Deeny and Others v Gooda Walker Limited and Others

QUEEN'S BENCH DIVISION (COMMERCIAL COURT)

[1996] LRLR 183, The Times 7 October 1994, The Independent 5 October 1994, (Transcript: John Larking)

HEARING-DATES: 4 October 1994

4 October 1994

INTRODUCTION:

This is a signed judgment handed down by the judge, with a direction that no further record or transcript need be made (RSC Ord 59, r9(1)(f), Ord 68, r 1). See Practice Note dated 6 July 1990, [1990] 2 All ER 1024.

COUNSEL:

G Vos QC, J Gaisman and D Lord for the Plaintiffs; B Eder QC, M Templeman, S Cookerill and S Bryan for the Defendants

PANEL: Phillips J

JUDGMENTBY-1: PHILLIPS J

JUDGMENT-1:

PHILLIPS J:

INTRODUCTION

1988, 1989 and 1990 were bad years for Lloyd's. In each of those years the market as a whole made a loss. The loss for 1988 was calculated by Chatset Ltd., who publish a Lloyd's 'League Table', at £510 million, or 13.7 % of the net premium income. In 1989 these figures were £2,063 million or 52 % of net premium income and in 1990 they were £2,915 million or 55.2 % of net premium income. These losses were not borne evenly by the Names at Lloyd's. The first and second Defendants, to whom I shall refer collectively as "Gooda Walker", managed between them four syndicates which fared particularly badly: 164 and 290, managed by the First Defendant and 298 and 299 managed by the Second Defendant. The Plaintiffs in this Action were the majority of the Names on those syndicates. They number 3,095. Between them they claim to have lost sums totalling some £630 million. They contend that these losses have been inflicted upon them because the manner in which the underwriting was conducted by the active underwriters for these syndicates was incompetent.

Whether that contention is well founded is the principal issue in this Action. If it is, the Plaintiffs will have established breach of a duty of care owed to them not only by the first and second Defendants, who managed the syndicates, but by the Plaintiffs' individual Members' Agents, who make up the other 69 Defendants in this Action. That this consequence will follow has been made clear by the decisions on preliminary points of law reached in this and parallel Actions by Saville J, the Court of Appeal and the House of Lords. I propose to summarise at the outset the effect of those decisions, incorporating the summary relating to the structure of Lloyd's with which Lord Goff introduced his speech in this case on 25 July 1994:

Every person who wishes to become a Name at Lloyd's and who is not himself or herself an underwriting agent must appoint an underwriting agent to act on his or her behalf, pursuant to an underwriting agency agreement. Underwriting agents may act in one of three different capacities.

(1) They may be members' agents, who (broadly speaking) advise Names on their choice of Syndicates, place Names on the Syndicates chosen by them, and give general advice to them.

(2) They may be managing agents, who underwrite contracts of insurance at Lloyd's on behalf of the Names who are members of the Syndicates under their management, and who reinsure contracts of insurance and pay claims.

(3) They may be combined agents, who perform both the role of members' agents, and the role of managing agents in respect of the Syndicates under their management.

Until 1990, the practical position was as follows. Each Name entered into one or more underwriting agency agreements with an underwriting agent, which was either a members' agent or a combined agent. Each underwriting agency agreement governed the relationship between the Name and the members' agent or between the Name and the combined agent in so far as it acted as a members' agent. If however the Name became a member of a Syndicate which was managed by the combined agent the agreement also governed the relationship between the Name and the combined agent acting in its capacity of managing agent. In such a case the Name was known as a direct Name. If however the Name became a member of a Syndicate which was managed by some other managing agent, the Name's underwriting agent (whether or not it was a combined agent) entered into a sub-agency Agreement under which it appointed the managing agent its sub-agent to act as such in relation to the Name. In such a case the Name was known as an Indirect Name.

Gooda Walker acted as combined Members and Managing agents for 230 of the Plaintiffs. In respect of the remaining Plaintiffs, their role was simply that of Managing Agents, acting on behalf of the Names who had joined their syndicates. The Plaintiffs who are indirect Names claim against Gooda Walker in tort and against their individual Members' Agents in contract. The Plaintiffs who are direct Names claim against Gooda Walker in contract and in tort. Whether in contract or in tort the nature of the claim is identical. It is for damages for a failure to exercise reasonable skill and care in conducting the business of underwriting on behalf of the Names. The claims relate to losses suffered by members of Syndicate 298 in the underwriting years 1988 and 1989, by members of Syndicate 299 in the same years, by members of Syndicate 290 in 1989 and 1990 and by members of Syndicate 164 in 1989.

Two preliminary issues have been determined in this Action in respect of the 1988 and 1989 years.

(1) Did Gooda Walker owe a duty of care in negligence to all the Names on their syndicates, whether or not they were in contractual relationship with them?

(2) Are the Members Agents contractually liable to the Names for any failure to exercise reasonable skill and care on the part of Gooda Walker, as Managing Agents?

Both these questions have been answered in the affirmative by Saville J, the Court of Appeal and the House of Lords. Before Saville J and the Court of Appeal a subsidiary issue was canvassed. Was it necessary when considering the performance of Gooda Walker's duty of care to have regard to the individual circumstances of each name, or was a uniform standard of performance owed to all the names? Saville J and the Court of Appeal held that the latter was the case. The premise underlying that decision is that objective standards of skill and care apply to the business of actual underwriting and those standards are the same, whether the duty of care is owed in contract or in tort.

After 1989 Names have entered into contracts both with Members Agents and with Managing Agents so that the Members Agents are no longer contractually responsible for the way in which the actual underwriting is performed. I have, nonetheless, to determine whether or not Gooda Walker were negligent in respect of the underwriting for Syndicate 290 in 1990, for on that issue may depend the result of a claim against Members' Agents by the Syndicate 290 Names in respect of 1990. That claim has been severed from this trial and will be determined at a later date.

The contractual duty owed by a Members Agent to exercise, through the Managing Agents and their servants to whom the task is delegated, reasonable skill and care in the conduct of the underwriting is only one of the duties owed by a Members Agent to its Names. Two of the Members Agents impleaded in this Action, the 36th and 38 Defendants, have each been sued by an individual Name for breach of a different duty - the duty to exercise reasonable skill and care when advising a Name which syndicates to join - see Brown v KMR Services Ltd Sword Daniels -v- Pitel (Unreported). In those Actions, which were tried together, it was common ground (though it may not be in this Action) that the duties owed by a Members Agent to its Name included the following:

(a) to advise the Name which syndicates to join and in what amounts,

(b) to keep him informed at all times of material factors which may affect his underwriting,

(c) to provide him with a balanced portfolio and appropriate spread of risk; a balanced spread of business on syndicates throughout the main markets at Lloyd's,

(d) to monitor the syndicates on which it places the Name, and to make recommendations as to whether the Name should increase his share on a syndicate, join a new syndicate, reduce his share, or withdraw.

(e) to keep regularly in touch with the syndicates to which the Name belongs, and

(f) to advise and discuss with the Name the prospects and past results of syndicates on which he could be placed.

In those Actions each Plaintiff claimed that his Members Agent had negligently advised him to join syndicates that involved a high degree of risk when the Plaintiff had requested to be placed on syndicates that involved a low degree of risk. The syndicates of which complaint was made were those which wrote London Market excess of loss ("LMX") business, including the Gooda Walker syndicates. It was not alleged that the underwriting carried on for these syndicates had been negligent - simply that the nature of the LMX business written involved a higher degree of risk than other types of business. In those Actions, the Defendant Members Agents unsuccessfully sought to challenge the assertion that LMX business was necessarily high risk, or alternatively, that they should reasonably have appreciated that it was high risk. In this Action it is one of the cornerstones of the Defence that LMX business was, and was generally known to be, high risk - indeed Mr Eder, QC, for the Defendants, has chosen to describe it as dynamite. It is quite plain from the evidence before me that this was not a perception of LMX business shared by many of the Defendants at the time. I have no doubt that there are many Plaintiffs who would understandably have felt outraged had they heard the plea being advanced on behalf of their Members Agent that Names had no cause for complaint because the type of business they had chosen to write was well known to be dynamite. The Plaintiffs have, however, for reasons which I can well understand, chosen to limit their allegations in this Action to the simple charge of negligent underwriting. It is the Names contention that, in underwriting on their behalf, the Gooda Walker underwriters disregarded certain fundamental principles of underwriting, with the result that the Names were subjected to excessive exposure to risk. Part of the debate in this Action has related to the extent to which it is proper for an XL underwriter deliberately to expose his syndicate to risk of loss. It is the Names' case that, in so far as this is proper, it can only be in circumstances where the underwriter makes it clear to the Names the extent of that exposure. In so far as I have to consider how the nature of the business of the syndicates was, or should have been, perceived by the Names, there is no room for distinguishing between the knowledge of one Name and the next. That question must be judged having regard to the knowledge and information that would have been generally available to Members Agents and their Names and on the premise that the Names were receiving competent advice from their Members Agents.

The Gooda Walker Group.

Until 1 January 1990, both Gooda Walker Ltd ("GWL") and Gooda & Partners Ltd ("GPL") performed the dual role of Members and Managing Agents. GWL was the Managing Agent for the non-marine syndicates, 290 and 164, and GPL for the marine syndicates, 298 and 299. GWL acted as Members Agent for 150 Plaintiffs, and GPL performed the same role for a further 80 Plaintiffs. Mr Anthony Gooda was Chairman of both companies, but was primarily concerned with the members' agency side of the business. Mr Derek Walker was the Managing Director of GWL and in charge of the non-marine underwriting. He was an ordinary Director of GPL. On 1 January 1990 the group structure was reorganised. Both Companies were placed under a holding company, Gooda Walker Holdings Ltd., of which Mr Gooda was Chairman. Mr Walker became Chairman of GWL, which took over responsibility as Managing Agent for all the group's syndicates. GPL became the Members Agent for all the group's names, with Mr Gooda as Chairman. Gooda Group Management Services Ltd provided administrative services to all the group's companies.

In the summer of 1991 Lloyd's withdrew its trading licence from all companies in the Gooda Walker group. Both GWL and GPL were placed in voluntary liquidation, and have no assets other than a contingent claim on error and omission ("E & 0") underwriters, who have purported to avoid liability on grounds of non-disclosure.

The Syndicates

Syndicate 290 was a non-marine syndicate writing almost exclusively excess of loss business. Mr Walker was the active underwriter for the syndicate and his son, Barrie Walker, was his deputy.

Syndicate 164 was a non-marine syndicate, whose active underwriter was Mr Edward Judd. Authority to write excess of loss business was, however, delegated to Mr Walker under a split stamp arrangement. This provided that a fixed percentage of the major categories of excess of loss business written by Mr Walker would be divided in fixed proportions between syndicate 290 and syndicate 164. In 1988 the split was 70:30. In 1989 this was reduced initially to 75:25 and, with effect from 30 March 1989, to 80:20. Mr Walker also made himself responsible for obtaining reinsurance cover to protect syndicate 164's excess of loss exposure. Mr Walker readily accepted that he had to bear the responsibility for the results of syndicate 164's excess of loss underwriting.

Syndicate 298 was a marine syndicate writing a broad range of excess of loss business, covering hull, cargo, liability, rigs, and whole accounts. It wrote a substantial amount of XL on XL business. Mr Stan Andrews was its active underwriter from the time that it started writing this type of business in 1983 until November 1989.

Syndicate 299 was a marine syndicate which wrote business at two boxes. At the cargo box Mr Malcolm Maxwell, the deputy underwriter, wrote cargo business and a very small amount of excess of loss business. At the hull box Mr Anthony Willard, the active underwriter wrote the hull business. The syndicate wrote a broad range of business which had always included an excess of loss account. With effect from July 1987 Mr Robert Hawkes, who was Mr Willard's deputy underwriter, took over the writing of the excess of loss account under Mr Willard's supervision.

LMX Business

The business of insurance provides a commercial mechanism for the mutual sharing of risk. Each insured pays a premium for insurance cover. The premiums thus paid by the many provide the resources for payment of indemnities to the few who sustain losses as a consequence of insured perils. Market forces should ensure that the level of premiums is sufficient to provide for payment of claims, the various expenses incurred by the insurers, and to leave a margin by way of reasonable reward for the insurer.

Conventional reinsurance is part of the market mechanism for this sharing of risk. It enables an insurer who has accepted a larger risk than he wishes to retain to share that risk with other insurers.

The balancing of premiums received on the one hand against indemnities paid on the other is threatened when a natural catastrophe strikes that causes losses to thousands or even millions of individual insureds. An individual insurer, particularly one who carries on business in a particular geographical area, may find that as a consequence of such a catastrophe claims greatly exceed premiums received. Against such a risk the insurer may wish to take out excess of loss protection.

The letters LMX stand for London Market Excess of Loss. The letters thus describe both the place where the business is transacted and the nature of the business. The place is the London insurance market. This embraces both Lloyd's and the companies which carry on the business of insurance in London. Excess of loss describes a particular type of reinsurance business. Indeed, as a matter of strict terminology it is perhaps incorrect to describe it as re-insurance; see the comments of Hobhouse LJ in Toomey -v- Eagle Star [1994) 1 Lloyd's Rep. 516 at pp 522-3. Insurers limit their exposure to losses by various types of reinsurance. They may reinsure a specific portion of a specific risk. They may reinsure a proportion of the business that they write, or a category of it, by a quota share treaty. Excess of loss reinsurance is designed to put a cap on an insurer's exposure to a particularly large individual risk, or an aggregation of risks that are all exposed to the same event. The insurer reinsures his liability for the consequences of a single event to the extent that this exceeds that part of the risk which he chooses to retain for his own account - his retention. It is customary to place this cover in layers , each layer being identified as a sum in excess of a figure which is the sum of the insurer's retention and the underlying layers of excess of loss cover. A low level layer of excess of loss insurance may be liable to be impacted fairly frequently by events, such as windstorms, which are relatively common occurrences. In general, the higher the layer of reinsurance the more serious, and the less common, will be the event that impacts the cover. Low layers of cover are sometimes referred to as 'working layers'. Excess of loss insurance of high layers is commonly referred to as 'catastrophe insurance' on the basis that the cover will only be likely to be impacted in consequence of a catastrophe, whether caused by man or by nature. Excess of loss business is commonly described by the letters XL.

If the layers of excess of loss protection of individual insurers are reinsured by a wide range of reinsurers, who are writing excess of loss protection for different types of business in different geographical areas, excess of loss reinsurance can form part of the mechanism for the overall sharing of risk. If, however, there are relatively few reinsurers who are prepared to write excess of loss reinsurance, the result will be not the dispersal but the concentration of risk. This is what occurred in the London Market in the 1980's. This period saw a steady and substantial increase in capacity at Lloyd's coupled with an increasing demand for excess of loss reinsurance - fostered in part by the growth in insured values of individual risks such as oil platforms. Excess of loss insurance had been the creation of Lloyd's after the great San Francisco earthquake of 1906 and in the 1980s a disproportionate share of excess of loss business was coming to London.

In a period without major catastrophe, excess of loss underwriters can make a good profit. In the first six years of the 1980's there was only one major global catastrophe - the hurricane 'Alicia' in 1983 - and the full effects of that took quite a few years to be felt in the LMX market. Prior to that there had not been a major global catastrophe since Hurricane Betsy in 1965. New Names and old were keen to share in the profits that it was believed could be made by writing excess of loss business.

The years 1980-1988 saw membership of Lloyd's rise from 18,552 to 32,433 Names, both increasing the stamp capacity of existing syndicates and resulting in the forming of new syndicates. This increase in stamp capacity was accompanied by a substantial growth in the excess of loss business being written at Lloyd's.

Mr Outhwaite, who was called by the Defendants as an underwriting expert, informed me that, of the 80 syndicates which commenced business between 1982 and 1988, 52 wrote some excess of loss business, including some which specialised in excess of loss, and only 28 wrote no excess of loss business. Mr Outhwaite also provided me with figures for 1989 of those of the Lloyd's syndicates and London companies which specialised in excess of loss business - that is those which made such business the major part of their book, those which wrote excess of loss as an adjunct to their main book and those which wrote no excess of loss business. Of the marine syndicates 33 fell into the first category, 67 into the second category and 19 into the third category. Of the non-marine syndicates 59 fell into the first category, 71 into the second category and 5 into the third category. Of the insurance companies which were members of the Institute of London Underwriters, 26 fell into the first category, 56 into the second category and 11 into the third category.

In a Report to Lloyd's made in 1992, Sir David Walker identified 87 syndicates which specialised in LMX business in 1989. These did not wholly tally with Mr Outhwaite's categorisation, but broadly supported the overall number of specialists in this field.

Mr Outhwaite said that it was almost impossible to make any greater degree of categorisation than that which he had attempted. In his statement he described the position as follows:

"Within each of the above categories, each syndicate wrote a different mix of business, depending on the outlook and judgment of the underwriter. Some wrote predominantly excess of loss reinsurance of other excess of loss underwriters (ie. excess of loss on excess of loss and/or reinsurance of excess of loss on excess of loss, so-called 'spiral' business); others wrote none (concentrating on excess of loss reinsurance of direct accounts). Some underwriters concentrated on lower layers, others on upper layers, depending on the underwriter's view of the relationship between the premium and liability. Some specialised in 'back-ups' (an excess of loss policy would normally have an upper limit each and every loss, subject to a limited number of reinstatements - a back-up policy was one which was designed to provide cover only after the reinstatements on another ('front-in') policy had been exhausted) - it was a question of where they saw the best bargain. The approach to underwriting manifested by the individual underwriters varied (in my experience according to the philosophy and nature of the syndicate, which the particular underwriter was best (indeed, uniquely) placed to judge."

Although syndicates 298 and 299 were marine syndicates, the excess of loss business that they wrote was not limited to marine risks. Mr Jewell, an underwriter who gave evidence for the Plaintiffs, explained how marine syndicates came to write non-marine business. The standard marine policy on cargo extends to cover shore risks before and after shipment. These are known as incidental non-marine risks. The grant of such cover led to some marine syndicates writing a small proportion - up to 10 % - of non-marine risks, notwithstanding that they were not incidental to marine risks, and this became an accepted feature of the marine market. So far as the reinsurer is concerned, the non-marine element in his book may significantly exceed 10 % for the following reason. The marine reinsurance syndicate may choose to underwrite pure non-marine reinsurance as its 10 % incidental non-marine. This means that the total non-marine element in its book of business will be greater than 10 %, because the 90 % of its account that is reinsurance of marine syndicates may itself contain the 10 % incidental non-marine business of those syndicates. When such a reinsurance syndicate buys its own whole account protection, the reinsurer will be subject to all that non-marine exposure. If that reinsurer, in its turn, retrocedes its risk to a further reinsurer which also includes in its book 10 % incidental non-marine, the non-marine content of its account will be even greater. This phenomenon explains the exposure that syndicates 298 and 299 proved to have to non-marine risks.

The Spiral

Spiral business plays a crucial role in this case. It is not clear on the evidence how many of those who wrote LMX business were prepared to write spiral business. I assume that most, if not all, of those who specialised in LMX business did so. All four of the Gooda Walker syndicates wrote spiral business. Mr Walker told me that his syndicate was one of only five or six major leaders of LMX business who were likely to be found on every LMX slip. They must all have written spiral business. The working of the spiral was complex, and whether by diagrams or in words it is only possible to attempt to describe it in a simplified form. My attempt is as follows.

Many syndicates which wrote XL cover took out XL cover themselves. Those who reinsured them were thus writing XL on XL. They, in their turn, frequently took out their own XL cover. There thus developed among the syndicates and companies which wrote LMX business a smaller group that was largely responsible for creating a complex intertwining network of mutual reinsurance, which has been described as the spiral. When a catastrophe led to claims being made by primary insurers on their excess of loss covers, this started a process whereby syndicates passed on their liabilities, in excess of their own retentions, under their own excess of loss covers from one to the next, rather like a multiple game of pass the parcel. Those left holding the liability parcels were those who first exhausted their layers of excess of loss reinsurance protection.

So far as the individual syndicate was concerned, the effect of the spiral was to magnify many times the impact of a particular loss. That is because claims were repeatedly made in respect of the same loss as it circulated in the spiral. I was told that claims in respect of the Piper Alpha loss exceeded by a multiple of about 10 the net loss that was covered on the London market.

This gearing effect did not, of course, result in an ultimate payment of a greater indemnity than the initial loss. As the loss passed through the spiral, however, it impacted repeatedly on successive layers of reinsurance cover, and ultimately concentrated on those reinsurers who found their cover exhausted.

There were at least two significant ways in which spiral business was written:

XL on XL : This described the grant of excess of loss cover in respect of an excess of loss account.

Whole account : An underwriter who took out, without exclusion, excess of loss cover in respect of his whole account would thereby obtain excess of loss cover in respect of that part of his whole account which itself comprised excess of loss business.

The spiral effect of claims was diminished or extinguished by individual retentions, whether before reinsurance protection commenced or after it had been exhausted, by co-insurance and by 'leakage' to reinsurers outside the London market, so that the extent to which catastrophe claims spiralled depended to a degree on the size of the loss, or more precisely that part of it which entered the London market. Thus, the higher the level of the layer of excess of loss protection, the lower the risk that it would be impacted. The effect of the spiral was, however, significantly to reduce the comfort that could properly be derived from being exposed only to what appeared to be a very high layer of loss. Another effect was to transfer from the insurers to the brokers a very substantial part of the overall premiums in respect of a risk, for on each excess of loss reinsurance, brokerage fell to be paid at a rate of 10 % of the premium.

The Losses

While I have to resolve any issues of principal that arise in relation to damages, I am not concerned with the quantification of the Plaintiffs' claims. At this stage of the Action the losses sustained by the Gooda Walker syndicates are relevant largely as a demonstration of the extent of the exposure to which they were subjected by the underwriting. Precise figures are thus not important.

The position opened by the Plaintiffs on the basis of GW Run Off figures as at 31 December 1993 was as follows:

Syndicate/Year Stamp Capacity Loss

298/1988 £42,700,000 £ 70,200,000

298/1989 £44,300,000 £323,600,000 (open year)

299/1988 £41,600,000 £ 22,100,000

299/1989 £38,100,000 £ 40,100,000 (open year)

290/1989 £70,300,000 £271,300,000 (open year)

290/1990 £57,700,000 £ 72,700,000 (open year)

164/1989 £57,200,000 £ 50,300,000

These figures differ somewhat from the figures shown in the 1992 and 1993 accounts. At this stage I need not - and cannot - resolve the discrepancies.

In his opening, Mr Vos told me that the losses sustained in these seven syndicate years amounted to 15 % of the losses made by the entire Lloyd's market in these years and that in 1989 the four Gooda Walker syndicates lost 30 % of the entire market's losses.

The Catastrophes

It is common ground that the losses made by the Plaintiffs were attributable in large measure to their exposure to a series of catastrophes. It is the Plaintiffs' case that the underwriters were negligent in leaving their syndicates exposed to these catastrophes. It is the Defendants' case that the sequence of catastrophes was unprecedented and unforeseeable and that no blame is to be attributed to the Defendants for the fact that the Plaintiffs were not protected from their consequences.

In this Action the Plaintiffs have focused on seven catastrophes. They contend that the first two are significant in that their impact should have alerted the underwriters to the risk to which they were exposing their Names. The subsequent five catastrophes are relied upon by the Plaintiffs as evidencing the consequences of the Defendants' breaches of duty for which they are liable in damages. This is the position in relation to the consequences of these catastrophes that Mr Vos opened: The first two catastrophes were:

Hurricane Alicia on the 17-20 August 1983. The overall insured loss attributed to this catastrophe by Sigma catastrophe reports published by Swedish Re, totalled $675 million. The Plaintiffs contended that Syndicate 290 suffered losses of about $70 million against reinsurance protection of $62 million. Losses were also experienced by the non-marine syndicates, although their reinsurance protection was not breached.

UK Windstorms 87J on the 16th-l7 October 1987. Sigma's figure for this loss is $870 million. As at 31 December 1992 Syndicate 290 estimated that it had suffered a gross loss of £85 million in respect of this catastrophe against reinsurance cover of £55 million. Syndicate 164 estimated its gross loss at £38 million against reinsurance cover of £21 million.

The other five catastrophes, which I shall hereafter refer to as 'the Five Central Catastrophes' were:

The Loss of Piper Alpha on 6 July 1988. At the start of this Action Syndicate 298 had an estimated loss of $220 million from this catastrophe against reinsurance of $110 million. Syndicate 299 was estimated to have lost $80 million against reinsurance of $56 million.

Exxon Valdez on 24 March 1989. Syndicate 298 was estimated to have lost approximately $200 million in respect of this catastrophe against reinsurance of $127 million, leaving net losses of $73 million. Syndicate 299 was also affected, but losses had not yet exceeded available reinsurance.

Hurricane Hugo on 15th -22 September 1989. Sigma's figure for this insured loss was $4,100 million. This catastrophe was estimated to have caused Syndicate 290 a gross loss of $385 million against reinsurance of $170 million, resulting in an estimated net loss of $215 million. It caused Syndicate 164 an estimated gross loss of $120 million against reinsurance of $64 million, resulting in an estimated net loss of $56 million. It caused Syndicate 298 an estimated gross loss of $380 million against available reinsurance of $161 million, resulting in an estimated net loss of $219 million and it caused Syndicate 299 an estimated gross loss of $85 million against available reinsurance of $82 million, resulting in an estimated net loss of $3 million.

Phillips Petroleum on 23 October 1989. Sigma's figure for this insured loss was $1,100 million. It caused Syndicate 298 an estimated gross loss of approximately $175 million against reinsurance of $84 million, leaving an estimated net loss of $91 million. It caused Syndicate 299 an estimated gross loss of $50 million against reinsurance of $41 million, leaving an estimated net loss of $9 million.

Windstorm Daria 90A on 24 January 1990. Sigma's figure for this insured loss was $4,600 million. It caused an estimated gross loss to Syndicate 290 of £200 million against available reinsurance of £89 million, producing an estimated net loss of £111 million.

None of these figures are agreed and revised estimates of some of them have been made since the Plaintiffs' case was opened.

Issues on Liability

As I have indicated, the only issue that I have to resolve in relation to liability is whether or not the active underwriters exercised the skill and care to be expected of reasonably competent underwriters. The Plaintiffs have pleaded that Gooda Walker were not merely vicariously liable for the faults of the active underwriters, but in breach of their own duties to manage the underwriting. In the event, however, there is no shortcoming alleged against the companies themselves that will not, if well founded, also involve a finding of fault on the part of the active underwriters. What are the faults alleged? The Plaintiffs have pleaded detailed particulars of principles that they contend the active underwriters should have followed and which they failed to follow. I quote from para 22 of the Points of Claim, as amended:

In the circumstances, any participation in the LMX reinsurance market required extreme caution and the exercise of a high standard of skill. Any competent insurer participating would have:

a) written only a limited volume and/or proportion of XL on XL business or business which included or constituted XL on XL business (by virtue of being a whole account protection of a cedent whose business was or included XL on XL or XL business) and/or

(b) paid close attention on a continuous basis to the following vital facts, and/or

(c) taken steps to plan, write and protect its business accordingly.

In particular, any competent insurer should have:

(1) obtained, so far as possible, information as to the nature of the risks comprised in the business which he was considering writing; this was in practice extremely difficult, when incoming business might include retrocessions at several stages away from the original risk; accordingly, it was necessary to assume the worst, when assessing the likely elements in the make-up of a piece or portfolio of LMX reinsurance business;

(2) planned, calculated and monitored total aggregate exposures on the various elements of LMX business written;

(3) planned, assessed and monitored, in the light of the information available and in any case of uncertainty on a worst case basis, the PML on the various elements of such business. In this respect:

(a) all XL on XL business would necessarily have to be assessed as having a 100 %;

(b) whole account business (particularly in the marine market, where exclusions of LMX or XL on XL business were not common) was being sought by insurers as a cheaper alternative to XL on XL and/or had to be assessed on the basis that it was likely to contain a substantial proportion of XL on XL business which would aggregate with other business;

(c) it would have been prudent to obtain from prospective reinsureds details as to their reinsurance programmes, which would have provided insight into their views of PMLs on the business being written;

(4) planned and placed reinsurances, normally in advance of any commitment to writing business requiring such reinsurance, whereby the syndicate would be protected in full in respect of the PMLs assessed in respect of the relevant elements of its portfolio;

(5) refused the business rather than accepting it and foregoing full or adequate reinsurance, when and if, as a result of premium rates, brokerage (customarily deducted at the rate of 10 % on each LMX reinsurance) and other factors, the rate on offer in respect of the business would not enable the financing of appropriate reinsurance to protect the relevant PMLs. In assessing the appropriateness of a premium rate in this context, the nature of the spiral meant that the height of the excess point under any particular reinsurance was no guide to the risk of it becoming a total loss. In the event of a loss exceeding the original (frequently depressed) retentions and such limited co-insurance as might exist, the whole loss would spiral upwards and hit the uppermost layer of any reinsurance - unless some other participant(s) in the spiral happened to run out of such reinsurance first;

(5A) refused the business rather than accepting it in circumstances where the cost of reinsuring it:

(a) deprived the writing of such business of its commercial viability, and

(b) would have resulted in foregoing either sufficient reinsurance to protect the LMX business written or sufficient cover for the balance of the account or foregoing both;

(6) recognised, taking account of the above, that marine syndicates choosing to write an "incidental" non-marine portfolio would thereby be exposing themselves to the effect of the spiral arising out of non-marine as well as marine risks.

The Points of Defence dealt with these pleas by denial or non-admission and did not seek to raise an alternative case as to specific principles to be applied in writing LMX business. Instead the pleading listed a number of matters which were alleged to be material when considering the standard of care to be applied in underwriting. These were as follows:

(i) Names at Lloyd's knowingly accept unlimited liability;

(ii) Underwriting is a risk business;

(iii) Underwriters assume risk in consideration of premium, and it is no part of their business to reinsure all the risks that they have assumed.

(iv) Lloyd's Regulations had no requirements for recording aggregate exposure or calculating probable maximum loss.

(v) In the late 1980's the entire insurance industry was hit with a series of unprecedented losses and the risk of such a concatenation of catastrophes within such a period of time could not reasonably have been anticipated and/or was so remote as not to be a matter which a reasonably prudent underwriter would necessarily guard against.

Mr Eder's opening for the Defendants made their case clearer. Excess of loss business is, and is known to be, an area of insurance that involves a higher degree of risk and of reward. The reinsurance written provides protection against the consequences of catastrophes. The proper premium for assuming such a risk is a matter of judgment. If catastrophes do not occur excess of loss business is very profitable. If they do occur, losses will result, which should be balanced by good profits from the catastrophe free years. The losses made by the Plaintiffs in this Action were the consequence not of bad judgment, let alone negligence or incompetence in underwriting, but of the unprecedented and unforeseeable sequence of catastrophes.

The pleadings and the opening of each party thus disclosed a fundamental issue as to whether or not the principles relied upon by the Plaintiffs had any application in the field of LMX underwriting. The precise nature of this issue became clearer as Mr von Eicken, the Plaintiffs' underwriting expert, gave evidence.

Mr Von Eicken

Before turning to the evidence of Mr Von Eicken I must rule on a submission made by Mr Eder that Mr Von Eicken was not competent to give expert evidence in this case.

Mr Von Eicken was employed for 35 years by the Munich Reinsurance Company ("Munich Re"). During this period Munich Re grew into the world's largest reinsurance company. While with Munich Re Mr Von Eicken undoubtedly became an expert in the field of catastrophe excess of loss insurance. In 1963 he was responsible for drawing up a set of general principles to be followed by Munich Re in relation to XL reinsurance. These principles remained in force, subject to amendment from time to time, up to the time that Mr Von Eicken retired from the company in 1990.

In 1967 Mr Von Eicken was asked to set up Munich Re's London Office. Thereafter he spent the majority of each working week in London. He served as General Manager and Managing Director of the company's Main Representation Office. Mr Von Eicken stated that during his 23 years in London he had constant contact with Lloyd's. He met frequently many Lloyd's underwriters and had extensive dealings with Lloyd's at Chairman level. By exchanging views on the reinsurance market he gained a clear understanding of the way Lloyd's syndicates operated. He negotiated individual participations by Munich Re in the reinsurance of Lloyd's syndicates, and was also involved in a section within the Munich Re (headed by its Chief Executive) which decided on the design of Munich Re's own reinsurance protections (mostly Catastrophe XL), which involved discussions with Lloyd's brokers and the underwriters at Lloyd's who were writing the Munich Re's covers.

Despite Mr Von Eicken's wide experience of reinsurance in general and excess of loss insurance in particular, his participation in the London insurance market and his general knowledge of the manner in which Lloyd's operates, Mr Eder argued that he was not competent to give expert evidence in this case. This was because the evidence clearly established that it was Munich Re's policy that the company should not reinsure excess of loss accounts. Such business was specifically stated to be undesirable in almost every edition of the written principles which Mr Von Eicken had originated. There was one exception to this. The guidelines in force from 1983 to 1986 stated that such contracts were "conditionally desirable where they are customary (e.g. on the North American market or in the London market for Lloyd's syndicates and fringe companies"). Mr Von Eicken told me that he discovered that Munich Re had started to write reinsurance of Lloyd's XL business in the mid 1980s and that he had put a stop to this because it was contrary to all the principles that Munich Re considered to be an essential basis for business.

In the light of this evidence, Mr Eder argued that Mr Von Eicken had no personal experience of LMX business, and in particular no experience of writing XL on XL cover, so that he was not qualified to comment on the principles that should be applied by the reasonably competent underwriter in that market.

Mr Von Eicken would not accept that there were special principles to be applied to the writing of LMX business. His constant theme was that LMX business was subject to the same basic principles as all reinsurance, indeed all insurance, and that he was highly qualified to speak to those principles.

Mr Von Eicken criticised the Gooda Walker underwriters for failing to write a balanced book of business, failing properly to calculate and reinsure their exposure and for failing properly to rate the business that they wrote. In the conclusion to his Report he commented that if the basic principles that he had outlined had been followed, it would have been possible for the Names not merely to avoid their losses but to make profits by writing catastrophe excess of loss business during 1988, 1989 and 1990. He cited by way of an illustration of this possibility, the results of Mr TR Berry's syndicate 536. Under cross-examination it became apparent that this did not accurately express Mr Von Eicken's opinion. It was his opinion that it was impossible to apply his underwriting principles when writing XL on XL business. That business was so opaque that it was impossible to make an accurate estimation of exposure or to estimate the appropriate rate for the risk underwritten. The only way to be confident of making money by writing spiral business was by skilful arbitrage and this was not what reinsurance was about. Arbitrage apart, writing spiral business was akin to gambling. The following passages from Mr Von Eicken's evidence clearly expressed his views about spiral business:

"The unfortunate fact is that in order to engage in this type of cover, you really have to do with a fraction of the knowledge, and a fraction of the experience and a fraction of the information which you would be needing in order to conduct catastrophe excess of loss business, and I maintain that it is just an aberration of catastrophe excess of loss business and should probably not have been touched by anybody it was an extremely incestuous market. Everybody in it was dealing indirectly with themselves and it had to be known that it was not a market and it had to be known that it was not a viable market and it had to be known that it would not be a market that had any hope of staying alive and that the profits that were made could only continue to be made as long as the hope came true that no major catastrophe occurred."

This approach to the spiral had not formed part of the Plaintiffs' case. In his written opening Mr Vos said this about the spiral:

"The Plaintiffs' case should not be misunderstood. It is not the Plaintiffs' case that the underwriters were negligent simply because they participated in the spiral. The Plaintiffs say that because of particular features of the spiral, it is necessary to exercise particular care when writing an account within it."

In his final speech Mr Vos submitted that Mr Von Eicken was probably right in his view that the spiral market was an aberration. The Plaintiffs, however, did not need to go that far. Their case was based on the disregard by the Gooda Walker underwriters of specific basic principles of insurance practice. If Mr Von Eicken was right about spiral business, so much the better, for that reinforced the Plaintiffs' case.

Mr Eder for his part also argued that it did not matter whether the spiral market, viewed as a whole, was an aberration. Even if it was, a competent underwriter could properly enter that market in order to make profit for his Names. Mr Von Eicken had deliberately kept out of that market and was not competent to express opinions about the conduct of those who practised within it.

Mr Eder asked Mr Jewell, whose role as a witness I shall deal with shortly, about Mr Von Eicken's ability to give expert evidence in this case. He commented:

I believe the fundamental principles of reinsurance are the same throughout the world, and that they start with a fundamental understanding of the direct business. Therefore, although a European reinsurer, or an insurer who is buying - who is not buying or who is buying - his reinsurance from the London market, may not have firsthand experience of the way the market is operating because he has chosen not to operate within it, does not necessarily mean that he does not understand what is going on ...

Somebody outside the market with a good understanding of the fundamental business, may actually be able to see the wood from the trees clearer than those within.

I have reached the same conclusion as Mr Jewell. Mr Von Eicken is qualified to express an opinion as to the principles that should be followed when writing excess of loss business. He is also qualified to express an opinion as to whether those principles are compatible with the writing of spiral business. I have not been persuaded by the evidence that it is appropriate to ringfence that minority of London Companies and Lloyd's syndicates that specialised in writing excess of loss business, and to treat them as members of a separate market. They were members of the London insurance market, and the London market is itself part of an international insurance market. Mr Von Eicken has had very considerable experience of the international insurance market, of the London insurance market, of the customs and practices of Lloyd's, and of the writing of excess of loss business. The fact that it was the almost unbroken policy of the Munich Re to avoid writing XL on XL business does not disqualify Mr Von Eicken from giving expert evidence in this Action.

Before turning to the expert evidence, it is right that I should state the impression made on me by those who gave it.

Mr Von Eicken

Mr Von Eicken was not an ideal expert witness. He showed a keen appreciation of his own abilities and a contempt for any challenge to his views. This attitude was not justified, for in his report Mr Von Eicken had, in places, made sweeping statements of underwriting principle which could not be justified. He adopted, throughout, a vigorously partisan approach and could scarcely ever be induced by Mr Eder to make a concession, however clear it might be that a concession was due.

Mr Jewell

Mr Jewell was not called by the Plaintiffs as an expert. He is a Director and General Manager of City Run-Off Limited ("CR0"). CR0 has contracted with GW Run-Off Limited ("GWR0") to administer the run-off of the Gooda Walker Syndicates. From 1962 to 1979 Mr Jewell was employed, first by New Zealand Insurance Company Limited and then by Norwich Winterthur Reinsurance Company Limited, for the most part as a deputy marine underwriter. From 1979 to 1989 he was employed by Claremount Underwriting Agency Limited as an active underwriter for Claremount Syndicates. He underwrote all the reinsurance account, consisting of marine, non-marine and aviation XL business for syndicates, which had at their peak, a combined stamp capacity in excess of £100 million. The XL account was predominant.

In October 1989 Mr Jewell joined GPL and took over the running of Syndicate 299 from Mr Willard. When, at the end of 1989, Mr Andrews was asked to resign as active underwriter of Syndicate 298, Mr Jewell agreed to replace him on the understanding that he would not write any new business other than lines already promised. When, in the autumn of 1991 GPL and GWL went into voluntary liquidation, CR0 was formed by Bankside Underwriting Agencies Limited to handle the day to day running of the syndicates. Mr Jewell was one of a number of Gooda Walker staff who were retained by CR0.

Mr Jewell was called by the Plaintiffs to give factual evidence about, inter alia, the exposures, reinsurances and losses of the Gooda Walker syndicates. Mr Eder chose, however, to cross-examine him at length on matters of expert evidence. I am glad that he did, for Mr Jewell impressed me as a witness of high standing, ability and experience in the field of excess of loss reinsurance.

I found the evidence that he gave in relation to the practice of excess of loss reinsurance more compelling than that of the witnesses who had been called specifically to give expert evidence on this topic.

Mr Outhwaite

Mr Outhwaite commenced work at Lloyd's in 1957 at the age of 21. He worked as personal assistant and latterly as deputy underwriter to Mr Roy Merrett and became responsible for most of the excess of loss underwriting for his syndicate. In 1972 he took over as underwriter for the syndicate jointly with Mr Stepehn Merrett. From 1974 until 1989 he ran his own syndicate. This was a general marine syndicate with an excess of loss book of 25-30 % of premium income. The syndicate was known as a leader of excess of loss business. Mr Outhwaite has served on many committees at Lloyd's, including the Joint Excess of Loss Committee which was formed in 1987. This summary illustrates the depth of Mr Outhwaite's experience of Lloyd's underwriting and of excess of loss in particular. I found him, in general, an impressive witness. Nonetheless I did not find him wholly objective. While he did not demonstrate the partisan enthusiasm of Mr Von Eicken, he was, I felt, careful in so far as he could to avoid answers which might further the Plaintiffs' case.

Mr Fryer

Mr Peter Fryer has worked in the insurance industry from 1953 until his retirement last year. From 1969 to 1985 he was Managing Director of North Atlantic Insurance Company Limited. After that company was acquired by the British National Insurance Group he was first of all Managing Director and subsequently Deputy Chairman and Chairman of British National, responsible for the group's insurance and reinsurance operations throughout the world. From 1986 until 1993 he was Chief Executive of the Charter Reinsurance Company Limited of London. Mr Fryer was well placed to give expert evidence from the viewpoint of a London reinsurance company. I found him an impressive and on the whole a balanced witness.

The Principles Applicable to Excess of Loss Underwriting

It is not possible to resolve the issue of liability in this case without having regard to the peculiar features of spiral business. I propose, however, to consider those features in the context of the specific allegations made by the Plaintiffs. The first step must be to decide, with the assistance of the expert evidence, what, if any, basic principles of underwriting practice should have been adopted by those who were writing LMX business. I shall then consider whether, and if so how, those principles could be applied by a writer of spiral business. Finally I shall consider the specific allegations directed at each of the active underwriters.

Exposure and Balance

The Points of Claim allege in some detail the principles of underwriting practice that the Plaintiffs contend should have been followed, but do not clearly allege what the competent underwriter should have attempted to achieve by the application of these principles. Their case was, however, clarified by Mr Von Eicken's Report. This emphasised the importance of writing a balanced account.

A balanced account, adopting the terminology of Mr Von Eicken, is one where the claims that the underwriter has to meet, together with all his expenses, are covered by the premiums that he receives, so that an appropriate profit remains. According to Mr Von Eicken, this balance should be achieved by the excess of loss underwriter in two ways. First he should limit the exposure of his syndicate to the consequences of a single event by spreading the syndicate's business over a number of different classes of business (eg. rig covers) and over a number of different geographical areas. In this way the underwriter can ensure that the potential exposure to a single event is balanced by premiums from business not exposed to that event. To achieve this balancing exercise in practice the underwriter will have to restrict his exposure in respect of some classes of business or exposure areas. In order to do this he must be aware of his aggregate exposure in relation to each class of business or exposure area.

A single event will be unlikely to expose an underwriter to liability in respect of 100 % of his aggregate exposure in respect of a class of business or exposure area. For this reason it is possible to adopt a more realistic and sophisticated approach to the balancing of an account. The underwriter selects the most serious catastrophic event that he can foresee as a possibility for which he must make provision and estimates the maximum loss that he will be likely to have to meet should that event occur. This amount is described as the Probable Maximum Loss or PML. It is thus the PML, rather than the aggregate, to which the competent underwriter normally has regard when considering the balance of his account.

The other tool which, according to Mr Von Eicken, the excess of loss underwriter uses to achieve balance is reinsurance. By reinsuring his exposure to his PML, or part of it, in a cost effective manner he can help to ensure that if a catastrophe strikes his liability will not overreach the net premiums that he has retained.

It was, I believe, initially the Plaintiffs' contention that the competent excess of loss underwriter should achieve a balance of this sort year by year for each year of account so that, unless an untoward event occurred, each year would end in profit.

It has always been the Defendants' case that the type of balance that I have just described, achieved year by year, is not compatible with writing excess of loss business on any scale. In so far as it concerns the concept of achieving balance by dispersing the business written over different classes or geographical areas of risk, the Defendants' stance on this point was supported by a number of witnesses including, most authoritatively, Mr Fryer. He told me that it was impossible to write an excess of loss account of any significance covering catastrophe perils and to achieve internal balance by such means. Most of the demand for excess of loss cover came from the United States and Europe and insufficient business could be written in other parts of the globe to balance exposure to those areas. Charter Re had also tried to achieve balance by writing a mixture of marine, aviation and non-marine business and by writing first, second and third tier reinsurance. The problem of imbalance could be reduced in this way, but not removed. Mr Fryer agreed, however, that it was possible to smooth potential losses by reinsurance.

If it is impossible for an excess of loss underwriter to achieve internal balance without reinsurance, I cannot see how reinsurance can wholly remove exposure to loss if the reinsurance rates properly reflect the risk transferred. It was Mr Fryer's evidence that the excess of loss underwriter could not expect to achieve a balance between claims and premium income year in year out, but would have to accept that there would be loss making years to be balanced by premium income in other years.

Mr Jewell said that in 1987, 1988 and 1989, when the market was terribly soft, it was "highly possible" to devise a reinsurance programme which was completely comprehensive. He told me that in 1988 he had been able to buy reinsurance cover for his syndicate at a rate which enabled him to write business for his Names which did not expose them to risk but enabled them to make profit from premium differentials. He added, however, that to attempt to make profit without risk was not a professional move "because the professional move starts with how much exposure you are prepared to write".

At the end of the day Mr Vos retreated from the suggestion that an excess of loss underwriter should necessarily aim to achieve a balanced account year by year. His submission, rather, was that, if the underwriter was going to subject his Names to the risk of making an overall loss, he should aim to restrict the exposure to a predetermined limit and make it quite plain to Names that they were exposed to loss to that extent.

In the course of closing submissions I suggested to Mr Eder that it was common ground that a competent underwriter had to restrict exposure to an appropriate level. Mr Eder would not accept this. He submitted that there was no inherent limit on the exposure that an excess of loss underwriter could properly incur. This reflected a theme in his argument that Names who joined Lloyd's deliberately accepted unlimited liability and that Names who joined excess of loss syndicates knew that they were thereby indulging in underwriting that involved higher than normal risk. Mr Eder's submission received a degree of support from Mr Outhwaite. He said that it was theoretically possible to write a book of business in the excess of loss field where all covers were exposed to the same loss event and that there was nothing, in theory, to prevent this. This, however, was subject to the important proviso that this policy was made plain to the Names. So far as exposure was concerned Mr Outhwaite said that the governing principle was that the underwriter should not expose his Names beyond that level which he aimed for and predicted. This statement was not far removed from a passage in Mr Von Eicken's evidence when he agreed with Mr Eder that one could not criticise an underwriter for leaving an uninsured gap, whatever its size, provided that he did so on purpose, knew exactly what it was and informed his Names of what he was doing. Mr Jewell said that the starting point was that the underwriter should decide how much risk he was prepared to write and expose his syndicate or company to and that having taken that decision he had to restrict exposure to that level by reinsurance. This evidence was put to Mr Outhwaite who agreed with it.

My conclusions in relation to exposure are as follows. The fact that a Name who joins Lloyd's deliberately agrees to expose himself to unlimited liability does not mean that he anticipates or accepts that when he joins a syndicate the active underwriter will deliberately expose him to the risk of such liability. On the contrary the Name will reasonably expect the underwriter to exercise due skill and care to prevent him from suffering losses. In many categories of insurance the Name will reasonably expect the underwriter to plan to procure profits, year in year out. The fact that syndicates are reconstituted each year does not, however, make it mandatory for an underwriter to conduct business in this way. I accept Mr Outhwaite's evidence that underwriting must be conducted as an ongoing business. There is no reason in principle why an underwriter should not write business on the basis that net losses will be made in some years that are balanced by generous profit levels in the other years. If, however, an underwriter is deliberately to expose his Names to suffering losses from time to time, he must make sure that the Names are aware of this and of the scale of loss to which they will from time to time be exposed.

This conclusion is one which I would have reached as a matter of common sense if unassisted by expert evidence. A Name at Lloyd's, with the assistance of his Members Agent, will want to structure his underwriting business in a manner that accords with his means and with his attitude to risk. Syndicates at Lloyd's tend to specialise in different categories of business. Some categories of business involve less risk of loss, and commensurately less prospect of reward, than others. A Name needs to know the nature of the exposure that he is likely to run by joining a particular syndicate if he is to be able to structure his underwriting business in an appropriate manner.

It was the Plaintiffs' case that, in the absence of a specific warning, a Name would not anticipate that a particular excess of loss syndicate would write in such a way that loss making years would be a natural consequence from time to time. A loss making year would only result from some untoward event or sequence of events for which the underwriter could not reasonably have been expected to provide. The Defendants challenged this case. As I have already indicated, it was Mr Eder's contention that any Name should have known that writing excess of loss business could well result in very high losses - such business was "dynamite". The extent to which Names should reasonably have anticipated that by joining the Gooda Walker syndicates they rendered themselves liable to suffer significant losses from time to time is a matter to which I shall revert at a later stage of this judgment.

Planning

In his initial Report, Mr Von Eicken stated that a competent excess of loss underwriter would necessarily make detailed written plans of his underwriting policy and how he was going to achieve it. Aggregate limits, PMLs and reinsurance protection would all be planned in advance and those plans would be committed to writing.

Mr Outhwaite challenged this evidence. He said that an underwriter would certainly prepare detailed plans, albeit not necessarily in writing, when setting up a new syndicate. Thereafter there would be no formal planning. The underwriter would run the syndicate as an ongoing business, keeping his underwriting and his general strategy under review and reacting to changes in the market. Only if the capacity or composition of a syndicate changed significantly between one year and the next was it necessary to plan how the underwriting should accommodate these changes. When Mr Eder asked him whether it was not strange that underwriting, as a professional exercise, should be conducted on an ad hoc basis without formal planning he replied:

"I do not find it strange at all, it was the environment in which I was brought up in Lloyd's; it is how Lloyd's has always worked."

Mr Jewell said that he would not necessarily expect an underwriter of a syndicate of the same type as the Gooda Walker syndicates to have a detailed underwriting plan. He added, however, that in his opinion all should have such a plan. He said:

"I would have felt very uncomfortable for my own job, let alone for the Names that I was writing for, if I did not have a structure and a plan and an idea of what I was attempting to put together... I do not believe I could have purchased a reinsurance programme intelligently without having an idea of what I was going to write and what I was trying to protect."

Written underwriting plans have, I believe, now become the norm at Lloyd's. This was not the case during the period with which I am concerned. Absence of written underwriting plans cannot be treated as evidence of lack of competence. What was, in my judgement, important was not that a plan should be written down, but that a plan should exist and be followed. In particular, any competent underwriter had to formulate and follow a policy as to the extent to which Names were to be exposed to the risk of loss. Mr Eder suggested that an underwriter could properly make no attempt to limit his Names' exposure beyond the restriction imposed by Lloyd's regulations on premium income. Even if that were correct, and the model is a purely hypothetical one, an underwriter could not be competent in adopting such an approach unless he did so after careful consideration and unless he made it absolutely clear to his Names that this was his policy. In fact all the witnesses in this case accepted that an excess of loss underwriter would have a policy as to the amount of exposure to which his Names would be subjected. In my judgment it was a fundamental principle of excess of loss underwriting that the underwriter should formulate and follow a plan as to the amount of exposure that his syndicate would run. During the period with which I am concerned such a plan would normally - if not inevitably - involve restricting the syndicate's gross exposure by reinsurance in order to attain the planned level of net exposure. Mr Outhwaite accepted that the underwriter would buy reinsurance cover with a view to protecting exposure resulting from specific projected aggregates. He said, however, that such projections were not inflexible. The underwriter might accept more or less business depending upon how attractive the particular business was. I have no difficulty in accepting this, subject to two provisos. First the underwriter must be monitoring his business properly so that he knows what he is doing. Secondly he should not make so radical a departure from his policy on exposure as to betray the reasonable expectation of his Names.

Aggregates and PMLs

In order to monitor the exposure that results from the business he writes, the excess of loss underwriter must be aware of his aggregates. He has the advantage that each piece of business he writes is subject to an express limit of liability. To calculate his exposure to a single event he needs to know how many of the covers that he has written are exposed to the risk of a claim should that event occur. He thus has to divide into different categories the covers that can aggregate. In practice this is normally done by a system of coding the different categories. The more carefully the business is recorded under appropriately chosen codes the more confident the underwriter will be able to be as to the limit of his exposure to a single event. As I have already explained there will be some categories where it is unlikely, or indeed inconceivable, that a single event will result in a claim on every cover. In respect of those categories the true exposure will be, not the aggregate, but the PML. The estimation of the PML has to be made by the application of judgment to the data available.

There is no real dispute between the parties as to the need to assess PMLs, although there was some debate as to appropriate terminology. As Mr Eder put it in his final written submissions:

"It was common ground that an underwriter must form some judgment of his exposure to one loss event in order to decide what reinsurance to purchase."

This leads me to the topic of reinsurance.

Reinsurance Policy

In his final submissions Mr Vos contended that a competent excess of loss underwriter should adopt a reinsurance policy that included the following elements:

(1) The underwriter should know the PML that his syndicate will be exposed to in the event of the worse catastrophe.

(2) He should then work out what net exposure he is prepared to run in that event.

(3) He should reinsure the balance.

(4) He should retain a meaningful retention at the bottom (and preferably the whole retention should be at the bottom).

(5) He should match the reinstatements on his reinsurances to the reinstatements that he is allowing in his writings, so that he is protected against frequent catastrophes.

Mr Eder made no substantial challenge to the first two propositions. It was his case that they would form part of an on-going and flexible review rather than an inflexible plan. As to the third proposition, Mr Eder accepted that this is what the competent underwriter should attempt to achieve. Mr Vos had originally asserted that the competent underwriter would have his reinsurance in place before he started writing cover. That proposition was patently absurd for, if applied to all, no-one could write any business. Writing excess of loss cover and obtaining such cover are two operations which can neither be done sequentially nor synchronised. There must inevitably be a risk that the insurer finds that he cannot obtain as much reinsurance cover as he would wish at rates which he considers fair.

The fourth proposition was not common ground. It had originally been the Plaintiffs' case, supported by Mr Von Eicken's Report, that the competent excess of loss underwriter would decide how much he wished to retain and reinsure everything in excess of that amount up to his PML. Under cross-examination Mr Von Eicken conceded, however, that there was no reason why the insurer should not retain exposure at the top above his layers of reinsurance. This was subject to the proviso that he did so deliberately in the knowledge of the amount of that exposure. Mr Jewell had already given evidence to like effect. He said:

"The wise approach in my experience would be to have a core reinsurance put together that you know you can achieve and buy the rest of your protection either at the bottom or at the top end of your exposure. The important factor is to know what your exposure is and how much of it you are prepared to carry. It is almost irrelevant in my view as to where you carry that."

It seems to me that in principle this must be right. If an underwriter has determined to retain a specific net exposure, those layers of exposure which he protects by reinsurance should depend upon the view he takes of the terms of the reinsurance on offer. The evidence suggests, however, that it may be difficult in practice to obtain reinsurance cover if one cannot demonstrate that one is retaining a layer of risk below that cover.

MATCHING REINSTATEMENTS

Mr Vos' fifth proposition introduces a topic of some complexity. In his final speech Mr Eder described it as "almost a diversion - a variation on a theme" and commented, with some justification, that as the point had not been specifically pleaded, no one had really focused on it as a separate issue.

In that section of his witness statement which deals with reinsurance protection, Mr Jewell states:

"Reinsurance policies cover layers of exposure... Each policy will provide coverage for a particular sum insured (eg. £5 million excess £5 million) and may be reinstated a limited number of times. The number of reinstatements available indicates the level of horizontal protections."

The excess of loss underwriter has to consider not merely the extent of his vertical exposure to a single loss event, but the extent of his horizontal exposure - the number of events resulting in claims to which he may be exposed. This Action has concentrated largely on vertical exposure. The pleadings make no express reference to horizontal exposure and the expert reports do not focus on this topic. It is, however, one of importance having regard to the manner in which the Plaintiffs advance their claim to damages and to the Defendants' reliance upon a 'concatenation of catastrophes'. This section of my judgment is derived largely from evidence provided by Mr Jewell.

Horizontal reinsurance cover is classically provided by a clause in the policy of reinsurance which provides that, if a claim is made, the cover will be reinstated so as to protect against a further loss event upon payment of an additional premium. The policy may provide for one or a number of such reinstatements. This is, however, only one form of horizontal cover. A policy may provide cover, up to a limit for each event, for a sequence of events for a single premium. Back-up policies can be purchased which only take effect after a specific number of loss events have occurred. A 'top and drop' policy may be used either to provide top layer protection or to provide back-up protection at a lower layer.

A syndicate's horizontal exposure, and the reinsurance protection purchased to cover this, will depend upon the nature of the business written. If working layers or low layers are written it may be reasonable to expect a significant number of loss events and to buy reinsurance protection to cover them. Thus, for example, Syndicate 290 had in place, as at 1 January 1989, a fairly complex structure of whole account X/L cover at lower layers which included cover for $150,000 excess of $100,000 for 10 loss events after an initial 10 such events had occurred, and thereafter cover for $50,000 excess of $50,000 for a further 20 loss events. Syndicate 298 had in place, as at 1 July 1989, whole account cover providing for 24 reinstatements at the lowest level of $250,000 excess of $250,000. At levels such as this I cannot see any place for the concept of "matching reinstatements" advanced by Mr Vos. The concept does make sense, however, when one is considering true catastrophe insurance written in the spiral.

Mr Fryer pointed out the difference between the need to devise an appropriate pattern of horizontal cover and the more specific desirability of matching reinstatements when writing high level catastrophe cover:

In normal situations you get a gradient of losses which may be relatively flat, in other words a large number of small and medium sized losses, or it may be relatively speaking, a smaller number of medium and large losses. Now the number of reinstatements you need and at which level within the programme would vary dramatically between those two cases

Q The point is a more general one that if one is writing a book of high level catastrophe reinsurances, XL on XL, you will not as a general matter write your higher large covers with two reinstatements and place your reinsurance programme with one reinstatement.

A I would not, no.

An underwriter who makes a practice of granting to his fellow excess of loss specialists more reinstatements than he purchases from them will be subjecting his Names to an obvious area of unprotected exposure. Mr Outhwaite gave the following evidence about reinstatements:

"... in general terms the business that you are writing on excess of loss will be limited to reinstatements, either one or two. Generally speaking, your outward reinsurance will be similarly placed, that is to say, it will have the same number of reinstatements. In general terms you will find the reinsurances outward are done on the same basis as inwards, so to that extent reinstatements match. It is not necessarily that they will always match under practical circumstances, because as risks attach throughout the year reinsurances may be exhausted by some losses and because of different attachment dates, you have claims on other policies.

Thirdly, there are areas where you write business with more than two reinstatements. Working layers are habitually done with more and it may be that you take a view that you only protect yourself up to some of them and you are prepared to run the risk of reinstatements beyond that point.

Mr Vos: But you accept that it is desirable to write your book of business and reinsure it on the basis that the reinstatements match in the way that you have just explained?

A The way that I have explained this is what will normally happen, certainly.

Q It is desirable to do that?

A Yes.

Q It is obviously desirable, because it protects yourself against frequency of losses except in excess of your reinsurance?

A Yes.

In a subsequent passage of his evidence, Mr Outhwaite added these further comments:

Q I am suggesting that the numbers of [losses], although they may have been an unusual number together, it was something for which an insurer should have been prepared for?

A I quite disagree, you have to make some estimate of the probability of these things, as I pointed out earlier. If you think that the probability of a loss is one in ten during the course of the year, the probability of two of them occurring is only 1 in 100 during that year. So you get to the point where you do not take the possibility of the numbers of occurrences into account, you get beyond any ordinary sort of calculation.

Q Your position is that you should not be prepared, if you are a reinsurance underwriter, for more than one loss?

A I am not saying more than one loss, but there comes a point where it goes past what I would call reasonable estimation of a probability of loss.

Q Where is the point?

A I cannot tell you, it would depend very much on how underwriters see it.

Q Three losses?

A I do not know.

Q Two losses?

A I just do not know, Mr Vos, it is impossible to say.

Q The real point here, Mr Outhwaite, is this, that these losses that occurred were unusual in the sense that there were more than there had previously been in a short space of time, but they were not so unusual, were they, that it was reasonable for the underwriters to ignore the possibility of their occurring?

A I do not know the degree that you can say that they ignore it. I am saying that any underwriter has to make some estimation of what possible losses are going to occur, and you have to stop somewhere. Underwriting is a business where you take a risk."

This evidence was not inconsistent with answers given earlier by Mr Jewell in relation to this topic:

"Q How should an underwriter who has written policies with reinstatements take this into account in relation to his reinsurance programme?

A I stumble with the answer - what he should do is recognise the exposure. He should measure the exposure - I am sorry to be repetitive to some extent - and then decide how much of it he wishes to keep for his syndicate. It is easy to recognise the exposure.

Q All I am trying to do is to explore how that answer fits into the reinstatement problem.

A The unprecedented or whatever - the succession of losses can only impact the syndicate if it did not cover for those losses."

Counsel for the Plaintiffs put to Mr Willard, Mr Andrews and Mr Walker that it was desirable to match reinstatements. Each of them agreed to this. Mr Andrews and Mr Walker claimed to have done so. Mr Willard accepted that he had not.

My conclusions are as follows. The competent excess of loss underwriter had to give careful consideration not merely to his vertical exposure but also to his horizontal exposure. This was true whether he was writing high level catastrophe business in the spiral or low level reinsurance of direct business and it is axiomatic that the underwriter had to plan his pattern of reinsurance protection. In relation to true catastrophe reinsurance, to grant more reinstatements than those purchased would produce an area of obvious exposure.

I accept Mr Outhwaite's evidence that there might be circumstances in which an underwriter would deliberately take such a risk. Nothing suggests that such circumstances applied in the present case, or that any of the Gooda Walker underwriters were deliberately taking such a risk. They should, in my judgment, have been following a policy of matching reinstatements in relation to the catastrophe excess of loss business that they were writing in the spiral.

A policy of matching reinstatements in this way was not, however, calculated to cover all horizontal exposure to a sequence of catastrophes. The position was graphically described in the September 1991 edition of Chatset:

"In 1989, apart from Hugo, which hit most syndicates in both Marine and Non-Marine, there were five other major losses, namely Exxon Valdez, Phillips Refinery, Atlantic Richfield, San Francisco Earthquake, Newcastle Earthquake. Any one of those may have necessitated calls on a syndicate's general protections and most reinsurance programmes allow for two reinstatements of the original policy limit, ie. three losses. Once the policy has met those three losses the fourth and every subsequent loss 'cascades' over the top and all claims are paid gross. Policies are also written covering 'cascade' losses, to add to the misery of those underwriters writing them, who may themselves already have cascaded."

The manner of impact of competing claims on writers of spiral business with limited reinstatements is, as Mr Jewell explained, complex and potentially capricious, particularly in the marine market where it is the custom and practice that losses are taken in settlement date order for reinsurance purposes. A sequence of catastrophes can expose a syndicate to losses, not because claims from the individual catastrophe outstrip the vertical cover in place, but because they outstrip the horizontal cover. The September 1992 edition of Chatset summarised the position as follows:

"The problems caused by "Hugo" were only part of the story. There were other catastrophes in 1989 which may not have exceeded the vertical reinsurance cover, but the number of catastrophes impinging on reinsurance may have exhausted the number of reinstatements available."

In the course of the trial Mr Jewell gave evidence that claims resulting from catastrophes other than the Five Central Catastrophes were likely to exceed the reinsurance cover available and result in net losses - Enchova, Lockerbie and the Australian Earthquake are examples. It is not axiomatic that exposure to those catastrophes is a consequence of a failure to 'match reinstatements'. They indicate simply that horizontal cover was not adequate to cater for the series of catastrophes that occurred. The pleadings make no specific allegation in relation to deficiency of horizontal cover and, apart from a few questions to the underwriters about matching reinstatements, their evidence does not cover this area. The significance of this is a matter to which I shall revert.

Rating

The experts were agreed that it is a fundamental principle applying to all insurance business that the underwriter must satisfy himself that the premium received is commensurate with the risk assumed. If not so satisfied he should not write the risk. This basic principle is subject to pragmatic exceptions. Sometimes it will be politic to accept an unattractive risk as a loss leader to a broker who has more attractive business in his gift. The majority of business written tends to be renewal business. Such business, once rejected, will find another home and not return. This again may justify the short term expedient of accepting a risk at an unattractive rate. It has not been suggested that these exceptions to the general principle that the rate should be adequate for the risk have any significance in the present case.

The experts agreed that there were four elements that made up an adequate premium:

(1) The basic risk premium;

(2) A loading for random fluctuation;

(3) A loading for administrative expenses and brokerage;

(4) A loading for profit.

In most classes of insurance past experience provides a good guide for assessment of the level of premium necessary to cover these elements. Past claims statistics both of the market in general and, where appropriate, of the insured in particular, provide a reliable guide to the nature of the risk. Market forces tend to ensure that rates are adequate.

Rating catastrophe excess of loss business is much more difficult. The reinsurer has to assess the likelihood of a single event resulting in loss to the reinsured in excess of a particular level. Any sensible attempt at this exercise requires detailed information as to the nature of the reinsured's own book of business so that a view can be taken of the measure of exposure of that business to a single catastrophic event. Much more difficult is the assessment of the likelihood of such a catastrophic event. Catastrophes do not occur regularly and, as the world develops, the potential consequences of a natural catastrophe become more weighty.

In these circumstances the experts were all agreed that historical experience is of little assistance in rating if all the underwriter is looking at is a period during which there has been no serious catastrophe.

The difficulty of assessing an adequate rate for writing an excess of loss risk has, in the past, been alleviated by a phenomenon known as "payback". If a claim is made under an excess of loss cover the premium is subsequently increased so that, in the space of a few years, the loss that has been paid is recouped.

In October 1988 a Working Party under the chairmanship of Mr Graham Lyons produced a Report on "Excess of Loss Reinsurance of Lloyd's Syndicates and London Market Companies". The aim of the Working Party was to set down the nature of the London market and the special considerations which apply to LMX business. In relation to rating the Report made this sanguine comment:

"Leaders of LMX business tend themselves to be London underwriters and have a close understanding of the various exposures written generally in the market and specifically by individual syndicates or companies. LMX premiums are reviewed each year by these leaders who take into account all the factors in setting the new annual premiums for each excess of loss contract. The leaders have questionnaires designed to draw out each reassured's exposures, both on individual risks and in the aggregate on the various classes of business written."

Such a questionnaire would provide detailed information about the underwriting business of the reinsured. The Report included suggestions by an actuary as to how excess of loss business might be rated. I did not understand that the actuarial approach suggested was one that at any material time was adopted in practice by excess of loss underwriters. The actuary drew attention to the problems inherent in using payback as a means of adjusting rating.

In his Report Mr Outhwaite had this to say about rating:

"On excess of loss business, premium is normally expressed as a 'rate on line' - ie. as a percentage of the total liability undertaken by the reinsurer on a particular piece of business. Rating is not an exact science: it is not possible in any particular case to say that a premium is right or wrong. It is a matter for the underwriter's judgment whether the risk is worth taking at the premium offered. The premium would be set by the lead underwriter on a slip in accordance with his assessment of the risk, rates in the market generally, and what he thought the business offered would bear. Others would then be approached by the broker to complete the slip and would subscribe or not according to their own evaluation of the risk, often much influenced by their view of the lead underwriter".

In my judgment any competent underwriter should have satisfied himself upon reasonable grounds that the premium adequately reflected the risk before writing excess of loss business. This is subject to some comments I shall have to make in due course about arbitrage. The underwriter's judgment, to which Mr Outhwaite refers, should, it seems to me, have involved considering the type and level of catastrophes that would result in a claim on the cover and the likelihood of such a catastrophe occurring. This was not an easy exercise, even in the case of direct reinsurance, but unless it was attempted the rating exercise was no more than guesswork.

The Problems Posed by Spiral Business

The reinsurer of a direct account is in a position to obtain the data that he needs in order to undertake the difficult task of assessing the risk. The risk is relatively transparent. The writer of first tier excess of loss reinsurance can, at least in theory, obtain data from the reinsured about the nature of his account. But beyond that stage the risk becomes totally opaque. One thing is certain and that is that the same event may give rise to claims under each excess of loss book that has been reinsured. All the XL on XL business aggregates with a PML of 100. What is uncertain is the level of catastrophe that will enter the spiral and burn through the individual syndicate's layers of reinsurance cover. In such a situation it seems to me that Mr Von Eicken is plainly right to say that the underwriter lacks the necessary data to make an informed assessment of the risk.

I am also persuaded that Mr Von Eicken was justified in describing the spiral as an aberration. Its effect was to concentrate the losses that came to London under direct excess of loss covers on a small proportion of those writing excess of loss business. The evidence in this Action demonstrates that at least some of those who were left bearing the losses had not intended to hold themselves out as insurers of last resort. The cost in brokerage of putting in place the complex contractual latticework of the spiral must have transferred from the underwriters to the brokers an inordinate proportion of the original gross premium that came to London.

The existence of the spiral enabled direct excess of loss insurers to write business in the confident expectation that they could obtain reinsurance cover. The evidence that I have heard suggests that this led some of them to be less than scrupulous in assessing and rating their own business. This led Mr Walker in 1988 to lead a market initiative which resulted in a requirement that those reinsuring excess of loss business should retain an exposure of 10 % by way of co-insurance. This must have dampened the spiral effect, but only to a limited degree.

The losses with which this Action is concerned destroyed the spiral and brought the market to its senses. Before Hurricane Betsy a spiral had developed in the London market and according to Mr Walker the losses that followed that catastrophe demonstrated the effect of the spiral. The lesson had been forgotten by the 1980's.

Rating Structure

It is clear on the evidence that the growth of the spiral in the 1980's was accompanied by the development of an irrational rating structure. This was due I believe to an approach to rating which ignored the manner in which the spiral greatly reduced the extent to which risk reduced as the layer of cover rose. The practice was first to rate the lowest layer and for subsequent layers to be rated at progressively lower rates, each rate being a reduction on the rate of the preceding layer. In the result the premium rate, at least of the upper layers, fell far below what might have been adequate to reflect the risk. The highest layers were, for a time, being written at 2 % or even 1 % rate on line. The steps by which the rate reduced to this modest percentage from the rates charged on the lower layers were far too steep and failed to have regard to the manner in which once a loss entered the London market it could spiral up through it.

Mr Eder sought to persuade Mr Jewell to agree that the rating structure reflected the view at the time of the whole market as to appropriate rates, and that the defect in the rating structure only became apparent with hindsight. Mr Jewell would not accept this. He said that many underwriters who wrote spiral business did not understand the difference between reinsuring direct business and writing XL on XL. They did not, however, represent the whole market or even, necessarily, the majority of the market. Many may not have written business at all because they recognised the irrationality of the rating. Mr Jewell said that he himself recognised it and used it to his advantage.

Should the Gooda Walker Underwriters have Appreciated the Nature and

Effects of the Spiral?

Some LMX underwriters undoubtedly had an appreciation of the spiral, and made that appreciation known. One of these was Mr Emney, Director and Chief Underwriter of Charter Re. In October 1987 he gave an address to the Institute of London Underwriters pointing out the gearing effect of the spiral on claims in respect of Alicia.

The Lyons Report quoted the following views on the LMX market expressed by an LMX underwriter in a talk to the LMX working party on 14 June 1988:

"XL-on-XL. This is reinsurance of other reinsurers' LMX business. The same loss inevitably goes back and forth creating a spiral. The main beneficiary is the broker with his 10 % brokerage. There is a gearing element: it is worth writing if the premium rates are higher than the cost of outwards reinsurance.

An LMX underwriter tries to get a decent handle on his exposures. His aim will be to protect to the top of his aggregate exposure, if possible, depending of course on what this will cost. Exposure can be under-estimated eg. 87J (the October 1987 UK hurricane).

If an underwriter can obtain an 'edge' (ie. if his net position is such that he is expected to make a profit) he will exploit it. However, the margins for profitability are very small, particularly when the brokerage of 10 % each time is considered. Hence in general there can only be a very few winners in the market - most players will be losers.

The information currently provided via the standard questionnaires is not great. An LMX underwriter really needs much more detailed information particularly on exposure. Without this information the underwriter's role is effectively entrepreneurial."

In April 1988, shortly before the Piper Alpha disaster, Mr Outhwaite delivered a paper at a reinsurance conference which proved to be prophetic. It is so germane to the issues that arise in this Action that I propose to quote some lengthy extracts from it:

"... there is no doubt that initially at any rate LMX reinsurance appears to increase the size of the market significantly. An underwriter would be prepared to accept a certain amount of catastrophe reinsurance from the United States and, estimating its PML, would accept more if he in turn was reinsured and I have no doubt this is just what the major responsible syndicates and companies do.

Catastrophic losses would have passed into the LMX market and here one comes across an extraordinary phenomenon. It is very rare to find an LMX underwriter who attempts to run any significant liability at all. The only market for a reinsurance of an LMX underwriter is to all intents and purposes the LMX market itself. The amounts placed overseas are insignificant and are in any case often reinsured on a similar basis back in London. There are at least two major firms of LMX specialist brokers who pride themselves in placing all their business within the London market. LMX underwriters have on the whole been conspicuously successful at a time when the underwriters of what one might call primary risks have been in difficulty. The explanation for this obviously lies in the fact that there has not been a catastrophe loss large enough to test the market to destruction. I am not an adherent to the conspiracy theory of history, as it is largely based on paranoia, but it does seem to me that there is an almost universal ostrich-like unwillingness to face facts.

In a paper at a previous seminar, I gave an example from the drilling rig market. The maximum insured value of a platform on the North Sea is around $2 billion. At least 60 % of that is insured in London net, and looking through the list of marine companies and Lloyd's I made an estimate of what each would maintain his retention was should the platform be a total loss. On the most generous basis this would not exceed one hundred million dollars. There is no dispute that if there were such a total loss, the London market would pay $1.2 billion net. Where will the other one thousand one hundred million dollars come from? It will come from the LMX market. It is an oddity that in the ordinary way a large risk is placed widely around the market in order to spread the load, it then gets spread further into the reinsurance market and indeed initially that is what would happen in the example I am quoting. However, once the claim was large enough, it would reach the point where every LMX underwriter believed he was fully reinsured, the loss would become increasingly concentrated amongst the few major syndicates and companies operating in the retrocession LMX market and would continue to go round and round until eventually each one exhausted his reinsurance protection and the loss would become net. In the event of a major windstorm where the initial loss to London could be much greater, the effects would be even more severe, but precisely the same thing would apply.

If this is true, which it clearly is, why is this not recognised? There are several reasons, most of them concerned with the perfectly natural reluctance to rock the boat. How does the underwriter justify the position? Firstly one can pretend that it is not right. He can say that the market came through Alicia unscathed (although I estimate that that is looking somewhat doubtful at the moment) and that the aviation market came through 1985 with flying colours. Of course this is a question of quantum. Had Alicia been only twice as large no one would be in any doubt that what I am saying is correct. There has not been a collapse of a platform in the North Sea. Secondly an informed underwriter can say yes this is all very true in general, but I myself am adequately protected. I do not believe that any underwriter who is accepting significant amounts of LMX reinsurance of excess of loss accounts can be adequately protected. As the liability accumulates, every policy written would have to be catered for under an outgoing reinsurance which is unlikely to cost significantly less. Thirdly, an underwriter may accept that a substantial amount of the liabilities he is accepting are not protected by reinsurance on the basis that it is unlikely that a loss of sufficient magnitude would happen at all, and if it did it is acceptable that his Names or shareholders could suffer a loss. That is a perfectly reasonable approach, provided that the Names and shareholders are aware that this is the case.

From various discussions I have had with Lloyd's Agents I am fairly certain that they do not believe that any underwriter in Lloyd's is operating on this basis. I would enter a slight caveat here. I am sure of this in the ease of marine underwriters, but slightly less so in respect of the other markets. There are certain minimum premium levels below which it is just not possible to place high level non marine catastrophe reinsurance or aviation reinsurance, and the rates seem to me to reflect sensible appreciation of the risk, and are at a level where it is quite reasonable to take on a significant net liability. On the other hand, in the marine market high level catastrophe rates are absurdly low, and are based solely on comparison and precedent and bear no relationship whatsoever to the risk being run. Indeed there are many examples where non marine XL reinsurance is being included with marine exposure and is being placed at a mere fraction of the cost (say around a quarter) of what the non marine market itself would require, and without any co-insurance warranty ...

I cannot believe that it is healthy, either for the LMX market itself or for the direct market, for the position to have developed to the stage where a large part of it would collapse in the event of a major catastrophe. Such a collapse would affect a vast number of companies and people, underwriters, brokers and the like, undermine confidence in Lloyd's which looking at a not altogether improbable combination of events, may default on a Lloyd's policy for the first time in its history with a substantial withdrawal of membership and capacity. Of course, it may never happen, but because there is the possibility that it may, is why there is insurance in the first place. So, for an underwriter to work on the assumption that it will never come to pass is hardly tenable."

Mr Outhwaite told me that in this paper he had deliberately exaggerated his views in order to be provocative. In my view they give a valuable picture of the ability of an underwriter to appreciate the nature and the inherent dangers of spiral business.

The irrationality of the rating structure was also appreciated by some. The views of the underwriter quoted in the Lyons Report included the following comments on rating:

"Rating is usually in relation to exposed premiums expressed as a rate on the premium base...

The actual rates charged for LMX on LMX and for LMX excluding LMX on LMX (eg. a generals programme) are of a similar pattern. For LMX-on-LMX the rates vary from about 45 % rate on line for the lower layers of a programme to around 5 % to 10 % rate on line for the top layers. This may not in fact be correct. For LMX-on-LMX a claim has to be rather larger to hit the bottom layer than for a straight generals programme. However, due to the incestuous nature of the LMX market and the spiral effect, once a claim has entered the LMX-on-LMX market there is a fair chance that it will go right through the programme. Hence the rate on line graph might need to be much less sloped and the "correct" rates could well be from 30 % to 15 % or even a fixed 22.5 %!"

The underwriter's appreciation of the nature of spiral business, as evidenced by these extracts, contrasts with that of Lord Strathalmond who was called by the Defendants in order to give expert evidence of the viewpoint of the Members Agent. He told me that prior to Piper Alpha he did not appreciate the existence of the spiral, let alone its implications.

In my judgment the Gooda Walker underwriters should have shared the appreciation of the spiral enjoyed by Mr Emney, by the Lyons Report underwriter and by Mr Outhwaite. This was a business in which they chose to specialise and they should have given the most careful thought to its nature. I shall deal later with the standard of skill and care to be expected of those underwriters. As a broad proposition, any professional who engages in a particular speciality can be expected to demonstrate the level of skill and care appropriate to that speciality. No reason has been suggested by the Defendants why the Gooda Walker underwriters should not have made the same appraisal of spiral business as Mr Outhwaite and I can think of none.

Could a Competent Underwriter Write Spiral Business?

Mr Von Eicken's view was that a competent underwriter should not write spiral business at all. During the material period Mr Outhwaite came close to following this course, for his spiral business was essentially restricted to a modest amount of back-up cover. He told me that he reinsured in excess of his PML's and wrote a balanced account.

I propose at this stage to quote some extracts from Mr Outhwaite's Underwriters Reports, for they exemplify the attitude that the Plaintiffs contend should be taken by the prudent excess of loss reinsurer. In 1986 he wrote:

"Excess of loss reinsurances continue to contribute a major part of our premium income. We underwrite a very wide-ranging account encompassing almost every class of business. Generally we do not accept reinsurances of other excess of loss underwriters as this type of policy will accumulate in the event of a major catastrophe."

In 1988 he wrote:

"A substantial part of the cost [of reinsurance protection] relates to the purchase of "catastrophe" excess of loss protection. In most years no claim will be made on these policies. They are however fundamental to the success of the syndicate. The limit of the cover purchased is fixed by reference to our estimate of the largest loss the syndicate is likely to suffer."

In 1987 he wrote:

"Excess of loss reinsurance policies form a major section of our accounts. We write a deliberately broadly based account: it includes purely marine contracts, covering individual sections of an account or a whole account, non-marine, aviation, satellite and even an occasional motor or livestock policy. Almost the only risks that we will not consider are reinsurances of other excess of loss accounts."

In 1989 he wrote:

"The account is written across areas of the business, with the intention of achieving as great a spread of business as possible so that no individual claim in any particular area can have too much influence on the overall results. This is in marked contrast to some excess of loss accounts underwritten in London where the normal practice is to accept huge aggregate liabilities in any given area (often many times the syndicate's premium income capacity) and therefore rely entirely on their own reinsurances (or on there being no catastrophe loss) ... A substantial part of the [reinsurance] cost, relates to the purchase of catastrophe excess of loss protection. In most years no claims will be made on these policies, they are, however, fundamental to the success of the syndicate. The limit of the cover purchase is fixed by reference to our estimate of the largest loss that the syndicate is likely to suffer. It is a matter of judgment and there can be no certainty that the amount is absolutely sufficient, it must be borne in mind that to buy too much protection can be as harmful to the long-term profitability of the syndicate as to buy too little."

These statements suggest that it was the degree of exposure involved that disinclined Mr Outhwaite to write XL on XL business. He told me that he stopped writing that business in the eighties because he did not think the rate was good enough. I asked him to explain how he assessed the rate and this is what he said:

"Excess of loss on excess of loss is, as has been made clear by virtually everybody here, a very opaque business. It is very difficult to imagine there is such a thing as a proper, a correct rate. You accept business, however, in the knowledge that it is not just for one year, but it is in the context of a number of years, and if you think that in one year a risk is worth taking, having taken all the things that you know of, worth taking at a rate of, say, 10 per cent, you are also comforted by the fact that if things happen you will probably get 20 per cent or 25 per cent next year for the same risk. So it has its compensations. But it seems to me, if you get to a point where you are offered risks at rates of 2 per cent, or something like that, that it was not, in my view, worth taking the risk. Other underwriters were taking it, of course, because I was shown slips at those rates and I also did not think it was worth it in order just to reinsure them."

In the course of his evidence Mr. Von Eicken referred to a few underwriters who played the spiral game correctly. I asked him how they did this. He gave me this answer:

"If you were able to reinsure yourself out at ... a better rate than you took the reinsurances in, such as the classical case where you took in low level covers, at a high rate on line, and had a meaningful retention and a quota share incidentally, as Mr. Berry had, and then reinsured it out, obviously at a high level, at a low rate on line, you would be playing correctly."

In such a situation, providing there was a satisfactory differential, the actual rate did not matter so that the opacity of the risk did not prevent successful underwriting. Mr Berry underwrote for Syndicate 536 which commenced underwriting in October 1982 for the 1983 account. The account consisted of marine excess of loss business emanating from London and overseas. Mr Von Eicken exhibited Mr Berry's annual underwriting reports, starting with that made in 1989, as an illustration of a competent approach to writing excess of loss business. These reports indicate that for each year, up to at least 1991, Syndicate 536 made a profit. In his 1992 report Mr Berry made this comment about the spiral:

"The spiral does not increase the size of a loss, at the end of the day, it only redistributes it. In all matters of redistribution there must be winners and losers, and ultimately, those who understand the nature of the beast eventually succeed."

In the following year he reassured his Names:

"I know that we understand our business although some Names and Agents have, I fear, taken the incorrect view that no insurer understands it!"

The reports indicate the following aspects of Mr Berry's underwriting policy:

(1) A careful monitoring of exposure. In 1989 he stated: "I personally maintain a comprehensive accumulation exercise from which I calculate the amount of excess of loss protection the syndicate should prudently carry".

(2) A detailed reinsurance policy. Thus in 1990 he reported that there was a shortfall of 2.42 % of the total reinsurance coverage that he had sought to purchase.

(3) The avoidance of net exposure when he considered the premium rate inadequate. Thus in 1991 he recorded that, when writing the 1988 account: "I considered that at the general level of rates which then prevailed in the market it was not worth taking the syndicate into the risk taking business. I was looking at a risk/reward ratio which was 100 per cent and virtually no prospect of reward."

(4) Discrimination in respect of those to whom he was prepared to grant cover. He wished to be satisfied that they were not simply relying upon excess of loss protection to avoid the necessity of making a proper appraisal of the merits of their own business.

(5) A willingness to carry net exposure when he considered that the rate was favourable.

(6) Careful covering of his reinstatement obligations.

(7) Express notice to his Names of taking a deliberate decision to expose them to risk to take advantage of favourable rates.

Mr Berry's reports indicate that he wrote a substantial proportion of spiral business. It is possible that his syndicate remained profitable by luck rather than good judgment. His reports lead me to conclude, however, that this was not the case and that, despite the opacity of the business, it was possible for a competent excess of loss underwriter who followed the approach of Mr Berry to write a book which included spiral business. I am inclined to think, however, that this could only be achieved by someone who fully understood the spiral and who deliberately took advantage of the disparity of rate for low level and high level layers of business.

WAS THE UNDERWRITING NEGLIGENT?

Unlike Mr Berry's syndicate, the Gooda Walker syndicates made very substantial losses in the period 1988 to 1990. It does not follow from this that their underwriters were incompetent or negligent. It is now necessary to turn to consider in detail the manner in which they conducted their underwriting business.

The Standard of Skill and Care

In his Opening submissions, Mr Eder advanced the following principles which he contended applied in the present case:

(1) The standard of skill and care to be exercised by a member of a professional calling is the degree of skill and care ordinarily exercised by reasonably competent members of that profession or calling.

(2) The existence of a common practice over an extended period of time by persons habitually engaged in particular business is strong evidence of what constitutes the exercise of reasonable skill and care.

(3) In situations which call for the exercise of judgment, the fact that, in retrospect, the choice actually made can be shown to have turned out badly is not of itself proof of a failure to meet the necessary standard of care.

(4) The Plaintiffs cannot show a failure to meet the required standard of skill and care unless the error on the part of the underwriter was such that a reasonably well informed and competent member of the profession or calling could have made it.

I accept each of these propositions. They merit, however, a degree of elaboration. The first proposition does not remove from the Judge the determination of the standard of skill and care that ought properly to be demonstrated. As the authors of the third edition of Jackson and Powell on Professional Negligence point out at p 39:

"It is for the Court to decide what is meant by "reasonably competent" members of the profession. They may or may not be equated with practitioners of average competence... Suppose a profession collectively adopts extremely lax standards in some aspect of its work. The Court does not regard itself as bound by those standards and will not acquit practitioners of negligence simply because they have complied with those standards."

The fourth proposition is based on a passage of the speech in Lord Diplock in Saif Ali -v- Sidney Mitchell [1980) AC 198 at p 220:

"No matter what profession it may be, the common law does not impose on those who practice it any liability for damage resulting from what in the result turn out to have been errors of judgment, unless the error was such as no reasonably well informed and competent member of that profession could have made. So too the common law makes allowance for the difficulties in the circumstances in which professional judgments have to be made and acted upon."

This passage was dealing essentially with the question of judgment. The Plaintiffs' case is not that errors of judgment were made, but that judgment was not exercised at all in that the underwriters never acquired the data on which that judgment might have been based.

As to the second proposition, the particular business with which this Action is concerned is the business of underwriting. More particularly, this Action is concerned with one area of underwriting, excess of loss reinsurance. At the heart of the Action lies one aspect of excess of loss underwriting, the writing of spiral business. That was a business that developed rapidly in the period of the eight years or so that led up to the events with which this Action is concerned. Only a relatively small proportion of Lloyd's underwriters specialised in writing spiral business. The London market no longer writes spiral business - at least on the scale and in the manner which developed in the last decade. In these circumstances I do not consider that one can automatically regard the practices of those who wrote spiral business as constituting strong evidence of what constituted the exercise of reasonable skill and care. It is necessary to approach this case with the possibility in mind that, for many involved, a significant involvement in spiral business may not have been compatible with competent underwriting.

Mr Eder has urged me to avoid the dangers of hindsight and of being wise after the event. He is right to do so. Nevertheless the underwriters in this case were putting their Names at risk to the tune of many millions of pounds. The heavy responsibility that that entailed entitled those Names to expect that their underwriters would exercise an appropriate amount of wisdom both before and during the underwriting that they transacted on behalf of their Names.

Successful excess of loss underwriting, particularly if it included significant spiral business, called for an approach to underwriting which differed from that of conventional direct insurance and called for a high degree of skill. In the course of his final submissions Mr Eder seized on the statement of the underwriter quoted in the Lyons Report that "there can only be a few winners in the market; most players will be losers". He submitted:

"... your Lordship put to me that [the market] did not make sense. It was an aberration. But each individual underwriter trying to be as competitive as possible, was using that market to transfer risks from him to some other party, with a view to making a profit.

In hindsight, ultimately, if you have one major loss or, certainly as in 1989, more than one major loss, ultimately, as was predicted in that article, it is quite interesting, it was done at about the time of Piper Alpha, there will be more losers than winners.

If one is in the practice of arbitrage, as we are all guessing Mr Berry might have been, it may be that he would be one of the winners. But that does not mean that the market was negligent at the time, necessarily. Nor, indeed, that any one particular underwriter working within that market was negligent."

In my view to justify participation in the spiral market the underwriter had to have good reason for believing that he would rank among the winners. As Mr Berry remarked "those who understand the nature of the beast eventually succeed".

Before looking at the individual syndicates, I propose to consider certain allegations of breach of duty that apply to each of them.

Inter-Syndicate Reinsurance

The Plaintiffs plead, as an allegation of breach of duty, that the three Syndicates 190, 298 and 299 entered into contracts of reinsurance between themselves. This is alleged to be objectionable because of the proportion of Names who were on more than one Gooda Walker syndicate.

The following facts were put to Mr Andrews and not challenged in relation to the 1989 year of account:

Of the Plaintiffs,

230 Names were on Syndicate 298 alone;

359 Names were on Syndicate 298 and one other;

368 Names were on Syndicate 298 and two others;

413 Names were on all four syndicates.

The Points of Claim plead the following particulars of inter-syndicate reinsurance:

Syndicate 298 purchased from other syndicates $5.76 million of cover in respect of the 1988 year and $8.53 million in respect of the 1989 year.

Syndicate 299 purchased from other syndicates $2.56 million of cover in respect of the 1988 year.

Syndicate 290 purchased from other syndicates $1.45 million of cover in respect of the 1988 year.

Where reinsurance takes place between two syndicates on which there are common Names, the consequence is to dilute the transfer of risk which the reinsurance is intended to achieve. The extent of the dilution depends upon the degree to which the syndicates are made up of common Names. Where syndicates have a large proportion of common Names, so that the dilution of the transfer of risk becomes significant, it will usually make sense to avoid inter-syndicate reinsurance. I imagine that such a situation is only likely to arise in cases, such as the present, where a number of syndicates are managed by the same Agency. It does not seem to me that it is possible to condemn inter- syndicate reinsurance as constituting, per se, a practice which breaches the duty to exercise reasonable skill and care in conducting the business of underwriting. It must be considered having regard to all the relevant individual circumstances - not least its potential consequences for the Names involved. Those consequences are particularly difficult to assess where the inter-syndicate reinsurance formed part of the complex web of reinsurance transactions that created the spiral. All that one can say is that the inter-syndicate reinsurance that took place in this case accentuated the incestuous nature of the spiral transactions and diminished the comfort that the active underwriters could take from the apparent protection afforded by the reinsurance cover that they bought.

Reciprocal Reinsurance

I have reached a similar conclusion in relation to the pleaded allegation that the underwriters were negligent in doing

"a substantial volume of reinsurance business with other syndicates and/or companies who were at one and the same time placing reinsurances with them and accepting reinsurances from them, with the effect that reinsurances were written and placed which effectively cancelled each other out and/or contributed to the LMX spiral"

Evidence supported the allegation of a degree of reciprocal reinsurance. This did not, however, have the affect of cancelling out the mutual transactions, for the book of business in respect of which reinsurance was taken out was not identical in the case of each participant, nor were the layers identical. Reciprocal reinsurance did, however, contribute to the spiral. It demonstrated the working of the spiral in its most immediate and incestuous manner. It was not negligent per se but it contributed to the problems facing the Defendants in conducting spiral business.

The Individual Syndicates

I turn now to consider the underwriting carried on on behalf of the individual syndicates. In the course of doing so I shall summarise the information provided to Names in the annual Reports and Accounts in order to see the extent to which, if at all, these forewarned Names of the exposure to which they were subject.

SYNDICATE 198

Mr Andrews

Mr Andrews was born in 1930. He left school at the age of 14 and joined the marine department of Matthew Wrightson. He joined the Gooda Group in 1954 as an assistant on the marine box, working on the hull side, where he remained until 1967. In 1967 he moved to the Gale Group as the active underwriter for their aviation business. There he developed a particular interest in excess of loss cover. In 1974 he joined the Swiss company, Turegum, as their aviation underwriter, and became a recognised aviation excess of loss leader. He was also involved in the writing of Turegum's marine account. In 1980 he rejoined the Gooda Group to become active underwriter of the non-marine syndicates 290 and 387 during Mr Derek Walker's absence. He also wrote the excess of loss accounts for aviation syndicate 295 and non-marine syndicate 164. From 1983 he took over as the active underwriter of syndicate 298, which had previously been the baby syndicate of syndicate 299. The plan was to turn this into an independent syndicate writing marine excess of loss business. In the first year 130 Names joined the syndicate, giving a stamp capacity of £1,505,000. The syndicate rapidly expanded. In 1984 its stamp capacity had grown to £5.2 million. In 1985 to £11.3 million, in 1986 to £21.6 million and in 1987 to £32 million. In 1986 Mr Andrews started writing an aviation account, and developed a position as a leader of aviation risks.

Mr Andrews was not a satisfactory witness. He began his evidence with an apology to the Names for the losses which they had suffered. He went on to point out that he had himself suffered losses which had led to his ruin. He went on to say:

"Having said all that I cannot agree with the allegations made against me in this action by the Names. In my view I at all times maintained good practices on my syndicate and acted in what I reasonably perceived at the time to be the Names' best interests."

I have no doubt that Mr Andrews did, at all times, believe that he was acting in the best interests of his Names. He has, however, been unable to accept the fact that he might be responsible for their downfall. He stated more than once that the loss of Piper Alpha was "unthinkable". The evidence suggests that when that loss occurred Mr Andrews found himself unprepared to present to the Board of Gooda Walker or to Members Agents the clear impact of that loss upon his Names. He was no more prepared in the witness box to accept that any aspect of his conduct of the syndicate's business could be criticised without the benefit of hindsight.

Policy in Relation to Exposure

In his report Mr Outhwaite stated:

"I have considered the aggregate exposures accepted by Syndicate 298 in 1988 and 1989; considered the PMLs that it may be reasonable to adopt and the reinsurance purchased by the syndicate. The net liabilities retained by the syndicate are in excess of what I believe to be reasonable."

Mr Outhwaite went on to say that this view did not imply criticism of Mr Andrews. Net exposure was a matter of judgment. He went on to state:

"Mr Andrews, in common with many other underwriters, did not reinsure up to the extent of his assessment of the PML. It was his conscious policy to run a significant proportion of the risk written."

This last comment is not supported by the evidence. Mr Andrews made it plain that it was not his policy deliberately to subject his Names to the risk of the losses that they sustained. The excess of loss reinsurance that he placed to protect the syndicate was subject to a retention of $250,000. He told me that had a Name asked whether there was any likelihood that a loss would exceed that amount his answer would have been no, and he added:

"That was our concept as well, to be really honest. We did not think the Name would sustain a loss beyond the protection we had in place."

Earlier he had said that he was aware that there were exposed aggregates, but did not feel it necessary to refer to this fact in his Report to the Names "because I felt the PML factor of it was potentially covered by our reinsurance programme". I accept that Mr Andrews did not believe that his underwriting was exposing his Names to the risk of loss. The issue is whether that belief is one that any reasonably competent underwriter could have held.

It was Mr Andrews' evidence that he monitored his aggregates, assessed his PMLs and took out adequate reinsurance to cover those PMLs. All of these assertions are challenged by the Plaintiffs.

Aggregates

Mr Andrews kept no record of his aggregates. He did, however, meticulously record premium income and estimated signed lines in premium income analysis books. Mr Andrews told me that in the renewal system he would use these records in order to calculate aggregates. He took the books home at the weekend to do this. The method that he adopted was to divide the total estimated signed line for each class of business by the total estimated premium income in order to obtain a rate on line for that class of business. Thereafter he would apply that rate to premium income in order to arrive at the aggregate. This exercise was done in relation to the two classes of business which were most significant from the viewpoint of aggregation. These were class 10, which was XL on XL business and class 01, which was whole account business. Mr Andrews told me that if a whole account cover included a significant proportion of XL on XL business, he would code it as class 10. Class O1 was business which he considered to have little chance of being invaded by XL on XL losses. He considered that this was a substantially risk free code.

The Plaintiffs challenged Mr Andrews' assertion that he monitored his aggregates in this way. For reasons that I shall explain I do not think it matters whether he did or not. On balance I do not believe that he did. The following are my reasons.

(1) The syndicate records contain no reference at all to aggregates.

(2) Mr Andrews' description in the witness box of how he monitored aggregates was confusing and contained inconsistencies.

(3) Mr Andrews' deputy, Mr Wilson, told the Poynter Committee that aggregates were not measured. When asked about this Mr Andrews said that Mr Wilson might not have been aware that he was monitoring aggregates. I found that suggestion almost inconceivable.

(4) Had Mr Andrews wished to monitor his aggregates the obvious way to do so would have been to keep a running account of the exposure written rather than adopting the tortuous method described by Mr Andrews.

The possibility remains that Mr Andrews kept some form of check on the aggregate exposure of his class 10 business, which he considered to carry a high risk of aggregation. I can see no reason, on his evidence, why he should have monitored the aggregates of his whole account business, which he considered essentially risk free. He does not suggest that he monitored the aggregates of any other codes.

PMLs

It was Mr Andrews' evidence that he attributed a PML of 100 % to his class 10 business and a PML of 15 % to his class Ol business. If he kept a check on his class 10 aggregates, I can accept the likelihood that he would have assessed their PML at 100 %. I have said that I do not accept that Mr Andrews monitored his class Ol aggregates and it follows, of course, that I do not accept that he assessed the PML of this class at 15 %. The explanation that he gave me for choosing this figure I found unintelligible. The figure of 15 % first appears in the transcript of Mr Andrews' evidence to the Poynter Committee. When that transcript was received by Reynolds Porter, the Solicitors acting for Mr Andrews, they wrote querying the 15 %. I think it not unlikely that, as the Plaintiffs suggest, the origin of this figure is a typographical error.

If, contrary to my conclusion, Mr Andrews did attempt to calculate his aggregate exposure in respect of his class Ol risks and assessed his PML in respect of that aggregate at 15 %, this of itself demonstrated a high degree of incompetence. It was Mr Andrews' evidence that he believed that the pattern of his exposure did not change as his stamp capacity increased. It was pointed out to him that between the time of Piper Alpha and of Hurricane Hugo his XL on XL exposure increased by 70 % and his whole account exposure increased by 45 %. He agreed, with some reluctance, that he had no idea of the size of these changes and added that if they were the true figures he was extremely disappointed and had to agree that his system was not as strong as he thought at the time. He was also forced to agree that if he had monitored his aggregates accurately he would have discovered that in 1988 his class 10 exposure was $80 million and his class Ol exposure in excess of $200 million. Mr Outhwaite commented that the system for monitoring aggregates, as used by Mr Andrews, appeared to be useless.

It is hard to understand how Mr Andrews could have believed that the whole account business which he wrote and classed as Ol was virtually risk free, giving rise to a PML of no more than 15 %. It was put to Mr Outhwaite that Mr Andrews could not sensibly, as a competent underwriter, have assessed the PML of 15 % on his whole account business. Mr Outhwaite answered that he found it difficult to imagine how anybody could really come to an assessment of 15 % as a PML on that account and agreed that Mr Andrews was wrong to consider that that account was not spiral business. Mr Andrews said that the basis upon which he classified whole account business under class Ol rather than class 10 was his personal knowledge of the type of business that he was reinsuring. In a prolonged piece of cross-examination Mr Gaisman demonstrated a number of examples of whole accounts that had been classified as Ol notwithstanding that they included a substantial element of XL on XL business. Mr Andrews conceded that his categorisation of class Ol business had been proved wrong but suggested that he had been misled by the information provided by the placing brokers. I did not find this a satisfactory explanation for Mr Andrews' misappreciation of the exposure to which he was subjecting his Names.

In his witness statement Mr Andrews referred to a number of factors which led him to believe that it was only necessary for him to buy cover in respect of about 15 % of his whole account aggregates. The first was what he described as his cautious approach to categorisation, only placing a risk in class Ol if he thought that the chances were small that XL on XL losses would invade it. Added to this he thought that, as he wrote at a relatively high level which would not in any event be damaged except by losses of a significant size, his whole account writings ought to be largely safe from invasion. He said he also bore in mind the loss of business in the direct marine market which made the exposure to whole accounts that much less. Finally he thought that by covering 100 % of his aggregate exposure for class 10 risks he was being over cautious which gave an element of slack in his reinsurance programme. For the reasons that I have already given, Mr Andrews had no justification for thinking that his approach to categorisation was cautious. His reference to the security implicit in writing risks that are relatively high level echoes comments made by Mr Andrews on a number of occasions when giving evidence about writing at a high level risks which were safe. Mr Andrews' attitude to the degree of exposure involved in writing the higher layers was demonstrated by a passage in his evidence when he said he would not buy reinsurance at those levels unless he could get it at 2 % or 1 % on line. He did not think that there was any need to buy cover at that level because there was no exposure. Passages in Mr Andrews' evidence suggested that he felt that much of the business that he was being paid to write did not involve any significant risk. His statement that the loss of direct marine business had resulted in a reduction of whole account exposure was a point that I failed to understand. It suggested that he was being requested to provide whole account reinsurance cover for business that did not exist.

Whether Mr Andrews assessed the exposure that he needed reinsurance to cover by calculating PMLs or whether he adopted a less scientific approach I consider that he demonstrated incompetence in estimating that exposure.

Reinsurance

The consequence of failing properly to calculate his Names' exposure may have resulted in Mr Andrews failing to obtain the reinsurance that he intended should cover that exposure. I say may because I am not confident that Mr Andrews based the amount of reinsurance that he bought on any calculation of his Names' exposure. In this Action he has said that he set out to cover the PMLs he assessed on class Ol and class 10 of his business. He also stated that he aimed to purchase reinsurance to approximately 40 % of the aggregate of those two classes which produced a similar figure. It was put to Mr Andrews that he had, in the past, described his reinsurance cover as being two and a half times his income, or twenty times his previous largest loss. Mr Andrews accepted that he might have made these statements by way of reassurance to Members Agents, but was adamant that they did not represent reinsurance policy. In the end I was left in doubt as to the basis upon which Mr Andrews purchased reinsurance.

The Exposure

The parties are not agreed as to the precise exposure to which Mr Andrews subjected his syndicate. What has been agreed in each case is the reinsurance cover available to meet each of the Five Central Catastrophes, together with the cover that would have been available had it not been depleted by prior claims. When this is compared to GWRO's estimated ultimate gross losses, a fair impression can be gained of the syndicate's exposure. While the estimated ultimate gross losses are not agreed figures, they do not in any event represent the maximum potential exposure but only the actual exposure to the particular catastrophe:

Casualty Actual Cover Cover on a Estimated Ultimate

First loss Basis Gross loss

Piper Alpha $109,750,000 $109,963,000 $220,000,000

Exxon Valdez $127,450,000 $149,950,000 $200,000,000

Hurricane Hugo $161,225,000 $169,950,000 $380,000,000

Phillips Petroleum $84,050,000 $184,925,000 $195,000,000

90A $131,075,000 $134,694,000 $132,000,000

Failure to Match Reinstatements

In his final submissions, Mr Vos contended that the disparity between cover on a first loss basis and actual cover available in relation to Phillips Petroleum was the consequence of a failure to match reinstatements. Mr Jewell, when giving evidence, had explained the disparity in general terms by saying that by October, when this catastrophe occurred, many of the preceding events had eaten into the available reinsurance cover. To a degree these two explanations may be different sides of the same coin. As at 1 July 1989 Mr Andrews had in place layers of whole account reinsurance which appear to have been designed to cater for three catastrophes, inasmuch as they provide for two reinstatements of cover up to a high level. However, the first and second reinstatements do not fully reinstate the available cover in that there are significant gaps in the upper layers of cover and the second reinstatement provides cover up to only $120 million, compared to the $150 million limit of the original cover.

The scale of the claims on Syndicate 298 in relation to Phillips Petroleum, when contrasted with the limited cover available to meet those claims, suggests that the exposure is due at least in part to a failure on the part of Mr Andrews to match reinstatements. I am satisfied that Mr Andrews, in order to cover obvious exposure, should have had in place reinstatement cover extending to a sufficient height to protect against the Phillips Petroleum claims. Mr Jewell's evidence suggests, however, that part of the exposure to Phillips Petroleum is due, not to lack of vertical cover, but to erosion of cover at lower levels as a consequence of a sequence of loss events which outstripped the horizontal cover available.

Piper Alpha

The deficiencies of Syndicate 298's reinsurance cover were demonstrated when Piper Alpha was lost. Mr Andrews said that he had regarded the total loss of a North Sea oil platform as unthinkable. Mr Eder submitted that such a view could not be condemned as incompetent. I do not believe that a competent excess of loss underwriter could properly discount the risk of the total loss of a North Sea oil platform, while accepting premium to provide excess of loss cover against that risk. This issue is, however, of no importance. Mr Andrews accepted that there were other larger casualties the possibility of which could reasonably be foreseen - Hurricane Hugo is an example. Mr Andrews could not justify his reinsurance policy on the basis that he had bought cover against the most serious catastrophe that he could reasonably envisage.

In summary, subject to his deliberate low level retention, it was Mr Andrews' policy to cover his Names exposure against any event that was a foreseeable possibility by reinsurance. He signally failed to do so. However he set about assessing the vertical exposure that he wished to cover he failed to exercise the skill and care to be expected of the reasonably competent underwriter when making that assessment. His failure adequately to assess and buy cover for his Names' against such exposure was negligent.

Rating

Assessment of exposure is a crucial element when considering the adequacy of premium rate for an excess of loss risk. No competent underwriter could decide to leave the upper layers of his PML exposed without forming a view of the likelihood of an event which would impact those layers. Mr Andrews did not believe that he was leaving unprotected upper layers of his aggregate which were liable to be impacted from time to time by catastrophes. He thus never applied his mind to the rate that would be appropriate to reflect that risk. His attitude was that the higher layers of cover that he was writing under his whole account classification were virtually risk free. Thus he spoke of writing high layers of cover for specific clients where he felt comfortable writing them as a safe risk. He told me that the basic concept that he applied when considering rating was to relate the premium charge to the loss record of the client at the particular level of cover. The experts all agreed that this was an inappropriate basis for rating where the historical period under consideration was free of major catastrophes. When asked why those he was insuring were prepared to buy cover when there was no significant risk he replied that he felt that the exposure to loss was minimal but that "if they were prepared to pay why was I to say no".

As I understood Mr Andrews' evidence he led some aviation business but otherwise was not a significant leader of excess of loss business. He described how each layer was rated lower as one went up through the layers until one reached 1 % or 2 %. When asked whether he considered that this structure was justified in the case of spiral business he would not express a personal view but simply stated that all that he could say was that that was the way the market operated.

Mr Andrews' evidence satisfied me that he had never had a proper appreciation of the excess of loss business that he was writing. He built up that business at a time that was largely free of major catastrophes and, by basing his approach to underwriting on the past historical records of those that he was covering wrongly discounted the possibility of major catastrophes that should have formed a vital element in his assessment of the risk. This was compounded by a failure fully to appreciate the gearing effect of spiral business and the extent to which that business had invaded the whole account covers that he was wrongly evaluating as posing minimal risk. Mr Andrews' performance fell far short of reasonable competence. These comments apply equally to the 1988 and the 1989 years.

Information Provided to Names

The Annual Reports and Accounts made it plain that Syndicate 298 specialised in writing excess of loss business and that such business included a significant element of excess of loss on excess of loss. Mr Andrews did not believe that he was exposing his Names to the risk of a loss beyond the reinsurance protection that he had in place. He said that before Piper Alpha he considered that a loss of 25 % or 50 % on stamp would be a disaster. In these circumstances it is not surprising that he did not warn his Names that they were exposed to such a risk. The average profits of the syndicate over the years 1983 to 1987 were 6.5 %. Mr Andrews described this as not a bad return, but disavowed any suggestion that his syndicate was a high reward syndicate.

SYNDICATE 299

Mr Willard

Mr Willard was born in 1935. He first went to work at Lloyd's in 1951. He served for 10 years as an underwriting clerk for Syndicate 284 and joined Gooda Walker in 1968. In 1969 he was appointed deputy underwriter to Mr Fullick, a position that he held until 1983, when he replaced him as active underwriter of the syndicate. He is thus a marine underwriter of immense experience.

Mr Willard himself shared in full measure the losses made by Syndicate 299. It was plain that he keenly felt his responsibility for the losses that had been suffered by his Names. Mr Eder, in his final speech, described him as a broken man. He was certainly an emotional witness and at times showed deep distress. Some areas of his evidence I have been unable to accept but I think that they may well result from Mr Willard deceiving himself rather than attempting to deceive me.

On taking over from Mr Fullick, Mr Willard continued the same underwriting policy. This was to write a broad spread of marine business together with an excess of loss account. Between 1983 and 1988 stamp capacity more than doubled from £18.3 million to £41.6 million. Gross premium income did not keep pace. It increased from £13.9 million to £19.5 million. This reflected over capacity in the marine market. The excess of loss element in the book broadly kept pace with stamp capacity, until the years 1988 and 1989 when it virtually doubled as a result of reinstatements. Mr Willard was a cautious underwriter and he told me that he resisted pressure from Mr Walker to increase the proportion of excess of loss business that he was writing. Mr Hawkes, Mr Willard's deputy, took over writing the excess of loss business in 1987 under Mr Willard's supervision. He continued to follow the established policy. This was to write middle to upper layers of business, which included XL on XL and whole account including XL. Mr Willard remained responsible for the reinsurance programme.

Policy in Relation to Exposure

The 1988 year of account has closed with a total loss of £22,148,302, representing 53.28 % of stamp capacity and over 100 % of gross premium income. A major cause of this loss was exposure to Piper Alpha. At the time of closure this was estimated at $24 million. It had been no part of Mr Willard's policy to expose his Names to the risk of such a loss. He had believed that he was operating a cautious policy under which he had in place sufficient reinsurance to protect his Names against any foreseeable catastrophe. He did not suggest that the loss of a rig such as Piper Alpha was not foreseeable. His evidence was that he had not foreseen that the gearing effect of the spiral would take such a loss through his reinsurance protection. So far as Mr Willard is concerned, the essential issue is whether his failure to provide his Names with adequate reinsurance protection was attributable to a failure to exercise the skill and care to be expected of a reasonably competent excess of loss underwriter.

Aggregates

No criticism can be made of the recording or monitoring of aggregate exposure by the syndicate. An aggregate book was kept in which entries were made both alphabetically and by class of business. No historical aggregate record was kept, but at any moment in time it was possible to calculate the current aggregate exposure of the syndicate. Mr Hawkes said that he reviewed aggregates weekly.

PMLs and Re insurance Policy

It is the Plaintiffs' case that prior to the loss of Piper Alpha Mr Willard never attempted to assess his PMLs. It is the Defendants' case that Mr Willard assessed his PMLs and had regard to them when deciding how much reinsurance cover to obtain. Before turning to the evidence I proposed to set out the relevant extracts from the pleadings.

Points of Claim

"27(1) The active underwriter, Mr Willard, and/or his assistant, Mr Hawkes, appreciated the need to calculate and monitor total aggregate exposures on the various elements of LMX business underwritten, and kept a book in which they sought to do so.

(2) However, they did not appreciate the need to plan, assess or monitor PMLs on the various elements of such business and overall using, inter alia, the book of total aggregate exposures which they kept; and did not do so at all or in any appropriate way.

(5) If (as to which no admissions are made) Mr Willard and Mr Hawkes made any assessment of PMLs... they took a deliberate decision to run an unprotected risk in the region of $30 million in respect of the Syndicate's then PML.

Points of Defence

41. The First and Second Defendants aver:

a. that Messrs Willard and Hawkes monitored carefully the risks written by them using the book of aggregate exposures maintained by them;

b. In assessing their possible maximum loss they always erred on the side of caution. They applied a figure of 100 % to XL on XL exposure. Whole account exposure, on the basis that it included both cargo and XL exposure, was rated at approximately 75 % and the time account at approximately 70 %;

c. Having assessed the possible maximum loss, Mr Willard would calculate the level of reinsurance that would be obtained by reassuring up to 50 % of the syndicate's aggregate exposure. He would then compare the latter with the possible maximum loss and consider whether, taking into account historical experience and past claims experience, any loss was likely to go higher than the 50 % on aggregate figure. This consideration governed whether he considered it prudent to run the gap between the 50 % on aggregate figure and the syndicate's possible maximum exposure to a loss. Until 1988 no loss had gone higher than the second layer of whole account protection. Finally Mr Willard would cross check that the proposed vertical protection was at least 400 % of the premium income written on the XL account for the syndicate;

42. Paragraph 27(5) of the Re-Re-Amended Points of Claim is admitted. The First and Second Defendants aver that the decision was deliberate and reasonable and it is denied, if it is intended to be so alleged, that it was not one which a prudent and competent underwriter could reach.

47. Further, the First and Second Defendants aver that the decision as to the level of reinsurance to be purchased by Mr Willard was made on the basis of the consideration of a number of factors, including the following:

a. probable maximum loss;

b. historical experience;

c. the syndicate's past year's claims experience;

e. the cost of purchase of reinsurance;

f. the level of aggregate exposure;

g. the level of written premium income;

h. the quality of security available;

Further, the level of reinsurance was kept constantly under review."

In his witness statement Mr Willard said this about monitoring PMLs:

"In assessing the PML I erred on the side of caution. I applied a figure of 100 % to the xl on xl exposure. I rated whole account exposure at approximately 75 %, on the basis that it included both general and xl exposure and I rated other sections of the time account at approximately 70 %. 1 applied these factors and percentages in my calculations for the PMLs of Piper Alpha, 87J and Alicia on the LUAA form. The result was a "gap" between the PML figure and the 50 % of total aggregate and the underwriting decision which I faced was whether, given historical claims experience and the other factors referred to, it was prudent to run that "gap" in the event of any one loss. In 1988, the "gap" appeared to be in the region of $30m for a dollar loss. Prior to 1988, no loss had gone higher than a second layer of the whole account protection.

The "gap" was, until mid 1988, in itself part of a consistent policy. This can be seen by reference to the following table.

US$ June 1983 July 1988 Growth

Capacity 36.6 83.2 227.32%

Aggregate 41.7 100.2 240.29%

Reinsurance 22.5 55.8 248.00%

Gap 19.2 44.4 231.25%

Gap as a percentage

of stamp 52.45 % 53.37%"

In his expert Report, Mr Von Eicken drew attention to what appeared to be an inconsistency between this evidence and the evidence given to the Poynter Committee, to which I shall shortly refer. This led Mr Willard to "clarify" his policy in a supplementary statement:

"I would always attempt to have vertical reinsurance for any one loss of up to roughly 50 % of the total aggregate exposure across the whole of the syndicate's book... I then calculated what I saw as the syndicate's exposure to those classes which could be exposed and therefore accumulate on any one catastrophe and apply PML factors of 100 % for XL/XL, approximately 75 % for whole account and approximately 70 % for time account. This would result in a "gap" between the vertical reinsurance bought and the total aggregates on exposed classes. By the time of Piper Alpha the "gap" in question was approximately $30 million, being the difference between what I saw as the syndicate's then maximum exposure ($86m.) and the then vertical reinsurance available for any one loss (approximately $56m.)"

When he came to give evidence Mr Willard explained that he did not use the expression PML but WLOB (worst likely outcome basis). The figure of $86 million represented the WLOB. The overall aggregate exposure was approximately $30 million more than this.

There was an obvious conflict between Mr Willard's evidence that he believed that his reinsurance covered his Names against the risk of any foreseeable catastrophe and his evidence that there was a gap of $30 million between his reinsurance cover and his calculation of the WLOB. He was cross-examined about this and was unable to explain it. His statement that no past claim had gone above his second layer of reinsurance cover was not a satisfactory explanation. Mr Vos, however, made a more fundamental attack upon his evidence. Mr Vos suggested that he had never carried out PML, or WLOB, calculations, nor had he followed a policy of reinsuring 50 % of his aggregates. In the course of cross-examination Mr Vos put to him a remarkable history of confusion and conflict in relation to the evidence that he gave to the Poynter Committee. He did not initially tell the Committee that it was his policy to insure 50 % of his aggregates. On the contrary he stated "I always basically kept roughly a $30 million gap". He was asked:

"Did you take parts of excess of loss accounts and allocate different levels of PML to them? For example, would you say that the whole account might be 90 per cent to 100 per cent PML whereas cargo or hulls or rigs would be a lesser percentage?

To this he answered "No".

The Poynter Committee showed great, and understandable, confusion as to the nature of and the basis for Mr Willard's $30 million gap. They tried to sort this out by correspondence, in relation to which Mr Willard was assisted by his Solicitors, Reynolds Porter. Mr Willard provided a series of statements, not all mutually compatible, one of which suggested that he switched from basing reinsurance on aggregates to basing this on a PML at some time in 1988. In the witness box Mr Willard accepted that he had given the Poynter Committee a series of answers which were "complete rubbish", "complete nonsense" and "completely false". Mr Willard said that he was tired and sick and wanted to get the Loss Review Committee out of his hair.

If Mr Willard did indeed have a reinsurance policy based in part on PMLs, I can see no reason why he should not have been willing and able to explain this to the Poynter Committee.

Mr Hawkes, when discussing reinsurance policy, said to the Poynter Committee:

"We did not have preconceived definite formulas or theories. It was an on-going evolving reinsurance programme."

In evidence he said that he and Mr Willard were of the opinion that if they protected to roughly half the aggregates they would not cause a loss to an individual Name of roughly more than 50 % of his stamp capacity, and that this was a policy that they inherited from Mr Fullick.

I am quite satisfied that Mr Willard did not calculate PMLs and thus that they formed no part of his reinsurance policy. I believe, as was suggested to Mr Willard, that the $30 million gap was something that only became apparent to him in the latter part of 1988 after the Piper Alpha loss as representing what appeared to be the difference between his syndicate's exposure to that loss and its reinsurance protection. It is a fact that between 1983 and 1988 the syndicate's excess of loss protection remained at an approximate, though not precise, level of 50 % of the excess of loss aggregate exposure. Mr Willard told me that when he took over the syndicate he increased the reinsurance protection from 45 % to 50 % of the aggregate. I asked him why he increased it by 5 % rather than leaving it in place or increasing it by 10 %. He answered "It was the feeling I have to do something. I am the new underwriter, I would review it." I believe that Mr Willard probably continued to maintain reinsurance at that level on the simple basis that it had proved sufficient in the past. This was an unsound basis for what was intended to be a cautious policy of covering the Names' exposure to catastrophe by reinsurance. It demonstrated, in my view, a failure to appreciate the gearing effect of the spiral. Both Mr Willard and Mr Hawkes were asked about the spiral and neither satisfied me that they had properly appreciated its implications. In my judgment the exposure to which the Names were shown to be subject when Piper Alpha was lost resulted from lack of competence on the part of Mr Willard and constituted a breach of duty on the part of the Defendants.

Rating

Mr Willard stated that:

"The premium levels were based on the leading underwriter's assessment of the actual risk being covered. The premium charge for any layer could not be based on the premiums for the original risk, nor on any assessment of the original risk being covered. To have done this would have been completely impossible, impracticable and irrelevant. The effective risk that an underwriter is assuming in writing an excess of loss account is that of a major catastrophe occurring and no amount of analysis of the individual underlying risks can throw any further light on the size of the loss which may happen."

It seems to me that this evidence reflected a complete abdication on the part of Mr Willard of any attempt at rating assessment. The vast majority of the risks that Mr Willard and Mr Hawkes wrote were renewals. They wrote high layers of cover because "obviously there was less likely exposure to risk than there would be by writing at lower levels". My belief is that Mr Willard would look at the past history of the business that he was writing and, on finding it loss free, accept almost whatever rate was going. He even wrote business at as little as 1 % on line, although he said that this tended to be in order to oblige brokers as he considered this rate too low. When asked why it was too low his answer was that it was half the rate that he had been writing in previous years. My impression is that it did not occur to Mr Willard, at least before Piper Alpha, that a rate of 2 % on line might not be adequate for the risks he was writing. Mr Willard's approach to rating lacked the skill and expertise to be expected of a competent underwriter of excess of loss business.

The 1989 Year of Account

The 1989 run off account as at 31 December 1993 indicated a loss of approximately £40 million against a stamp capacity of about £38,100,000 - that is a loss of about 105 % on stamp capacity. By an amendment to the Points of Claim the Plaintiffs allege that the whole of this loss was attributable to breach of duty on the part of Mr Willard and Mr Hawkes in a number of different respects, not all of them mutually compatible. These can be summarised as follows:

(i) They failed properly to asses or monitor the excess of loss PMLs or to cover these adequately by reinsurance;

(ii) They spent too much on excess of loss reinsurance, leaving other areas of the account with inadequate reinsurance;

(iii) They continued to write marine business notwithstanding that rates had fallen too low to be profitable.

The extreme case advanced by the Plaintiffs was that, at the end of 1988, Mr Willard should have decided to cease writing further business altogether and put his syndicate into run off. As a less extreme alternative, the Plaintiffs suggested that Mr Willard should have given up writing excess of loss business and the less attractive marine business and restricted his underwriting to business that was profitable.

To close down a syndicate and condemn the Names to the costs of run off is an extreme and expensive step. Periodic fluctuations of profitability are a recognised feature of the insurance market. Between 1987 and 1989 marine rates were depressed. The Plaintiffs have failed to satisfy me that Mr Willard was negligent in continuing to write a marine book of business in 1989. To have demonstrated this would have involved a detailed examination of marine underwriting at the time in order to ascertain the standards and practices applied by the reasonably competent marine underwriter. The Plaintiffs did not attempt that exercise. The evidence in this trial has focused on excess of loss underwriting. In so far as Mr Willard suffered losses on his marine underwriting in 1989 I suspect that his experience was shared by many others in the market. I do not believe that either he or they are automatically to be condemned as negligent for continuing to write business through this difficult period.

Excess of Loss

Mr Willard's policy in relation to the exposure flowing from his syndicate's excess of loss business did not alter between 1988 and 1989. His aim was to ensure that his Names were fully protected against exposure to foreseeable catastrophes by an appropriate level of reinsurance protection. I have already commented on the degree of expertise necessary to write a successful book that included spiral business. Mr Willard lacked that expertise. He did not fully understand the implications of the spiral. He was not able to evaluate premium rating for excess of loss risks or the dangers inherent in and the opportunities offered by the rating structure of the different layers. He did not appreciate the need to assess and monitor his PMLs. Piper Alpha brought home to him the fact that he had failed to protect his Names against their exposure to catastrophes. His reaction was to remedy this situation by reducing the cover written and increasing the reinsurance protection purchased. He told me that his intention was to reduce the $86 million exposure to $76 million and to increase insurance cover by $20 million, thereby bridging the $30 million gap. That gap was, as I have found, the difference between the 1988 reinsurance cover of $56 million and Mr Willard's calculation of the syndicate's exposure to Piper Alpha of $86 million. It is not clear on the evidence precisely how Mr Willard and Mr Hawkes set about implementing this policy. In January and February 1989 Mr Hawkes made a number of calculations of PMLs, which were put in evidence, which seemed to indicate two alternative approaches to this task, one his and one Mr Willard's. Mr Willard said that he made no PML calculations after Piper Alpha and that he had no recollection of Mr Hawkes providing him with details of PMLs. Not without some doubt I have concluded that Mr Willard was not working by reference to precise PML calculations when buying reinsurance, but working by reference to the $30 million gap.

A minute kept by one of the Members' Agents records Mr Willard as acknowledging after Piper Alpha that he had not been very wise to write excess of loss cover on high layers at a rate as low as 1 % and that he was now writing business on less volatile layers, yet on 1 January 1989, in a written statement of underwriting policy, he recorded that "the marine excess of loss account will mainly involve middle and upper layers". The syndicate continued to write layers at relatively high levels and low rates, but Mr Willard found that buying reinsurance to attempt to cover his exposure was expensive. This was in part, no doubt, because having given notice of anticipated claims in relation to Piper Alpha, the premiums that he was being required to pay to renew cover were significantly increased. In March 1989 Mr Willard is recorded by a Members' Agent as having predicted that his reinsurance costs would be 27 % to 30 % of income. If he did this was, as he accepted, wildly out. After reinstatements, his excess of loss reinsurance costs amounted to some 67 % of his excess of loss premium income, including reinstatement income.

Mr Willard accepted that, in the event, so large a proportion of his excess of loss income was spent on reinsurance that the business was doomed to be loss making. He attributed this to the difference between the rate he obtained on the business which he wrote and the rate that he had to pay for reinsurance. He denied that this was a problem which he should have appreciated, saying at one stage that when he bought his reinsurance he anticipated writing business at more satisfactory rates and at another stage that when he wrote his business he did not appreciate how expensive his reinsurance was to be.

The reality is, in my judgment, that Mr Willard did not give due consideration as to whether, and if so how, he could increase his reinsurance protection as he planned and yet write a profitable excess of loss book. His actions were dictated by what he considered to be the urgent need to cover the vertical exposure to catastrophes that Piper Alpha had revealed. The additional expenditure on reinsurance exposed his Names, as he accepted, to the certainty of loss, but limited the extent of possible loss in a manner which, having regard to the catastrophes that subsequently occurred, was beneficial to his Names. One cannot divorce Mr Willard's actions in relation to the 1989 year from those which had given rise to the problem that he was attempting to mitigate. The fact that Mr Willard's 1989 Names were placed in a position where they were bound to make a loss resulted from a lack of competence in the conduct of the underwriting business.

Exposure

The Underwriting Report as at 31 December 1993 gives the following picture for 1989 in relation to major catastrophe claims:

[Editor's Note: There is a Table at this point in the judgment which we are unable to incorporate onto Lexis. Please see original judgment for Table.]

This compares with the previous years's report which showed an estimated loss in excess of available reinsurance in respect of Hurricane Hugo of $3 million, an estimated loss in excess of available reinsurance of $9 million in respect of Phillips Petroleum, but no net loss in respect of Exxon Valdez. Neither of these losses resulted from claims exceeding the cover available on a first loss basis, which was $82,470,000. Mr Vos suggested to Mr Willard that the losses resulted from a failure to match reinstatements:

"Q: The fact there were several losses in 1989 should really have been irrelevant, should it not, if you had matched the reinstatements in your reinsurance programme to the reinstatements in the business you were writing?

A: Which I would have done, had I thought that was likely to be the situation.

Q: But it is the easiest thing in the world to say, "Well, I know I am writing business with four reinstatements which means I can get caught by four or five losses"?

A: I think you are exaggerating, if you are describing my accounts.

Q: What, why am I exaggerating?

A: Four or five reinstatements on upper layers, I do not think that is very likely.

Q: Let us say you are writing business with two reinstatements?

A: Two maximum, I would have thought, at high level, yes.

Q: Assuming you are writing business with two reinstatements, you should be careful, should you not, to ensure that your reinsurance programme matches the reinstatements you have written?

A: In hindsight, I should have been, yes, but I was not anticipating that number. I can say no more."

In my judgment Mr Willard was at fault for failing to match reinstatements. He should have had in place two full reinstatements of vertical cover to protect against catastrophes. It does not, however, follow that the losses were all attributable to a failure to match reinstatements - a matter to which I shall revert when I deal with damages.

Information Provided to Names

Mr Willard believed that he was following a cautious policy and that his Names' excess of loss exposure was fully protected by reinsurance. Mr Vos took him through his syndicate Reports and Accounts and put this proposition to him:

"For a Name and an Agent looking at these accounts, they would have taken the view that your business was properly protected by reinsurance and that it was a cautious syndicate going about a careful business."

Mr Willard agreed. I consider that he was right to do so.

SYNDICATE 290

Mr Walker

Mr Walker started working at Lloyd's in 1943 at the age of 15 and has had 35 years experience of underwriting in that market. He started Syndicate 290 in 1974 with less than 100 Names. By 1980 there were 489 Names on the Syndicate. By 1989 there were 3,163 Names on the syndicate with a stamp capacity of some £70 million. In the early and mid-1980's Mr Walker made good profits for his Syndicate, which no doubt accounted in part for its expansion. He was an admirably candid witness. One or two aspects of his evidence were equivocal, but in general he courteously answered the questions put to him in a lengthy cross-examination in a straightforward manner.

Policy in Relation to Exposure

Unlike Mr Andrews and Mr Willard, Mr Walker deliberately followed an underwriting policy which exposed his Names to the likelihood of periodic losses when catastrophes occurred. It is the Defendants case that this was a perfectly legitimate policy, justified by extraordinary profits in the catastrophe free years. The Plaintiffs contend that Mr Walker's policy, and the manner in which he executed it, were negligent in the following respects:

1) He did not control his aggregates.

2) He did not attempt to balance his book.

3) He did not attempt to assess or monitor his PMLs.

4) He had no planned limit of exposure.

5) He gambled that the catastrophe would not occur.

6) He made no attempt to rate correctly the business that he wrote.

Aggregates

This was one area where Mr Walker's evidence was equivocal. He suggested at one point that he restricted the amount of business that he wrote in order to limit the size of his aggregate exposure. The reality was - as I believe in the end he accepted -that it was his premium income limit that restricted the amount of business that he wrote. He wrote enthusiastically almost up to this limit, with the result that reinstatement premiums carried him well over the limit.

Balance

Mr Walker did not set out to restrict his exposure by balancing different categories of direct excess of loss cover. Almost all the business that he wrote was, or contained, XL on XL.

He coded risk with a view to assessing profitability, not to calculating aggregates or PMLs. Each of the first six of the codes that Mr Walker kept were capable of being exposed to the same single event. Mr Walker said that he was aware of this and agreed that the consequence was that, in 1989, the Syndicate was exposed to a figure approaching $570 million in the event of the worst possible situation.

PMLs

Mr Walker made no attempt to estimate a PML. He said that he did not think that this was a viable thing to attempt until a loss had actually occurred. It was not an appropriate exercise for someone who wrote excess of loss spiral business.

Exposure

When considering Mr Walker's attitude to exposure it is necessary to distinguish between two interrelated questions. The first is the extent of his Names' exposure to a single catastrophe. The second is the likelihood of such a catastrophe occurring.

The Extent of Exposure

Mr Walker said that he did not start by working out "cast in stone" what he wanted his net exposure to be, but gave a variety of explanations of how he arrived at this. He said that on average over the years he spent not more than 40 % of income on reinsurance, but this was not a policy - he was flexible. He said that he tried to keep reinsurance at about one-third of his aggregate. He said that he bought as much reinsurance as he could get at a reasonable price. In his statement he said that his reinsurance policy was essentially based on running a gap which could give rise to a loss of 200 % to 300 % over stamp if a truly major catastrophe occurred, such as a Florida windstorm or a West Coast earthquake somewhere along the San Andreas fault line.

He said that he tried to work out a ratio between reinsurance costs and exposing the syndicate to about 250 % maximum exposure. In his 1988 Accounts, which were published in May 1989, he stated:

"I am always alerting the Members to the high risk nature of the syndicate and I must again emphasise to all my members not to write too large a premium on the syndicate; I usually recommend about £25,000 maximum for a Member. The year may come when we have to face a loss of up to 200 per cent"

Mr Eder described this as the strongest warning given by Mr Walker, by which I believe he meant the strongest written warning. Mr Walker in fact stated that he had given oral warnings to Members' Agents that the maximum possible loss might amount to 300 % of stamp. There is no recorded instance of Mr Walker warning of the possibility of a loss of this size and I am not persuaded that he did. Mr Vos took Mr Walker through his codes and demonstrated that exposure to a single catastrophe might, on a worst case scenario, involve a loss in excess of 300 % of stamp. In the event, I believe that Mr Walker underestimated the extent of his Names' maximum exposure to a single event. This I do not believe to be of great relevance in this case, for the Names have not suffered from a single catastrophe of maximum severity. Their losses flow largely from two successive catastrophes, each of which very significantly exceeded their reinsurance protection. Hurricane Hugo is estimated to result in losses of $228,000,000 and 90A in losses totalling $265,000,000. This leads to consideration of the likelihood of such losses occurring.

The Likelihood of Losses

Mr Walker made it plain that the potential loss against which he warned his Names was not one which he envisaged would flow from a catastrophe such as Hurricane Hugo or 90A He had in mind a catastrophe of more dramatic proportions. In his final written submissions Mr Eder made the following point:

"... the Plaintiffs have made much of the fact that Derek Walker's warnings were predicated on the occurrence of "the big one" or "Armageddon"; a loss of the order of San Francisco falling into the sea and of the fact that neither Hugo nor 90A were this loss, even in his own estimation. The short answer to this point is that it is based on a misconception. The Plaintiffs appear to be assuming that Mr Walker's warning to Names was that they could make either a loss of 250-300 % or no loss at all. This is plainly rubbish. His warning was that if the worst happened they could lose 300 %; if what happened was only extremely bad they would still lose, but not as much. This is exactly what happened in the event.

It is true that none of the catastrophes in question were as big as the catastrophe which Derek Walker had anticipated; but looking at each catastrophe individually the losses which resulted from them were nowhere near as large as he had predicted they could be."

This suggests that Mr Walker's warnings related only to the most severe catastrophe and that, when he gave those warnings, he was aware of a category of lesser, and less uncommon, catastrophes that would cause his Names severe losses, albeit not threatening the full extent of their exposure. I cannot accept this analysis. Mr Walker's warnings were directed at those casualties which might result in loss up to the limit of exposure. I believe that it was Mr Walker's view that only an Armageddon type catastrophe posed this threat.

In the course of lengthy and skilful cross-examination Mr Vos referred Mr Walker to the various factors which had, in fact, been responsible for his Names' exposure to Hurricane Hugo and to 90A, and obtained from him ready admissions that he was aware of them. He accepted that over the years a number of factors had increased the impact that natural catastrophes were likely to have on insurers and reinsurers. These included increases in population and the density of populated areas; increase in capital assets exposed to loss; an increasing propensity to insure property and inflation so that claims were larger in many terms. Mr Walker accepted that, having regard to his worldwide book of business, it was conceivable that he would have to face the consequences of two catastrophes of the severity of Hurricane Hugo in successive years. He accepted that it was possible to envisage more serious catastrophes than Hurricane Hugo. He was finally led to accept that he underwrote his account with the knowledge that any major catastrophe, including something like Hugo, would subject his Names to a very large loss indeed. Mr Vos relied upon this passage of cross-examination to support the submission that Mr Walker gambled on catastrophes such as Hurricane Hugo not occurring.

When Hurricane Hugo and 90A occurred, Mr Walker's reaction was not that these were likely to

cause very large losses to his Names. On the contrary, in each case he concluded that the claims would be contained within his reinsurance protection. I do not believe that Mr Walker had been through the exercise of attempting to calculate the type and level of catastrophe that might cause loss to his Names. He simply assumed that only the most extreme catastrophe would produce this result. There was no justification for making such an assumption.

RATING

Mr Fryer in his witness statement said:

"No alternative method [to historical experience] of assessing exposure is suggested, and I know of no measure which does not depend to some degree or other on past experience. Reinsurers approach to rating is not as simplistic as seems to have been assumed... The anticipated frequency and magnitude of catastrophe losses due to causes such as windstorm, earthquake and flood can only be estimated by an extrapolation of past trends adjusted for changes in the value of property exposed. Considerable efforts have been devoted by insurers and reinsurers with seismologists and meteorologists to improve risk assessment methods, but the answers of the scientists, in my experience at least, is that their advice is to look at the historic record over as long a period as possible and make intelligent estimates on the basis of those results subject to whatever adjustment may be indicated."

Mr Walker made no attempt at a statistical assessment of the chance of a relevant catastrophe. He said that it was impossible to forecast on this basis. I had waited with interest to hear what Mr Walker, as one of the leaders in the LMX market, would have to say on the subject of rating. The description that he gave in his witness statement was as follows:

"Syndicates and companies in the non-marine market were prepared to pay a high price for their excess of loss reinsurances and those prices rose steadily in the non-marine market all the way through the late 1980's. As one of the leading underwriters, I helped set those rates. On a relatively catastrophe free year, a well written excess of loss account could be profitable for Names.

Another important factor in non-marine excess of loss writing which assisted its profitability and viability was what was known as the "pay back". This was an arrangement whereby clients who had advised claims in the previous year of account would be charged for the next year's renewal, the differential to bring reinsuring underwriters into at least a break even situation or, we hoped, profit.

Rating Methods - Rate on Line

My rating method was, as I believe with the majority if not all xl underwriters, based on what was known as "rate on line" or ROL. I assessed a risk over a period. It would involve reviewing the client's previous claims history in respect of that particular account and that particular layer and I would rate accordingly. The ROL was the ratio the premium bore to the exposure. In my market, I was fully ware of all the major losses which had affected the market and what I really wanted to know was the effect which those losses had on the account in question.

The other factor of major importance was the identity of the client. The non-marine xl market was relatively small and tight, although there was a tendency in the late 1980's for new syndicates to enter it due to the high levels of return.

The trend of my underwriting in the late 1980's was to move away from the lower or "working" layers. This was because the lower or working layers were tending to be hit repeatedly each year and I regarded them as little more than money swapping. On certain lower layers I was finding that I just could not make money, even having taken a premium of 40 % on line. It therefore seemed sensible to move out of the lower layers. Historically, the higher layers had not been impacted by previous major catastrophes (with the exception of Alicia, a 1983 loss, which was beginning by 1988-89 to breach the top layers). My view was that by the mid to late 1980's writing lower layers almost guaranteed a loss."

In the witness box Mr Walker's evidence did not suggest that he played an active role in setting rates. Rather he gave the impression of having been a helpless captive of the market with little choice but to accept the rates on offer. Thus when it was put to him that it was not appropriate to rate a high layer at a low rate simply because the reinsured had a claim free record, he replied:

"I am afraid that is what the market did ... A lot of the top layers were coming down to quite low rates on line ... probably top layers were being placed at about 2 % in the non-marine market, because they had never been impacted ..."

When dealing with the rating structure whereby rates declined as the layers rose, he commented:

"At the end of the day, you could quite well say that in the spiral business, if the spiral goes to the top, then every risk should have the same rate... That is unfortunately not the way the market worked."

Summary

Mr Walker's policy was to underwrite so as to leave a substantial net exposure in the belief that only exceptional catastrophes would result in losses and that these would occur so rarely that the profits made by his Names in the good years would compensate for the occasional loss made in a bad year. Mr Walker had no basis for believing that the rates that he was charging for the higher layers that he was writing were sufficient to compensate for the occasional losses that he envisaged. Nor had he any idea of the size of the catastrophes that would precipitate such losses. He had not envisaged that catastrophes of the size of Hugo or 90A would result in a major loss to his Names. He believed that they were only vulnerable to significantly larger, and rarer, catastrophes than those. Mr Walker had, I believe, failed to appreciate the extent to which the spiral reduced the effective protection afforded by his own excess of loss cover.

High Risk / High Reward

The Defendants submitted that the profits that Mr Walker was making for his syndicate in the good years were sufficient to compensate for the losses that were liable to occur in the bad years.

In his final submissions Mr Eder placed before me calculations indicating an average rate of return for the ten years from 1979 to 1988 of 24.25 % after deduction of personal expenses. This was based on the syndicate accounts. Mr Vos had referred me to the Poynter Report, which relied for its figures on the Chatset League Tables. These showed the following rate of return together with the position of Syndicate 290 in the League Table:

1983 12 % 9 out of 101

1984 9 % 30 out of 102

1985 12 % 9 out of 101

1986 23 % 7 out of 102

1987 9 % 58 out of 116

1988 16 % 11 out of 115

The Poynter Report commented:

"It is apparent that Syndicate 290, except for 1987, consistently reported substantially better performance than the median return in the non-marine market."

The Report went on to consider the extent to which the syndicate's results were improved by time and distance policies. These are 'policies' which pay in a future year an 'indemnity' in an amount that exceeds the 'premium' by an amount which, as I understand it, reflects the value of the use of the premium over the period in question. Both the premium and the indemnity feature in the syndicate accounts in the year that the time and distance policy is taken out, so the result is to improve the profit/loss position shown in the accounts. Mr Walker accepted that the effect of his time and distance policies was that he was able to show a profit rather than a loss for 1981 and 1985 and that substantial increases were produced to the profit figures for 1983, 1986 and 1987.

The propriety of the use made by Mr Walker of time and distance policies has not been challenged in this Action. Mr Outhwaite told me that they are widely used at Lloyd's and that they have a legitimate role to play in a reinsurance programme. Thus the position remains that the results achieved by Mr Walker for Syndicate 290 were significantly better than those of most non- marine syndicates in the years up to 1989. It is, however, possible that this position was achieved in part by the skilful use of time and distance policies not matched by some of the other syndicates. Certainly, when the benefit of the time and distance policies is removed, it is not easy to recognise a pattern of a series of particularly profitable years being relied upon to balance the effect of the rare catastrophe.

Conclusion

Most of the criticisms made by the Plaintiffs in relation to the manner in which Mr Walker conducted his underwriting business are made out. He was deliberately running a net exposure to risk without monitoring the precise level of that exposure or correctly informing his Names of this. He had made no attempt to estimate how often his Names might have to face a year of loss and he mis-appreciated the level of catastrophe that risked bringing that result about. His rating was based not on any assessment of the earnings his syndicate needed to make in the good years to balance the losses in the bad, but on an acceptance of the structure and level of rates prevailing in the market. In these respects, the Plaintiffs' allegations of breach of duty are made out.

Information to Names

The Annual Accounts stated repeatedly that Syndicate 290 specialised in excess of loss non-marine reinsurance and that Mr Walker wrote and led catastrophe reinsurance. In addition the accounts included the following statements:

1980: "Major losses must be expected by this syndicate but such losses should be well contained within our protection".

1982: "We are always waiting for the large loss to hit the market and we should be in a strong position if this loss comes, with strong reserves behind the syndicate".

1983: "We had our largest loss since we started in 1983 ... this had produced large insurance claims as a result of this loss, it has enabled us to assess how a major catastrophe could effect our syndicate and because of this we have been able to get substantial increases on XL renewals for 1984. Our approach to reinsurance remains unchanged, we purchase substantial cover to protect the syndicate against a major catastrophe...."

1984: "We have in force a major reinsurance programme to protect us; this covers the whole account protection and is backed up by specific reinsurance for the different classes. The 1983 year had Alicia and the Winter Freeze... but we have more than adequate reinsurance protection and I believe we can still hope for a small profit.

The reinsurances protecting the syndicate are at all times under constant review and we believe our current policy covers the needs of the syndicate and is the best available.

We believe it is our duty to remind our Names of the nature of the class of business that we underwrite, and we must apologise if we have to constantly point out the higher risk nature of this syndicate. The only major loss since commencement of the syndicate is 1983 when Hurricane Alicia struck the Galveston area in the US Gulf and we are pleased to report that the gross loss to the syndicate is well within our reinsurance programme".

1986: "As we have advised all our Names in the past, this syndicate is a high risk by the nature of the business it writes and we have nearly reached the limit of reinsurance that is worth buying. We must always live with the problem which must apply to all syndicates that in the event of a major loss we could run out of reinsurance protection, but we have produced good results in the past 10 years and we can but await future losses which we hope will be contained within our reinsurance programme".

1987: "The 1987 account... will be effected by the major windstorm loss of October in the U.K. and Northern France. The final original loss must be in the region of £1.5 billion. We have been waiting for the large loss from the U.S. only to find it coming from the U.K. The syndicate purchased additional reinsurance in 1987 including specific non-dollar cover. It is early days to forecast the loss to the syndicate, but it should be contained within our reinsurances".

1988: "I am always alerting the members to the high risk nature of this syndicate and I must again emphasise to all my members not to write too large a premium on this syndicate; I usually recommend about £25,000 maximum for a member. The year may come when we may have to face a loss of up to 200 per cent. The 1987 U.K. loss is testing the system. Underwriters have to run a net account sometimes and this will happen to us in the event of a major earthquake, hurricane or flood. There is no way we can purchase enough reinsurance protection to cover our aggregate, but we do have an aggregate limit as opposed to a direct writer".

As Mr Walker accepted in cross-examination, with the exception of the 1988 accounts, these statements were not calculated to suggest to Names that they were being exposed to the risk of large losses in the event of a severe catastrophe. Lord Strathalmond commented, however, that as a Members' Agent these accounts would indicate to him that Mr Walker's syndicate was running a net exposure.

The 1990 Year of Account

Mr Walker's statement of underwriting policy, dated 17 November 1989, stated:

"The Syndicate will continue to specialise in Excess of Loss Non-Marine reinsurance. It will continue to write and lead catastrophe reinsurance, which covers fire, windstorm, flood and earthquake; this being reinsurance of Lloyd's Underwriters and world-wide insurance companies. It is my intention to substantially reduce the X/L on X/L business. This will come about in two ways.

The Non-Marine Underwriters, including ourselves, are being forced to retain, uninsured, approximately 20 % of their catastrophe income. This will remove the low level high rated business. There is also a co-insurance on all layers of between 5 % and 10 % and it will not be possible to place the high layers of X/L on A/L, so we will be keeping a very careful check on this class of business."

In his witness statement, he added:

"My 1990 writings therefore involved substantial reductions in aggregates. The reinsurance programme, however, did not change substantially until mid 1990 as the majority of high layer protections which we had on our generals ran through until 1.7.90."

In the event, the aggregates on the codes that included spiral business reduced by only $42 million from about $572 million to $530 million. This left Syndicate 290 exposed to 90A, on the Defendants' calculation, to the extent of $143 million, or 124 % of stamp.

Thus Mr Walker's policy and practice for 1990 was not significantly different from that for 1989. Just as for 1989 his conduct amounted to a breach of the duty owed by the Defendants to exercise reasonable skill and care in carrying on the business of underwriting.

SYNDICATE 164

It is common ground that, on liability, the Plaintiffs' case in relation to Syndicate 164 stands or falls with Syndicate 290. Mr Walker had assumed responsibility both for the excess of loss business written for Syndicate 164 under the split stamp arrangement and for buying reinsurance cover against the excess of loss exposure. Syndicate 164's accounts provided the information that excess of loss business was being written under a split stamp arrangement, but gave no warning of net exposure to loss. Mr Judd told the Poynter Committee that it was his understanding that the reinsurance arranged by Mr Walker would be sufficient to protect his syndicate against catastrophe loss.

GENERAL CONCLUSION

There are common features in the approach to the conduct of excess of loss business by Mr Andrews, Mr Willard and Mr Walker. Each of these underwriters had immense experience of the business of underwriting, having started work straight from school and spent all their working lives in the London insurance market. The approach of each of them to excess of loss underwriting was one that may well be appropriate in other fields of business - the reliance on past experience when estimating risk.

Past experience has to be treated with particular caution in the field of catastrophe excess of loss insurance, for the size and the incidence of catastrophes do not conform to a pattern. The growth of the LMX market in the 1980s and, in particular, the growth of spiral business, raised special problems in relation to the assessment of risk, exposure and rating, that called for special consideration. Some gave it that consideration. The Gooda Walker underwriters did not.

PERCEPTION OF RISK AND 'VOLENTI NON FIT INJURIA'

The Defendants raised a plea of 'volenti non fit injuria' to meet an anticipated allegation that the degree of exposure to which the Names were subject was negligent per se. To such an allegation it was the Defendants' case that all the Names knew or should have known that excess of loss insurance was high risk business, and that some of the individual Names were specifically aware of the extent to which they were exposed to the risk of loss. Consent to such exposure precluded the Names from seeking to base a cause of action upon it.

In the event Mr Vos did not contend and I have not found that the exposure to which the Names were subjected was negligent per se. The exposure in respect of which I have held the underwriters at fault was culpable because it was unintended, unplanned and unjustified by any proper analysis of risk. Even had Names been aware of the exposure, such knowledge would not have constituted acceptance of the circumstances that made it blameworthy or given rise to any defence. I propose, however, to make some general findings on the knowledge that Names had, or should have had, about the excess of loss business that was being written on their behalf.

Lord Strathalmond, who at the material time was a Director of Bain Dawes (Underwriting Agency) Ltd, told me that he appreciated that a specialist excess of loss syndicate specialising in catastrophe business would be of a higher risk or more volatile nature. There was a likelihood in years without natural man-made catastrophes that higher profits would be made, but if there were such a catastrophe larger losses could be made - claims would come in but they could possibly be reinsured out as well. A Members' Agent would have thought that two catastrophes impacting the same year of account might result in losses of 30 % or more of the syndicate's stamp capacity - losses of as much as 100 % would be almost unthinkable. He said that before the Piper Alpha disaster he had not heard the term LMX and that he did not know what excess of loss on excess of loss policies were - he was not aware of the spiral. He thought that his knowledge was fairly typical of Members' Agents.

Mr Jewell commented as follows:

"Excess loss syndicates were perceived as being high risk by most people within the market and that meant, in most people's perception, that a severe catastrophe or series of catastrophes might cause a syndicate of that nature to have losses that would represent something between 25 % and 50 % of stamp capacity. In the most extreme of examples, it would be considered almost unthinkable that it would get to a level of 100 %."

In my judgment these comments fairly reflected the perception that Names, competently advised by their Members' Agents, would have had of Syndicates 164, 298 and 299. Mr Andrews said that before Piper Alpha he would have been very sorry to lose 25 % or 50 % of the stamp. Mr Willard would not have considered that he was exposing his Names to even this degree of risk.

Mr Walker's warnings, to which I have referred earlier, were enough to draw to the attention of Members' Agents and his Names that his was a high risk syndicate and that a catastrophe could occur which would lead to claims that significantly exceeded the level of reinsurance cover. Mr Jewell was often present when Mr Walker described the nature of his syndicate. He gave this description to the Poynter Committee:

"He always told the syndicate and told the world "This is a high risk, high return syndicate". The sort of words he would say are "There are ten years of profit but when the big one happens we are going to have a loss"... he could not have emphasised more strongly the high risk nature of the syndicate. I have never heard anybody express it stronger than he did in front of the Names and/or agents."

Mr Vos asked Mr Jewell if he heard Mr Walker explain what he meant by "the big one". He said that he did - on many occasions:

"The sort of events that would be described would be California slipping into the ocean, a North Sea tidal wave washing out some rigs and flooding London or Europe, the Tokyo earthquake, or a jumbo jet on a skyscraper in New York. These are the sort of theoretical worst situations that would be described as producing a horrible loss to the syndicate."

Thus Mr Walker's warnings reflected his own view that only an extreme catastrophe posed a threat of heavy losses to his syndicate. As to the amount of those losses, Mr Jewell said that he did not recall the extent of the loss being discussed at all.

CAUSATION AND DAMAGES

The following order was made on the Summons for Directions:

"The issues to be determined at the trial be limited to issues of liability and questions of principle relating to quantum."

The primary question of principle that arises in relation to quantum is whether the Plaintiffs have made good their claim that all their losses in the relevant years have been caused by the Defendants' breaches of duty. If this claim is not made out, issues of principle arise as to the extent to which, if at all, the Plaintiffs have established causal nexus between breaches of duty and losses sustained, and the manner in which the relevant losses should be assessed. These questions overlap. Before they are reached, however, I have to address an argument advanced by the Defendants as to the scope of the claim which it is open to the Plaintiffs to advance.

The Scope of the Claim

The Defendants contend that the only charges that they are required to meet relate to the Five Central Catastrophes : Piper Alpha, Exxon Valdez, Hurricane Hugo, Phillips Petroleum and Windstorm Daria 90A At the most the damages for which they are liable are the losses resulting from those catastrophes.

To resolve this issue it is necessary first to consider the pleadings. The Points of Claim allege, almost exclusively, negligence in relation to the underwriting of excess of loss business. The exception is an allegation that Mr Willard and Mr Hawkes "failed to understand or to act upon the common knowledge that the marine market was suffering from over-capacity and inadequate rating, so that following underwriters with limited capacity, who had little to offer brokers, could not expect to write profitable marine business". This allegation was relevant to the submission that Mr Willard should not have continued underwriting in 1989. I have rejected that submission.

It follows that the Plaintiffs are restricted to recovering damages flowing from incompetence in relation to the writing of excess of loss business.

Do the pleadings restrict the Plaintiffs to claiming the losses flowing from the Five Central Catastrophes? In my judgment they do not. The Points of Claim first allege, in detail, the conduct that amounted to negligent underwriting. The allegations include:

Writing too much spiral business;

Failing to calculate and monitor aggregate exposures;

Failing to calculate and monitor PMLs;

Writing business at rates that would not finance appropriate reinsurance;

Writing high level business at inadequate premiums.

The pleading then alleges that during 1987 to 1990 a succession of major losses occurred, the worst of which are identified in an Annex to the pleading. This is a Schedule of ten of the worst catastrophes in the period in question. Finally the pleading alleges that all of the losses suffered by the Plaintiffs in relation to the relevant years, including run off costs, have been caused by the pleaded breaches.

Nowhere do the Points of Claim restrict the Plaintiffs' claim to the losses flowing from the Five Central Catastrophes.

Despite the breadth of the Plaintiffs' pleaded case, the manner in which they focused on the Five Central Catastrophes in the preparations for this trial, which included detailed discussions with the Defendants' representatives and a considerable measure of agreement in respect of the syndicates' exposure to the Five Central Catastrophes, and the manner in which Mr Vos referred to these catastrophes in his opening, could well have led the Defendants to believe that the consequences of these catastrophes formed the essence of the Plaintiffs' complaint.

On the seventh day of the trial Mr Vos interrupted Mr Eder's cross-examination of Mr Jewell in order to make it plain that the Plaintiffs' claim was not restricted to the losses flowing from the Five Central Catastrophes, but embraced all the losses that they have or will have sustained as members of the Gooda Walker syndicates in the relevant years. Mr Eder registered a protest at this formulation of the Plaintiffs' claim - a protest that he repeated on a number of occasions during the trial.

The Plaintiffs' case, as clarified by Mr Vos in the course of the trial, is a simple one. The Plaintiffs contend that the entire underwriting of each syndicate was conducted in disregard of basic reinsurance principles. They say that it is neither appropriate nor possible to attempt to identify individual aspects of the underwriting of each syndicate that were negligent and to attempt to assess damages by reconstructing the results that each syndicate would have enjoyed had those aspects been competently performed. Unless the Defendants can prove to the contrary, the Court should proceed on the basis that all the losses sustained by each syndicate were a consequence of the negligence.

This case amounts, in large measure, to a plea of 'res ipsa loquitur'. The Plaintiffs rely on the fact that there has been incompetent underwriting on the one hand and losses sustained on the other as giving rise to the inference that all the losses are attributable to the incompetence. I propose to deal with that case on its merits. It is a case which is open to the Plaintiffs on the pleadings and it is not a case which has, in my judgment, resulted in embarrassment or prejudice to the Defendants.

The Defendants have raised the following contentions in answer to the Plaintiffs' case:

1) Not all of the losses are attributable to excess of loss underwriting in the relevant years.

2) During the relevant period there was an unprecedented and unforeseeable 'concatenation of catastrophes'. A competent excess of loss underwriter could not reasonably have been expected to protect his Names against all loss in the face of these catastrophes. This series of catastrophes was, at least in part, an independent cause of the losses sustained by the Names.

3) The evidence demonstrates that significant heads of loss have been sustained by the Names which do not arise from the breaches of duty that the Plaintiffs have alleged in this Action.

In these circumstances the Defendants contend that it is both possible and necessary to distinguish between the losses flowing from such specific breaches of duty as the Plaintiffs may have made out and losses attributable to other causes.

The Causes of Loss

Excess of loss business did not constitute "the entire underwriting" of Syndicates 164 and 299. In so far as the losses suffered by those syndicates are attributable to the balance of their business, they cannot be laid at the Defendants' door.

The Plaintiffs' allegations in this Action have been of negligence in writing excess of loss business in respect of the relevant years. It follows that the Plaintiffs can have no right to recover damages in respect of losses which do not flow from the writing of excess of loss business in respect of those years. Thus, for instance, losses flowing from the deterioration of business written in prior years, transferred by reinsurance to close, cannot be recovered. The evidence indicates that some of the losses sustained by the Plaintiffs fall into this category.

For these reasons alone, the Plaintiffs' claim that they are entitled to recover all the losses that they have sustained cannot succeed.

The 'Concatenation of Catastrophes'

Statistics

Mr Harold Clarke, who heads the general insurance practice of a firm of actuaries called Bacon & Woodrow, gave evidence of what a reasonable actuarial forecast of insured catastrophe losses would have been in the 1980's, based on an analysis of the catastrophe losses in earlier years. His analysis was based on a statistical model, which produced figures for what was to be "expected" in the years in question. Mr Clarke commented "the model uses standard distributions as this is most likely what would have been used in practice by reinsurers". I do not believe that any of the Gooda Walker underwriters spoke of carrying out statistical exercises of the type conducted by Mr Clarke and Mr Walker expressly stated that he did not do so.

Mr Clarke introduced his model with the following explanation:

"Catastrophe claims are intrinsically random in nature in that one cannot say when a catastrophe will occur or how severe it will be. However, over a longer period an underlying pattern (or distribution) may be observed. I have constructed a model of catastrophe claims which estimates the pattern in catastrophe claims.

Modelling catastrophes allows

i) the projection of expected future losses

ii) the establishment of confidence bands for those projected losses

iii)the estimation of the chance that the number of disasters for a particular period will exceed or be less than a specified count, and

iv) the projection of the future occurrence rates of disasters producing losses within specified ranges."

It is important to appreciate that what a model such as Mr Clarke's indicates is "expected" to happen in any given year is unlikely to happen in practice. The expectation is simply a statistical possibility based on past random events.

Mr Clarke produced tables for the expected amount of losses flowing from claims in excess of $100 million, $250 million and $500 million world wide and in Europe and North America. As he found that catastrophes occurring in North America and Europe account for 87 % of all catastrophe insured loss, I shall simply set out one set of results - the world wide figures:

Claims in Excess of $100m

Year Expected amount Actual amount Expected less

$ m $ m actual $ m

1987 2,830 3,264 (434)

1988 3,617 3,322 295

1989 3,870 9,323 (5,453)

1990 5,890 17,509 (11,619)

Claims in Excess of $250m

Year Expected amount Actual amount Expected less

$ m $ m actual $ m

1987 1,182 1,208 (26)

1988 1,261 2,470 (1,209)

1989 1,667 8,308 (6,641)

1990 4,007 14,990 (10,983)

Claims in Excess of $50Om

Year Expected amount Actual amount Expected less

$ m $ m actual $ m

1987 741 870 (129)

1988 682 2,140 (1,458)

1989 1,375 6,700 (5,325)

1990 3,968 13,016 (9,048)

Mr Clarke also carried out, in the course of the trial, calculations to ascertain the statistical probability of the actual loss levels experienced, resulting from claims above the same selected thresholds. These were his results, worldwide:

Cutoff: US$ l00m

Year Expected Actual Likelihood that

amount amount (A) loss exceeds A

1987 2,830 3,264 1 in 5 years

1988 3,617 3,322 1 in 3 years

1989 3,871 9,323 1 in 31 years

1990 5,890 17,509 1 in 42 years

Cutoff: US$ 250m

Year Expected Actual Likelihood that

amount amount (A) loss exceeds A

1987 1,182 1,208 1 in 3 years

1988 1,261 2,470 1 in 10 years

1989 1,667 8,308 1 in 105 years

1990 4,007 14,990 1 in 47 years

Cutoff: US$ 500m

Year Expected Actual Likelihood that

amount amount (A) loss exceeds A

1987 741 870 1 in 3 years

1988 682 2,140 1 in 20 years

1989 1,375 6,700 1 in 125 years

1990 3,968 13,016 1 in 47 years

Mr Clarke concluded from these results that:

"the total amount of catastrophic loss in 1989 and 1990 was larger than could have been expected, given the experience up to that time".

His figures amply justify that conclusion, whether the expectation is statistical or the expectation of the reasonably competent underwriter. This is perhaps more readily appreciated simply by considering the Sigma catastrophe statistics, which provided Mr Clarke with much of his data. I have extracted from his tables the following figures, using values that have been adjusted to accord with the value of the dollar in 1986 in order to enable more meaningful comparison:

Casualties exceeding $100m $250m $500m $1,000m

1969 2 1 1 0

1970 5 4 3 0

1971 2 0 0 0

1972 6 3 0 0

1973 2 1 1 0

1974 5 3 2 0

1975 6 1 0 0

1976 3 2 1 0

1977 5 2 0 0

1978 7 2 2 1

1979 9 4 3 1

1980 5 4 0 0

1981 2 0 0 0

1982 9 2 0 0

1983 3 2 2 0

1984 7 3 0 0

1985 13 4 1 0

1986 5 0 0 0

1987 13 2 1 0

1988 7 3 2 1

1989 12 8 3 1

1990 27 13 8 2

In 1991, when writing to Members' Agents about Gooda Walker losses, Mr Jewell commented:

"At the root of the losses lies the fact that the business written was always at risk to high exposure catastrophes. It is well known that, in the years involved, the insurance markets of the world were faced with an unprecedented and unforeseeable concentration of such events in the short space of three years."

Under cross-examination by Mr Eder about this statement Mr Jewell said:

"Each of these events was not unforeseeable. What took the market by surprise is that they happened within such a short period of time."

The Defendants have sought to rely upon the unprecedented sequence of catastrophes to justify the vertical exposure left by the underwriters to each individual catastrophe. Had that exposure been a deliberate and calculated exposure based upon an assessment of the likely incidence of the catastrophes that would impact upon it, this defence would have invited consideration. That, however, was not the position and the unprecedented number of catastrophes that occurred provides no answer to the case that the Plaintiffs have made out in relation to vertical exposure. Nor is the sequence of catastrophes any justification for a failure to match reinstatements.

The unprecedented sequence of catastrophes does, however, provide an answer to the Plaintiffs' contention that I should infer that all their losses are the result of incompetent underwriting - even if one focuses exclusively on losses flowing from excess of loss underwriting.

I have already referred to the manner in which a sequence of catastrophes or loss events can cause losses which are attributable neither to inadequate vertical cover nor to a failure to 'match reinstatements'. The unprecedented and unexpected series of catastrophes that occurred between 1988 and 1990 was, in my judgment, a phenomenon which was capable of resulting in net losses to some, at least, of those writing excess of loss business in a competent manner. It would not be right to infer, from the shortcomings demonstrated in the underwriters' approach to vertical exposure, that losses flowing from horizontal exposure are the consequence of incompetence.

The Losses Sustained by the Syndicates

Syndicate 290

Inadequate vertical reinsurance cover caused substantial losses to Syndicate 290 in respect of two catastrophes. Hurricane Hugo and 90A In the case of each of those catastrophes, the reinsurance cover that Mr Walker had in place to meet the maximum loss was essentially intact, but was greatly exceeded. There were, however, in the relevant period many other, individually less significant, losses. The GWRO report for 31 December 1992 records estimated losses after reinsurance recoveries of $9 million in respect of Phillips Petroleum, $11 million in respect of US Winter Weather and $13 million in respect of Australian Earthquake. In the case of each of these the estimated ultimate gross loss was well below the insurance cover that would have been available on a first loss basis. The Underwriting Report commented:

"The complexities of projecting estimated final gross loss positions and the available reinsurance on each of these disasters are manifold. The problem is compounded by the diversity of the account, the frequency of loss and the fact that the majority of reinsurance policies, both written and purchased, had limited reinstatements...

The Syndicate is not only seriously exposed to individual major losses which have exceeded the amount of reinsurance purchased. Due to the diversity of the account, and as a result of the prior exhaustion of large elements of the reinsurance programme, there is the potential for material amounts of more modest market losses, including numerous aviation losses, to be retained net by the Syndicate."

The Report goes on to comment on potential long tail exposures - presumably inherited under reinsurance to close from prior years.

A year later, the Report recorded a deterioration in the account, due in part to:

"a deterioration in catastrophe losses which largely arises from smaller losses exhausting available reinsurance protection."

In the course of cross-examining Mr Jewell, Mr Eder put to him Schedules of claims paid by Syndicate 290 in relation to the 1989 and 1990 years, as at 31 March 1994. These Schedules showed a large number of paid claims, without indicating the extent to which they were protected by reinsurance cover.

Similar Schedules were not available for Syndicates 298 or 299, but Mr Jewell said that the picture for 298 would be very similar and for 299 "probably this times ten". Mr Eder suggested that the Syndicates had performed well in containing these many event within their reinsurance protections. Mr Jewell replied:

"It is a question of judgment whether they were contained and whether those bits that were not contained within the reinsurance programme were contained within the premium income, and the bit that fell out of the side of that - if it produced a loss for the syndicate. I would not suggest that that is necessarily something that a syndicate should achieve."

This evidence leads me to conclude that a significant part of the losses sustained by Syndicate 290 were caused not because of vertical exposure to catastrophes but because of a large number of loss events which lay below or extended beyond the scope of the syndicate's horizontal cover. The extent of horizontal exposure may have been attributable to incompetent underwriting, but there has been no exploration of this area of the case in the pleadings or in evidence and the Plaintiffs' simple contention that negligence is to be inferred is not made out.

Syndicate 164

Precisely the same considerations that I have developed in the case of Syndicate 290 apply to the results of the excess of loss business written on behalf of Syndicate 164. Excess of loss constituted only part of the business of Syndicate 164 and part of the overall losses of the syndicate may be attributable to other business. In this context I note that the Underwriting Report for 31 December 1992 records reserves being made in respect of anticipated losses on Lloyd's Underwriting Agencies Errors and Omissions policies.

Syndicate 298

Inadequate vertical reinsurance cover is estimated to have caused substantial losses to Syndicate 298 in relation to each of the Five Central Catastrophes. Mr Jewell's evidence suggests, however, that part of the losses was due not to a failure to have in place sufficient vertical cover and matching reinstatements, but to the erosion of lower layers of cover by other loss events. The Plaintiffs have not made out their entitlement to recover losses caused in this way.

I note that the GWRO underwriting report tables gross paid and outstanding claims in relation to several aviation losses amounting to $105 million and warns:

"As a result of the prior exhaustion of large elements of the reinsurance programme, there is potential for material amounts of those losses being retained net by the syndicate."

Mr Andrews was an aviation specialist and had in place what appears to be quite a sophisticated specific reinsurance programme in relation to his aviation excess of loss account. There is no basis upon which I could properly infer that aviation losses suffered by Syndicate 298 result from incompetent underwriting.

Syndicate 299

1988 Year of Account

In 1988, which is a closed year, Syndicate 299 made losses of approximately £22,100,000. Of this a major factor was the loss resulting from vertical exposure to Piper Alpha. As at 31 December 1992 this was estimated at $24 million. That estimate has subsequently been revised downward by $6 million because claims were made against the syndicate in relation to another loss - Enchova - under cover which it had been anticipated would be used to make claims in relation to Piper Alpha. This does not, however, affect the impact that the Piper Alpha had on the 1988 Names at the time of closure.

The Underwriters Report as at 31 December 1990, written by Mr Jewell, said this about the 1988 year:

"It is difficult to justify a loss situation. However, by means of explanation it is worth stressing the fact that, although severe damage to the account was caused by the notorious "Piper Alpha" disaster, there are many other features that contributed to the downfall of this year.

Capacity, both real and created, reached a peak in the World's insurance and reinsurance markets. The resultant scramble for business drove premiums down to totally uneconomic and unrealistic levels.

In addition to Piper Alpha there were many very significant losses, all of which impacted on the account. It could certainly be argued that all of these claims were foreseeable and quotifiable, it is also a truism to point out that the series of claims of the size was unprecedented and therefore certainly not predictable from historical data."

The excess of loss business written by Syndicate 299 was only part of a very broad range of business. I have in mind Mr Jewell's comment that the list of claims on this syndicate would look like the schedules of claims produced in respect of Syndicate 290 times ten. The negligence established by the Plaintiffs in relation to the syndicate's vertical exposure to catastrophes does not give rise to the inference that all the 1988 losses suffered by this syndicate were attributable to negligence.

The 1989 Year of Account

Part of the losses sustained in the 1989 year were attributable to a failure to have in place matching reinstatements to provide cover against catastrophes. Part of the catastrophe losses may be attributable to the erosion of horizontal cover by the series of loss events. Much of the losses are attributable to business other than the excess of loss business written for that year.

Summary

This review of the losses suffered by the individual syndicates demonstrates, in each case, significant causes of loss that cannot be attributed to the negligence that the Plaintiffs have established.

THE APPROACH TO THE ASSESSMENT OF DAMAGES

Having ruled against the Plaintiffs' simple contention that all their losses are recoverable as damages, I have to determine the approach to be adopted in assessing those damages that are recoverable.

The Test

The Plaintiffs claim in respect of breaches of identical duties arising in contract and in tort. On the facts of this case the approach to the assessment of damages is identical, whether the claim lies in contract or in tort.

In The Albazero [1977] AC 774, [1976] 3 All ER 129 at 841 of the former report Lord Diplock summarised the general approach to damages of English law as follows:

"The general rule in English law today as to the measure of damages recoverable for the invasion of a legal right, whether by breach of a contract or by commission of a tort, is that damages are compensatory. Their function is to put the person whose right has been invaded in the same position as if it had been respected so far as the award of a sum of money can do so. Such an award can readily do so in the case of mercantile contracts, since the purpose of the parties in entering into them is to make a money profit. So where the wrong for which suit is brought is the breach of a mercantile contract the measure of damages for the breach is generally the financial loss that the plaintiff has sustained by reason of the Defendant's failure to perform the contract according to its terms."

Applying this approach to the present case the Plaintiffs are entitled to that award of damages which will place them in the same position as if the underwriting carried on on their behalf by each syndicate had been competently performed. That basic test is easy to state but difficult to apply. My task is to define, in as far as I am able as a matter of principle, the methods that should be adopted, the matters that should be considered and the matters that should be disregarded when applying the basic test.

The Plaintiffs' Alternative Case

The Plaintiffs anticipated that their primary case on damages might not succeed and that I might not be persuaded that all their losses were attributable to breach of duty on the part of the Defendants. To meet this possibility they advanced a novel alternative approach to the assessment of damages. I should consider the results achieved by syndicates specialising in excess of loss business in the relevant years and from these deduce the rate of profit or loss achieved by a typical competent excess of loss syndicate ('the paradigm syndicate'). I should then award such damages as would place the Plaintiffs in the same position as if they had been members of that syndicate. The Plaintiffs called Mr Whewell to give expert evidence as to the appropriate actuarial approach to be adopted in selecting the specialist syndicates whose results would form the basis for constructing the results of the paradigm syndicate and the appropriate method of calculating those results.

The Defendants did not accept the validity of this approach to the assessment of damages but, lest I should be minded to adopt it, they in their turn called expert evidence challenging both the selection of the specialist syndicates whose results would form the basis of the exercise and the manner of conducting that exercise.

The novel approach to the assessment of damages suggested by the Plaintiffs has the attraction of avoiding an alternative exercise of much greater complexity. That cannot, however, justify its adoption. There is no stereotype excess of loss syndicate. To create one by taking some form of average or mean of syndicates carrying on different forms of excess of loss business, combined in greater or lesser degree with other business, would be an artificial and unrealistic exercise. Nor would it be realistic to proceed on the premise that had each of the Gooda Walker syndicates conducted its excess of loss business without the deficiencies in approach that the Plaintiffs have demonstrated, their results would have been identical. I can see no alternative to approaching the assessment of the damages sustained by the Names on each of the four syndicates on a syndicate by syndicate basis.

The Damages Recoverable

To attempt to reconstruct the position that the Names would have been in had the Gooda Walker underwriters adopted a competent approach to vertical exposure to catastrophes would be an impossible task. I believe that the reality is that if they and all others who were writing excess of loss business had adopted a competent approach to underwriting, the LMX market would not have existed in the form in which, and with the capacity with which, it in fact developed.

It is for the Plaintiffs to prove that the losses that they have sustained have been caused by the Defendants' breaches of duty. In this Action the Plaintiffs have concentrated, almost exclusively, on the attitudes and actions of the Gooda Walker underwriters in relation to vertical exposure to catastrophes and the consequences that these have had in relation to losses flowing from the Five Central Catastrophes. In the case of each catastrophe, losses have been suffered because the underwriters failed to put in place - whether on a first loss or reinstatement basis - cover that extended sufficiently high to cover the claims arising out of that catastrophe. The Plaintiffs have satisfied me that the resultant exposure was neither an unforeseeable consequence of excess of loss underwriting, nor a consequence that was foreseen by the underwriters as a possible result of a deliberately calculated risk taken on behalf of the Names. It was a consequence of an incompetent disregard of important principles of excess of loss underwriting. In my judgment the Plaintiffs should recover by way of damages such sums as will put them in the same position as if this exposure had been protected by reinsurance. To an extent this is an artificial approach, for in some cases such reinsurance would not have been obtainable. To have regard to such a situation would not, however, absolve the Defendants from liability in damages, but set in train an alternative enquiry of greater complexity which would be unlikely to result in any lesser liability on the part of the Defendants. Accordingly I consider that the test which I have adopted is one which combines relative simplicity with a fair and correct approach in principle.

Subject to what I shall have to say about consequential losses, the Plaintiffs have not satisfied me of their entitlement to recover losses attributable to other causes. In particular they have failed to establish that, in so far as losses are attributable to the impact of other catastrophes or loss events, those losses are the consequence of incompetent underwriting. Thus, in broad terms, the Defendants' submissions as to the limit of their liability in damages succeed, not as a pleading point but because the Plaintiffs have failed to establish any wider liability.

Assessment of Loss

The starting point in the assessment of loss must be to compare the claims relating to a catastrophe with the reinsurance cover available to meet those claims. Not all the shortfall will, however, be recoverable. The loss recoverable will be that attributable to the inadequacy of the vertical extent of the cover. Thus the following will not be recoverable:

1) The consequences of deliberate retention of low layer exposure, co-insurance, or any other form of deliberate retention of exposure. The Plaintiffs have not established that this exposure exceeded that which might have resulted from competent underwriting and have not made out their entitlement to recover losses flowing from it.

2) Losses resulting from short placements by brokers of reinsurance protection ordered. I do not consider that such losses can be attributed to incompetent underwriting.

3) Losses resulting from the erosion by other loss events of reinsurance cover that would otherwise have been available as protection against the Five Central Catastrophes.

The loss recoverable should be further reduced by a percentage to reflect a notional premium that would have been payable for the reinsurance that should have been in place. The rate should be based on the rates actually paid for the upper layers of cover.

Loss Not Established

Rating

I have held in the case of each underwriter that there was a failure to give proper consideration to the adequacy of the rates paid for the upper layers of business written. In assessing damages on the premise that additional high level reinsurance cover would have been available to the syndicates at the prevailing rates, the consequences of this shortcoming will be largely provided for. I find it impossible to formulate a basis for any further award of damages in relation to rating.

The 1989 Year of Syndicate 299

I have held that as a consequence of incompetence on the part of Mr Willard, the 1989 year of Syndicate 299 could not expect to make a profit. The extent to which reinsurance costs exceeded those which might have been carried in a reasonable year is, however, not clear. In the event 1989 was a bad year - both for excess of loss and for marine business. The additional reinsurance placed proved beneficial. I can see no basis for awarding the Plaintiffs more than the losses attributable to exposure to the Five Central Catastrophes.

CONSEQUENTIAL LOSSES

In addition to the losses calculated on the basis outlined above ("the primary losses") certain consequential losses will be recoverable.

Exchange Losses

Exchange losses will be recoverable to the extent that they are consequential upon the primary losses. I do not consider such losses too remote, even when brought about by an event as dramatic as the devaluation of sterling. This type of loss is a normal and foreseeable consequence of suffering a primary underwriting loss.

Some of the exchange losses are attributable to a failure on the part of some of the Names to pay cash calls when due. The Defendants contend that such losses are too remote. Prima facie this would seem correct, but I am not persuaded that this would necessarily be so in all circumstances. The point was not argued and I propose to reserve it for further argument.

Interest

Similar considerations apply to interest. Interest ought to be recovered by Names on their primary losses from the dates that those losses were sustained. Where Names failed to pay cash calls it appears that they may have been debited with interest at a higher rate than the Court would award. As with exchange losses, I propose to reserve issues relating to interest for further argument.

Personal Expenses

In principle personal expenses will be recoverable to the extent only that they have been increased by primary losses.

Run-Off Costs

Similarly run-off costs will be recoverable to the extent only that they have been increased by the primary losses. In this context I find that the incurring of those primary losses was the proximate cause of the syndicates ceasing to trade and being placed into run-off.

STOP-LOSS POLICIES

I now turn to consider arguments advanced by the Defendants as to the significance of stop-loss policies.

Some of the Plaintiffs will have taken out stop-loss policies which will have indemnified them, in whole or in part, in respect of the losses which they seek to recover in this Action. Some of the Plaintiffs will not have taken out such policies. Some of those who have taken out stop-loss policies will have done so on the advice of the Members Agent, that they are now suing.

Some of those who have not taken out stop-loss policies may have disregarded such advice. These various possibilities underline a number of submissions made by the Defendants in respect of stop-loss policies.

Quantum

The Defendants contend that in assessing damages a Plaintiff must give credit for any stop-loss recovery made. It is logical to consider this submission before proceeding to deal with a number of points made in relation to liability.

The Defendants' submission on quantum runs counter to the well established principle that where a Plaintiff takes out an insurance policy against the risk of injury or loss, the monies paid under that policy are not to be taken into account when assessing the damages payable for causing such injury or loss: Bradburn v GW Rly (1874) LR1O Ex.1; Parry v Cleaver [1970] AC 1, [1969] 1 All ER 555. The law considers the contract of insurance to be collateral, or res inter alios acta. Mr Eder submitted that in this case stop-loss insurance was not collateral. It was part of the underwriting strategy of the Name designed to reduce the risk of loss. Where loss was reduced in this way the Members' Agent could not properly be made to pay compensation - particularly where the stop-loss insurance was taken out on his advice - See Eley v King & Chasemore (1892) 2 EG 109.

This argument was advanced in Brown v KMR Services Ltd, where it was more pertinent as the claim was in respect of negligent advice, not negligent underwriting. Gatehouse J dismissed it as contrary to authority. My reaction is the same. The principle in Parry v Cleaver plainly applies on the facts of this case. Stop-loss recoveries have no relevance to the measure of the Plaintiffs damages. In so far as any damages recovered relate to losses in respect of which Plaintiffs have already been indemnified by stop-loss insurers, the recovery will be made for the benefit of stop-loss underwriters - see Napier & Ettrick v Kershaw Limited [1993] 1 All ER 385, [1993] 2 WLR 42 for the recognition by the House of Lords of rights of subrogation in an analogous case.

Causation

The Defendants contended that any Name who was advised to take out stop-loss insurance but failed to do so was thereby responsible for his own loss to the extent that the stop-loss insurance would have provided protection. His failure to follow advice broke the chain of causation. The contention inevitably fails in the light of my ruling that insurance recoveries do not reduce the measure of damage. Whether or not stop-loss insurance was taken out has no relevance to causation.

Voluntary Assumption of Risk

The Defendants have contended (a) that any Name who deliberately did not take out stop-loss insurance was voluntarily assuming the risk of suffering the loss against which such insurance would have provided protection and (b) that any Name who did take out stop-loss insurance thereby demonstrated a voluntary assumption of the risk insured against. Both contentions are equally bad. Stop-loss insurance covered loss, whether caused by negligence or not. No decision to take out such insurance or to refrain from taking out such insurance could amount to a voluntary assumption of the risk of negligence.

Taxation

Difficult issues arise in relation to the effect of taxation on the measure of damages. These became apparent at a late stage of the Action and were not fully developed in argument. It was agreed that they should be reserved for further argument.

DISPOSITION:

Judgment accordingly

SOLICITORS:

Wilde Sapte; Elborne Mitchell