reading of the 31 August letter and the third presentation that the company market as well as the two lead syndicates have received effective notification relating to the implementation of S@FI products generally, not limited to DOS. The appellants have therefore had a measure of success on this appeal, but whether it is ultimately of any value to them I cannot say.

On a purely formal point, the first and second Lloyd’s lead syndicates and the Lloyd’s following market are not represented by separate named underwriters but are grouped together as the first respondents, although their interests are different and the outcome in relation to them is different. The parties should consider what is the most appropriate form of order in these circumstances.

Rix LJ has set out comprehensively the relevant facts and policy terms. I agree with his conclusion that the second presentation was an effective notification to the two lead Lloyd’s underwriters of circumstances which might give rise to claims stemming from the manner of implementation, as distinct from the essential nature, of “fiscal engineering products” which had been marketed by S@FI, including but not limited to so-called Discounted Option Schemes. I also agree with him that the third presentation had the same effect in relation to the relevant companies (the second to fifth respondents).

It is common for a proposal form for professional indemnity insurance to ask the proposer to state whether he is aware, after inquiry, of any circumstances which may (or perhaps “are likely to”) give rise to a claim against the would-be insured and, if so, to provide details. Even if such a question is not asked, information about potential claims (unless trivial or their likelihood can be dismissed as remote) is likely to be material to the prospective insurer and therefore disclosable in any event. The prospective insurer is then likely to exclude cover in respect of any claims which may arise from circumstances disclosed to him prior to the policy being agreed.

In order to secure protection for the insured against such claims it is also standard for professional indemnity policies, which basically provide cover against the risk of claims being made against the insured during the policy year, to contain a provision enabling the risk of a later claim to attach to the policy where it arises from circumstances of which the insured becomes aware and gives notice to the insurer. It is not merely the insured’s awareness of the circumstances, but his giving of notice of them to the insurer, which causes the risk to attach to the policy.

These two features of professional indemnity insurance fit together. Their essence is simple and well understood by the market, but they can give rise to a variety of problems, especially when there is a change of insurer between different years.

In the present case the relevant notification of circumstances clause required the insured to give to the underwriters notice in writing as soon as practicable of “any circumstance of which they shall become aware during the [policy] period…which may give rise to a claim against them”.

There are two parts to that phrase: the awareness of a circumstance, which is a pure matter of fact, and the characterisation of the circumstance as one which may give rise to a claim against the insured. The question of construction which has been argued in this case is whether the insured needs to be “aware” that the circumstance may give rise to a claim against him and, if so, what degree of appreciation of risk is required. Is the test subjective or objective, and, if subjective, what is the subjective requirement?

It is a curious feature that in this case the second presentation purported to be a notification on behalf of the insured of circumstances which might give rise to a claim, but the insurers deny that it was effective because the insured lacked the necessary awareness to give such a notification. It is more common for such an argument to arise where the insured has not given a notification of circumstances, which the insurer says ought to have been given because the risk of a claim was objectively plain, whether or not the insured subjectively appreciated it. However this reversal of the customary roles (for which there is understandable tactical reason) is irrelevant to the question of construction.

Looking at the practical context in which a notification of circumstances clause comes to be relevant, I do not believe that the correct answer to the question is to say simply that the test is subjective or that it is objective.

The question whether a circumstance may give rise to a claim is not a matter of simple knowledge, a question of fact of which a person may or may not be “aware”; rather, it involves a degree of crystal ball gazing , an estimation of the likelihood of a claim.

At one end of the spectrum, there may be cases in which an insured seeks to notify a circumstance which is too vague or remote to be reasonably capable of being regarded in itself as a matter which might give rise to a claim. This is not as unlikely as it might sound, because an insured at the end of a policy period may have an incentive to give a notification in the widest possible terms for which there may be no real justification. The insurer would be entitled to refuse to accept such a purported notification.

In the middle of the spectrum, there may not uncommonly be cases in which different people, possessed of the same knowledge, might reasonably form different views about whether a claim was a real possibility as distinct from a remote risk. In such cases an insurer could not reject a notification of the circumstance, but nor could an insurer complain if the insured did not give such a notification.

At the other end of the spectrum are cases in which any reasonable person in the insured’s position would recognise a real risk of a claim. If so, the insured would be duty bound to give notice of it to a prospective insurer. He would also in my view be bound to give notice of it to the current insurer if the terms of the policy required him to give notice of any circumstance of which he became aware and which might give rise to a claim.

In short, in my judgment the right general approach to a policy clause which entitles an insured to give notification of a circumstance which may give rise to a claim, and thereby cause the risk to attach to that policy, is to treat the right as subject to an implicit requirement that the circumstance may reasonably be regarded in itself as a matter which may give rise to a claim. The right general approach to a policy clause which goes further and imposes a duty on the insured to give such a notification is to treat it as implicitly limited, not only by the requirement that the circumstance may reasonably be regarded as a matter which may give rise to a claim, but to a circumstance which either the insured notifies or which any reasonable person in his position would recognise as a matter which may give rise to a claim and therefore requiring notification to the insurer.

In the present case Kidsons might have reacted to the concerns expressed by Mr Torrance at various stages in different ways. I am satisfied that the matters to which he drew attention were capable of being seen as circumstances which might give rise to a claim. Kidsons might after considering them have concluded (a) that they disagreed with him and did not consider that there was any real risk such as to merit notification to the underwriters, or (b) that there was a real risk of claims arising both from the nature of the schemes and from the manner of their implementation, or (c) that there was no real risk of claims in relation to the essential validity of the schemes, but that there was a potential exposure to claims from the manner of their implementation which merited notification.

The second presentation was plainly an attempted notification of circumstances which might give rise to a claim and was understood by the underwriters to be such. Its validity has to be judged at that time. As Rix LJ has recounted, Mr Kealey QC accepted (correctly in my opinion) that the language of the 31 August letter drew underwriters’ attention to Mr Torrance’s concerns about implementation of the “fiscal engineering” schemes which was not limited to Discount Option Schemes. However, it gave no indication of any circumstance which might give rise to claims relating to the essential validity of the schemes. On the contrary, the underwriters were told that they had been validated by counsel and the warning note expressed in the letter related only to their practical implementation. I agree therefore with Rix LJ that the presentation was effective as a notification of circumstances which might give rise to claims relating to implementation of the schemes but not otherwise. (The second presentation was to the Lloyd’s first and second lead underwriters. The attempted notification to the companies was the subject of the third presentation, to which the same essential reasoning applies although the documents presented were not all identical.)

As I read her judgment, the judge’s reasons for finding that the second presentation was ineffective have two linked strands. First, the circumstances notified were “too vague and nebulous”. Secondly, the 31 August letter “gave away as little as possible” and was deliberately written in “coy and restrictive terms”. Rix LJ and Buxton LJ have come to different views about whether the judge was right to treat the presentation as ineffective on those grounds.

As to the first point, I agree with Rix LJ’s observations in paragraph 84.

The second point is more worrisome. The judge did not mince her words about the 31 August letter and Mr Kealey echoed her criticisms of it. But he also said expressly that he was not alleging that the Kidsons were in breach of any duty of good faith, although at times he came close to doing so. I would readily accept a proposition that where a policy contains a provision for the insured to notify a circumstance which may give rise to a claim and thereby attach the risk to the policy, it is impliedly incumbent on the insured to see that any such notification is a fair, if summary, presentation of what the insured knows. (This would be consistent with the expert evidence in this case referred to by Rix LJ at paragraph 63.) If the insured chooses to be deliberately misleading or economical with the truth, so that the insurer is not given a fair picture of what the insured knows, I can see a good arguable case for saying that the notification should be treated as invalid. But like Rix LJ, I would not wish to express a firm view on the matter when it has not been fully argued.

In agreement with Rix LJ, I do not consider that the judge’s decision on the invalidity of the second presentation can be sustained on the coyness ground. As Rix LJ has pointed out, the core of Mr Kealey’s submission on the second presentation was not that by the time of the second presentation the insured was aware of circumstances which might give rise to claims relating both to the essential validity and the implementation of the S@FI products, but chose to give a notification in a deliberately disingenuous form, so as not to alert the underwriters to the more fundamental risk but to use language which could nevertheless later be relied upon as wide enough to cover it. Mr Kealey’s case was that (a) as a simple matter of construction, the 31 August letter was sufficient to alert the underwriters to a risk of possible implementation claims, but nothing more; however, (b) at the time of the second presentation the insured were in fact unaware of any circumstance which might then give rise to a claim or claims (whether relating to the essential validity or to the implementation of the schemes). As to (a), I agree with Mr Kealey’s submission about the construction of the letter. As to (b), I have concluded that at the time of the second presentation, the circumstances of which Kidsons were then aware could reasonably be regarded as circumstances which might give rise to the limited risk to which the presentation alerted them. That is sufficient, absent an allegation of bad faith which Mr Kealey disavowed, to make the presentation effective for that limited purpose.

I agree with Rix LJ that the judge’s finding that the fourth presentation (to the Lloyd’s following market) was not “as soon as practicable” is unassailable for the reasons which he has given. The question is therefore whether, although the notification was out of time, it is saved by GIC (b) of the Lloyd’s policy or clause B7 of the ICA policy.

I have found this the most difficult part of the case because of the way in which the policies are drafted. Rix LJ has set out the relevant provisions and the problems to which they give rise. The root cause of the problems is that each policy is a patchwork of provisions, which have no doubt been largely drawn from other policies but do not all fit well together.

Kidsons, Camerons and Millers submitted that if the insured became aware during the policy year of a circumstance which might give rise to a claim, it could be validly notified at any time thereafter; and that the only adverse consequence for the insured of the notification being late (possibly months or even years after the end of the period of the policy) would be if the lateness resulted in “prejudice to the handling or settlement” of any claim or loss, in which event the amount of the indemnity payable would be reduced to such sum as in the underwriters’ opinion would have been payable in the absence of such prejudice.

In my judgment that is a hopeless contention. The notion that a risk may attach to the policy by reason of something known to the insured but not to the insurer, and which the insured may decide to disclose to the insurer after the expiry of the policy at any time of his own choosing, is so contrary to the way in which a professional indemnity policy would ordinarily be expected to work that it would need the clearest words to produce that effect.

The natural assumption in the case of a claims made professional indemnity policy is that claims or risks of claims (where the insured becomes aware of a threatened or possible claim and gives notice of it to the insurer) may attach during the policy but not after it.

In passing, it is pertinent to observe that GC 6 is a specific exception. It provides that any claim first notified to the insured prior to the expiry date of the policy will be deemed to be dealt with under the policy provided that it is properly notified to the underwriters within 15 calendar days of the expiry date. The reference to a claim being notified to the insured must in my view be a reference to the receipt by the insured of notice from a party of an intention to make a claim against him (to which GC3 applies). For if a claim notified to the insured prior to the expiry date were the same as a claim made prior to the expiry date there could be no question of it being “deemed” to be dealt with under the policy. I infer that GC3 is intended to provide an exceptional, short extension of the period in which the risk of a threatened claim may attach to the policy for the protection of an insured who receives a notification of an intention to claim on the eve of the expiry of the policy. But the significant point is that the contractual extension which enables such a risk to attach to the policy after the end of the policy period is both express and short.

It is also relevant to note that the prejudice which may be taken into account by the underwriters under GIC (b) is limited to prejudice to the handling or settlement of the claim or loss. The post expiry attachment of risks could well cause the insurer to suffer prejudice in much broader ways, by increasing the potential size of risks incurred but not reported, with consequent problems for the insurer in assessing its liabilities, arranging reinsurance, etc.

That said, the logical construction of GC4 and GIC (b), and the corresponding ICA policy clauses, set out by Rix LJ has the potential to produce anomalous and harsh results, as he has explained.

In these circumstances I was attracted by an argument which it appeared that Mr Michael Harvey QC was going to advance that any notification of circumstances under GC4 had to be prior to the expiry of the policy.

It is true that GC4 does not say so. But I have suggested above that the natural assumption in the case of a policy of this nature is that claims or risks of claims may attach during the policy period, but not retrospectively, and the argument would be that the language of GC4 “shall give notice as soon as practicable” is not intended or sufficiently explicit to permit such a notification to be given after the expiry of the policy (by contrast with the explicit wording of GC6). It might also be argued that the wording of the second sentence of GC4 “any loss or claim to which that circumstance has given rise which is subsequently made after the expiry of the [policy] period” is in contrast with the implicit requirement that the notification must have been given before the expiry of the policy for the risk to attach to the policy.

If that argument were right, then GC6 could be given the construction for which Kidsons, Camerons and Millers have contended, without creating unsatisfactory anomalies and without turning a claims made policy into a policy under which claims or risks of claims may attach for an indefinite period after its expiry.

However, none of the parties sought to advance the argument. Mr Harvey may not have formally abandoned it as a fall back proposition, but he certainly advanced his submissions in a different direction. In presenting their cases the parties may have had good reasons for regarding the potential argument sketched above as flawed or for not wishing to pursue it – possibly because it would lead inevitably to the dismissal of the appeal against the judgment given in favour of the Lloyd’s following market.

Whether by the route identified by Rix LJ or on the argument sketched above, I am fully satisfied that the result in this case must be the same, ie that the fourth presentation was out of time and ineffective. I do not think in the circumstances that it would be right for me to adopt an interpretation which would make no difference to the outcome of the case and which has not been tested in oral argument. (Nor would it be sensible to spend additional time examining the corresponding wording of the ICA policy.) On the arguments as they have been presented I am content to concur with the judgment of Rix LJ and would wish to reserve my opinion whether there may be an alternative way of interpreting these unsatisfactory policies.

I respectfully agree with all that has fallen from my Rix LJ, save as to the view that he takes of the effect of the letter of 31 August 2001 as an effective notification under GC4, either standing on its own or when presented to underwriters with other documents in October 2001.

The judge found as a fact, and was well entitled so to find on the basis of the compelling evidence set out in her paragraphs 182-184, that, far from the purpose of the letter having been to give underwriters full and proper notification of possible relevant claims, it was designed by its authors to say as little as possible on that subject. I entirely agree, as did the judge, that it is possible for a document that is subjectively intended not to be a relevant notification nonetheless objectively to be such a notification. However, we should approach with caution a claim that expert and experienced practitioners in this market have succeeded in doing by accident what they by design intended not to do. And that difficulty is, as the judge found, reflected in the wording of the letter itself. The judge was entirely justified in saying, in her paragraph 177, that

When reviewing the expert evidence in her paragraph 174 the judge held, in her subparagraph (ii), that it was the responsibility of the assured, in conjunction with his broker, to tell the insurer what he knows in “fair, comprehensive and comprehensible terms”. That proposition was not formally agreed, but I did not understand it to be challenged, nor could it have been. The judge set out in her paragraphs 178-185 a detailed account of how far the 31 August 2001 letter fell short in that respect; and she returned to that subject when setting out the objective construction of the letter in her paragraph 205 (iii)-(ix). Small wonder, therefore, that the recipient of the letter, Mr Elwes, did not understand it to be a notification of a claim (judgment, paragraph 186); and there was ample objective basis for his reaction, the judge having held at her paragraph 205(x) that

This (objective) defect in the letter cannot be cured by its having been subsequently used as part of a presentation to the claims department, supported by the bordereau. The objective force of the letter was fixed by its terms and the circumstances in which it was written; and it is agreed that the effect of the notification could not be extended by what accompanied it.

In my view, therefore, it is impossible in the light of this evidence and these findings to rehabilitate the 31 August 2001 letter as an effective notification of the circumstances that were known to the insureds. I would therefore dismiss the insureds’ appeal on this point.