Ace European Group & Ors v Standard Life Assurance Ltd

The Appellants, whom I will call “the Insurers”, were for the policy year 2008/2009 the professional indemnity insurers of the Respondents, Standard Life Assurance Limited, to whom I shall refer as either “SLAL” or “the Assured”. The policy, or strictly a primary policy and three excess policies, afforded cover in total of £100M above a self-insured deductible of £10M. The indemnity given was in respect of the Assured’s civil liability for any third party claims made against the Assured during the policy period arising out of the provision of or failure to provide financial services by the Assured. Both “civil liability” and “financial services” are defined terms. Civil liability is defined to mean:-

“(a)

a legally enforceable obligation to a third party for compensatory damages in accordance with an award of a court or tribunal by whose jurisdiction the Assured is bound; or

(b)

a legally enforceable obligation to a third party for compensatory damages acknowledged by an agreement made, with the consent of the Underwriters, such consent shall not be unreasonably withheld or delayed, between the Assured and third party in settlement of a claim; or

(c)

any compensatory damages pursuant to any award, directive, order, recommendation or similar act of a regulatory authority, self regulatory organisation or ombudsman or following arbitration or other alternative dispute resolution processes whose findings are binding upon the Assured.

…

Compensatory damages shall include civil compensation or damages, compensatory restitution, any other compensatory payment of money or delivery of property of any kind and any settlement agreed by the Underwriters.”

The indemnity provided by the insuring clause includes the Assured’s defence costs and expenses, claimant costs and (if the Assured is liable therefor) co-defendant costs. The insuring clause also provided:-

“This Policy shall also indemnify the Assured for Mitigation Costs.”

“Mitigation Costs” is also a defined term. Mitigation Costs are defined as:-

“Mitigation Costs shall mean any payment of loss, costs or expenses reasonably and necessarily incurred by the Assured in taking action to avoid a third party claim or to reduce a third party claim (or to avoid or reduce a third party claim which may arise from a fact, circumstance or event) of a type which would have been covered under this policy (notwithstanding any Deductible amount).”

In 2009 SLAL faced the prospect of multiple claims arising out of its provision of financial services, namely, the manner in which it had sold and operated its Standard Life Pension Sterling Fund, hereafter “the Fund”. Between 14 January and 10 February 2009 SLAL faced mounting and damaging criticism of its conduct in this regard. The immediate catalyst for that criticism was that with effect from 14 January 2009 SLAL adopted a new method of valuing the asset backed securities in the Fund, different from that which had hitherto been used, thereby producing an immediate one-off one day fall in value of units in the Fund of around 4.8%. The Fund had been marketed to many investors as a temporary home for short-term funds in a manner which suggested that it was invested in cash or the equivalent of putting money on deposit. It was a surprise to many to find that their money had been invested in asset-backed securities, let alone that a new and more reliable method of valuing them needed to be adopted. It is no exaggeration to say that SLAL faced something of a crisis. In order both to avoid and to reduce the claims, both actual and potential, and to avoid or reduce further damage to the Standard Life brand, on 10 February 2009 SLAL resolved to make a lump sum payment into the Fund which would have the effect of restoring its value to what it had been before the revaluation. It resolved also to compensate those who had sold units in the Fund between 14 January and 11 February 2009 on the same basis, i.e. by making a payment to them which restored the value of their units to what it had been before the reduction. The lump sum payment into the Fund, £81,999,935 and payments to customers in respect of units sold between 14 January and 11 February, £19,862,113, in total £101,862,048, was referred to as “the Cash Injection”.

SLAL sought to recover the Cash Injection, and some associated payments to which I will refer hereafter, from the Insurers. Insurers denied liability.

The key issue at trial in the Commercial Court was whether the Cash Injection fell within the definition of Mitigation Costs. Insurers argued that all or some of the Cash Injection should properly be regarded not as made for the purpose of avoiding or reducing claims, but rather as made for the dominant purpose of avoiding or reducing potential brand damage. The judge, Eder J, found that the Cash Injection was reasonably and necessarily incurred by Standard Life in taking action to avoid or reduce third party claims of a type which would have been covered under the Policy. The judge also found however that another, equally efficacious intended objective of making the Cash Injection was to avoid brand damage.

The Insurers argued at trial that, consistent with those findings, there should be an apportionment of the Cash Injection between “the insured and uninsured interests at risk and sought to be preserved by the Cash Injection”. SLAL had, submitted Insurers, assessed the brand damage if they did not make the Cash Injection as likely to be of the order of £300M. SLAL had, moreover, assessed the potential liabilities to pay third party claims that were averted by the Cash Injection as substantially exceeding the policy limit of £100M. It followed, argued Insurers, that in broad terms the amount of the Cash Injection which should be regarded as properly referable to the avoidance of claims within the policy limit of £100M was 25%.

Eder J held that no apportionment was required and that SLAL was entitled to recover the full amount of the Cash Injection, subject only to the deductible. At paragraph 185 of his judgment he said this:-

“. . . It is my conclusion that in order to succeed and so far as this part of the definition of Mitigation Costs is concerned, SLAL must show that the costs claimed were expected and intended to avoid or to reduce third party claims of the stipulated type; and SLAL will be entitled to recover in full even if such costs were also incurred for some other purpose.”

It is against that conclusion that Insurers appeal. It is common ground for the purposes of this issue that the Cash Injection met the Policy’s definition of Mitigation Costs.

There is a subsidiary ground of appeal to the effect that part of the Cash Injection should be irrecoverable as representing a payment to persons who have no claim because no misrepresentation had been made to them as to the nature of the Fund. Reinstatement of the Fund to its 14 January value had to that extent conferred on such holders of units in the Fund a windfall which could not be regarded as a payment reasonably and necessarily incurred in taking action to avoid or reduce third party claims.

The judge was faced with a major contest on the facts and the issues which arise on appeal represent only a fraction of the coverage and other legal issues which he had to resolve. His judgment is very long, running to 90 single spaced pages and 264 paragraphs. The curious can find it at [2012] EWHC 104 (Comm). In order to set the issues on this appeal into their factual context it is necessary only to put a little flesh on the bones which I have already sketched out. In doing so I have borrowed heavily from the parties’ helpful skeleton arguments prepared for the appeal.

The Fund was launched in July 1996 and, as already indicated, marketed as a temporary home for short term funds, with some literature referring to it as being invested in “cash” and even as being the equivalent of putting money on deposit. However, by 2007 the Fund’s assets included a substantial proportion of asset backed securities. By early 2008 the Fund was valued at approximately £2.3 billion, around 50% of which was invested in asset backed securities and floating rate notes.

From late 2007 and especially after the collapse of Lehman Brothers in September 2008, market prices of asset backed securities began to fall and securities of that type became increasingly illiquid, making their valuation more and more subjective. Units in the Fund fell in value from time to time during this period. With effect from 14 January 2009 Standard Life switched to new sources of prices, resulting in a one-off one day fall in the value of units in Fund of around 4.8%. The overall fall in value of the Fund over its entire life from launch up to and including 14 January 2009 was about 9.8%, equivalent to about £190M.

Following the one day fall in value on 14 January 2009, Standard Life received widespread criticism in the media, a mass of complaints and claims from individual customers and Independent Financial Advisors, “IFAs”, and close scrutiny from the Financial Services Authority. The key issues raised were that the Fund had been misleadingly described in the marketing literature and that customers and IFAs had been led to expect that it was a secure investment. Between 14 January and 10 February 2009 SLAL closely considered its position in light of the adverse reaction of customers, IFAs and the media. By early February 2009 SLAL had come to appreciate that unless it acted quickly to reverse the price fall, it was confronted by the likelihood of a very substantial loss of present and future business affecting retail customers and corporate clients. That is what I have already described, as did the judge, as damage to the brand.

Intensive consideration by SLAL of the question how best to respond to the issues raised by customers, IFAs and by the FSA threw up two options. Option 1 was to set up a scheme to compensate customers and to invite all customers to make claims for compensation which would be assessed on a case by case basis. Option 2 was to restore the one day 4.8% fall in the value of the fund, while also considering any individual claims for further compensation over and above that fall in value. An investor to whom the Fund had been mis-sold might not be satisfied by restoration of the Fund’s 14 January 2009 value and might seek compensation in respect of his further loss sustained over the earlier period. It was common ground at trial that these were the only two possible options in the circumstances that Standard Life then faced.

It is important to understand the relative cost of these options as they were perceived at the time. Needless to say the options were not and could not be costed in an accurate manner. Rather, SLAL had certain “reference points”.

Thus the maximum cost of Option 1, before taking into account administrative costs, would be £124M. This, broadly, represents 64% of the £190M figure to which I have already referred at paragraph 11 above. Roughly 64% by value of those invested in the Fund had received misleading literature on the basis of which they could mount a claim for mis-selling. The working assumption was that, if invited to claim, customers would not limit their claims to the one-off fall in value sustained in January 2009 but would claim in respect of the loss in value of the Fund attributable to its unexpected investment in volatile securities. Hence the figure of £124M, being 64% of the overall fall in value of the Fund since launch. That figure in turn assumes a 100% response rate, i.e. that all those entitled to claim would make a claim. There were in fact relatively few investors in the Fund, in the sense that 4% of the customers by number held 50% of the value of the Fund, 30% by number had 90% of the Fund. SLAL needed to take into account the prompting to make a claim that customers would receive from intermediaries and the media and, potentially, from claims management companies.

However Option 1 also had attached to it a further less tangible cost in terms of brand damage. It was seen as inevitable that Standard Life would lose both business and reputation if Option 1 were adopted. That brand damage was costed in broad terms at £300M. £240M of this was made up of lost new business commission, on the assumption that sales, as compared to 2008, would be 30% down in 2009, 15% down in 2010 and 5% down in 2011. The remaining £60M comprised the anticipated cost of additional brand marketing over the course of the next 36 months.

Option 2 had one fixed and one unknown element of cost. The Cash Injection itself was 4.8% of the Fund prior to the one day reduction in value, i.e. £101,862,048, paid in part as payment into the Fund to restore its value and in part as direct payments to those customers who had disinvested in the Fund since 11 January, restoring them to the position in which they would have been had they disinvested immediately prior to the one-off reduction. The second element would be the cost of compensating on an individual basis any customer with a valid claim to recover the additional fall in value of the Fund over and above that represented by the one-off one day reduction in value. However the view was formed that implementation of Option 2 would generally reduce the levels of indignation which had built up amongst policy holders and their advisers, reinstatement of the Fund value being expected to be perceived as “the right thing to do”. Certainly the CEO, Sir Sandy Crombie, thought that if Option 2 were implemented the cost of additional claims over and above the reinstatement of value to the January 2009 level would be confined to single digit millions of pounds. The thinking was therefore that implementation of Option 2 would lead to a much reduced level of claims for compensation for the additional losses sustained over the life of the Fund than would be likely to result from adoption of Option 1.

SLAL’s management presented these two options to the Standing Committee of the Board of Standard Life plc at a meeting on 10 February 2009. As it happens there was excluded from the position paper presented to the Standing Committee any express mention of the reference cost of £124M attaching to Option 1. It was thought that there was a danger of spurious accuracy. The figure was however buried in Appendix G to the position paper, a Note from the Actuary, although there was at that point added to it, without explanation, £15M for the cost of administering a mis-selling unit, the resultant figure being rounded up to £140M. I mention this because Mr George Leggatt QC, for SLAL, was at pains to emphasise that some of the decision makers may not have had in mind the reference cost figure attributed to claims if the value of the Fund was not to be reinstated. I doubt if this much matters in the light of the judge’s findings. Certainly the view of management to which some attendees of the meeting on 10 February 2009 were privy was that Option 1 was likely to be more expensive than Option 2 even ignoring brand damage. That was undoubtedly the view expressed by the CEO at the meeting. At this meeting the Standing Committee accepted the management’s recommendation to adopt Option 2 and to reject Option 1. Pursuant to that decision, the total amount paid to restore the 4.8% fall in value of the Fund, the Cash Injection, was £101,862,048. Further payments have since been made to individual customers who pursued complaints for additional losses. As at 25 January 2012 these payments totalled £4,831,390, bringing the total of all “remediation payments” made by that date to £106,693,438.

Against that background I should next record the judge’s findings on the two main issues which were a matter of fierce debate before him. It was the Insurers’ case at trial that (a) the Cash Injection was made with the dominant purpose of avoiding or reducing brand damage and (b) that the Cash Injection was not “necessary” to avoid or to reduce third party claims when considering the definition of mitigation costs. The Insurers lost on both these points and they have been refused permission to appeal against the judge’s findings in relation thereto. The findings therefore against the background of which Insurers pursue this appeal include the following:-

“218.

However, contrary to the Insurers’ submissions, it seems equally clear to me that, irrespective of brand damage, the decision was taken at the meeting on 10 February and the payment of the Cash Injection was made in order to avoid or to reduce third party claims of a type which would have been covered under the Policy within the meaning of “Mitigation Costs” as defined in the Policy.

. . .

245.

For all these reasons, it is my conclusion that the Remediation Payments including the Cash Injection were made in order to avoid or to reduce third party claims of a type which would have been covered under the Policy within the meaning of “Mitigation Costs” as defined in the Policy. In summary:

(i)By the time of the meeting on 10 February 2009, Standard Life had already received a considerable number of third party claims and faced the prospect of many more if no action was taken to avert them.

(ii)It is common ground that the only possible alternative to Option 2 (i.e. Option 1) would have involved writing to all customers with holdings in the Fund acknowledging that they may have grounds for complaint and informing them how to complain. That action, if the 4.8% price fall had not been reversed, would have encouraged claims rather than doing anything to avert them.

(iii)It is clear from the evidence that Standard Life believed – entirely correctly as subsequent events proved – that reversing the 4.8% price fall would remove what was believed to be the key cause for complaint for many customers and reduce the levels of indignation which had built up among customers and their advisers – with the result that far fewer of them would pursue claims for compensation (including claims for more than the 4.8% fall) than if Option 1 were adopted.

(iv)Standard Life also obviously believed that the quantum of any claims would be reduced if the complainants had already been compensated to the extent of the 4.8% fall.

(v)

Thus, in Standard Life’s belief – and on any reasonable view of the matter – the number and value of third party claims was expected to be very substantially less if the 4.8% price fall was reversed by adopting Option 2 than if no such action was taken and Standard Life simply wrote to all customers informing them how to complain.

246.

In addition, there is no doubt in my mind that the Cash Injection was reasonably incurred. However, that is not sufficient for the purposes of the definition of “Mitigation Costs”. The question remains to consider whether such payment was also “necessarily incurred” which is an important part of the definition of “Mitigation Costs”. I have already considered the scope of this part of the wording of the definition earlier in this Judgment. As there stated, I accept the Insurers’ submission that this sets a “high threshold” albeit one which must have regard to the realities. In that context, I recognise and accept that there was no legal necessity to make such payment. However, the existence of a legal obligation is not, in my view, the relevant test of “necessity” and the relevant “payment” does not have to be made to discharge a particular liability of a particular third party claimant. I also recognise and accept that at least at the time when the Cash Injection was made and absent any “recommendation” from the FSA, it was no doubt open for Standard Life to have adopted Option 1, waited for the claims to come in and then considered each claim in turn. That was certainly a possible course of action and, to that extent, I also recognise that the adoption of Option 2 and the immediate payment of the Cash Injection might be said to be not “necessary” in that sense. However, even recognising the “high threshold” imposed by such wording, it is my conclusion that the “realities” did mean that the Cash Injection was necessarily incurred to avoid or to reduce relevant third party claims. As discussed above, it seems to me important to read the various parts of the definition as a whole. In other words, it was necessarily incurred bearing in mind the stipulated purpose. For the avoidance of doubt, I should repeat that, as stated above, I accept the Insurers’ submissions that the Cash Injunction (sic) [Injection] was also incurred in order to avoid or to reduce “brand damage”. However, in my view, both intended objectives were equally efficacious and, for the reasons stated above, this does not affect SLAL’s entitlement. . . . .”

I should stress that the judge’s reasons for these conclusions and his discussion of the evidence from which he distilled them are exhaustively set out in the many preceding paragraphs. Reference to “the realities” was no doubt a reference to the circumstance that there were only two possible options and that Option 2 was going to be the cheaper of the two in terms of claims.

The judge had already concluded, at paragraph 185 of his judgment, which I have reproduced at paragraph 7 above, that if SLAL could show that the costs claimed were expected and intended to avoid or reduce third party claims of the type stipulated in the definition of Mitigation Costs then SLAL would be entitled to recover them in full even if such costs were also incurred for some other purpose. Accordingly, he held SLAL able to recover all the Remediation Payments, including the Cash Injection, without apportionment subject only to the single deductible of £10M.

I turn to the main argument on the appeal. Mr Alistair Schaff QC for the Insurers submits that unless the policy otherwise provides, there is a rule of law which requires in cases such as the present an apportionment of the mitigation costs by reference to the respective insured and uninsured interests at risk and sought to be preserved by that expenditure. Here neither brand damage nor liabilities to third parties in excess of £100M in aggregate were insured interests. He proffered three alternative apportionment calculations, depending upon whether one took into account £190M of maximum potential liabilities, £124M of perceived liabilities or £100M of liabilities reduced or avoided by the Cash Injection. Each calculation, because of the influence within it of the figure of £300M attributed to brand damage being avoided or reduced by the Cash Injection, resulted in SLAL’s recovery being less than 25% of the £96M odd which it has recovered under the judgment.

I believe that this argument founders at the first hurdle, which is that it is untenable as a matter of construction of the Policy. Under the Policy the Insurers undertake “to indemnify the Assured for Mitigation Costs”. The judge has found that the Remediation Costs including the Cash Injection meet all of the criteria which are required in order to satisfy the definition of Mitigation Costs in the Policy. These are therefore costs which the Insurers have undertaken to pay. Any apportionment of the costs would involve the Insurers failing to honour their promise to indemnify the Assured for Mitigation Costs. As Mr Leggatt pointed out, and as is obvious, the promise is not to pay some part or element of what may constitute Mitigation Costs but to pay all costs which are Mitigation Costs, subject of course to any relevant Policy exclusions and limits. The whole payment was incurred for the relevant purpose and I can see no basis upon which it can be said, consistently with the plain words of the insuring clause, that because the payment also achieved and was intended to achieve another incidental objective so it is in some part irrecoverable as Mitigation Costs. On the facts of the present case the argument for apportionment is I think manifestly unfounded when one bears in mind that the Cash Injection was one indivisible sum which had to be paid in full in order to restore the value of the Fund. Option 2 could not be performed by a payment of less than the full amount of the Cash Injection, since only by payment of that amount could the value of the Fund be restored. Given the undertaking to pay Mitigation Costs, there is simply no basis upon which Insurers could perform their promise other than by payment of the full amount, subject of course to any applicable limits.

I agree with SLAL that a useful analogy is to be found in the principle enunciated in J J Lloyd Instruments v Northern Star Insurance Company Limited, the “Miss Jay Jay” [1987] 1 Lloyd’s Rep 32 to the effect that where a loss is caused by an insured peril, the entitlement of the insured to an indemnity is not affected by the fact that there is another equally effective cause of the loss which is uninsured.

Mr Leggatt also drew to our attention a series of cases concerned with the undertaking, common in liability policies, to pay costs and expenses incurred in the investigation and defence or settlement of claims made against the insured. The cases were Capel-Cure Myers Capital Management v McCarthy [1995] LRLR 498; New Zealand Products v New Zealand Insurance [1997] 1 WLR 1237; Thornton Springer v NEM Insurance [2000] Lloyd’s Rep IR 590 and John Wyeth v Cigna [2001] Lloyd’s Rep IR 420. In each of those cases, albeit in different contexts, there arose the question whether costs which had been incurred to defend both insured and uninsured claims were recoverable in full or fell to be apportioned. In each case the court rejected the principle of apportionment. It is right to say, as Mr Schaff points out, that defence costs, where recoverable, are dealt with by a specific term in the policy and that the question in each case turned upon the construction of that clause, as Lord Clyde emphasised in giving the advice of the Judicial Committee of the Privy Council in the New Zealand Products case at page 1242. There is however an obvious analogy between acceptance of a liability for costs incurred in defending claims and acceptance of a liability for the cost of avoiding or reducing claims. I also acknowledge that in none of those cases was it suggested, as it is here by Mr Schaff, that apportionment should be introduced by analogy with the rule which prevails in relation to sue and labour clauses in marine property insurance. That argument was not addressed. Nonetheless, the approach to the construction of these clauses does in my view again afford a useful analogy which is supportive of the view I have expressed above.

For my part I would be content to dispose of this appeal on the short ground that as a matter of construction of the insuring clause apportionment is simply unavailable because it would result in insurers paying less than they have by unambiguous language promised to pay. It is inconsistent with the clear language of the policy.

However, out of deference to the interesting and learned argument deployed by Mr Schaff in support of the wider principle, I propose briefly to explain why in my view there is in liability insurance no place for a principle of the sort which he advocates.

The starting point of his argument is that the supplementary provision enabling the Assured under a liability policy to recover mitigation costs is analogous to a sue and labour provision traditionally found in marine policies on ship and goods. The provision which appeared in the original form of the Lloyd’s S G Policy was as follows:-

“In case of any loss or misfortune it shall be lawful to the assured, their factors, servants and assigns, to sue, labour and travel for, in, and about the defence, safeguard and recovery of the said goods and merchandises, and ship, etc., or any part thereof, without prejudice to this insurance, to the charges whereof we the assurers will contribute.”

Section 78 of the Marine Insurance Act 1906 is sub-titled “Suing and Labouring Clause”. It provides:-

“(1)

Where the policy contains a suing and labouring clause, the engagement thereby entered into is deemed to be supplementary to the contract of insurance, and the assured may recover from the insurer any expenses properly incurred pursuant to the clause, notwithstanding that the insurer may have paid for a total loss, or that the subject-matter may have been warranted free from particular average, either wholly or under a certain percentage.

(2)

General average losses and contributions and salvage charges as defined by this Act, are not recoverable under the suing and labouring clause.

(3)

Expenses incurred for the purpose of averting or diminishing any loss not covered by the policy are not recoverable under the suing and labouring clause.

(4)

It is the duty of the assured and his agents, in all cases, to take such measures as may be reasonable for the purpose of averting or minimising a loss.”

Section 78(3) thus prescribes averaging in the case of expenditure incurred to avert damage to both insured and uninsured property. Mr Schaff points out that the archetypal example of where this occurs in the marine property context is that of underinsurance. Thus, if £100 is paid in sue and labour expenses to save from loss property valued at £400 but insured only for £100, 25% of the expenditure is attributable to the insurers’ interest in the property and 75% is attributable to the insured’s uninsured interest in the property.

Mr Schaff suggests that underinsurance is just one example of a situation in which apportionment is appropriate to reflect that it is both insured and insurer who are intended to, and do, benefit from the expenditure in question. He points to some observations made by Rix J, as he then was, in Royal Boskalis Westminster N.V. and Others v Mountain and Others [1997] Lloyd’s Reinsurance LR 523. That was a case concerned with marine property insurance against war risks, the relevant policy containing a sue and labour clause. The Dutch insured were the owners of a valuable fleet of dredgers engaged in dredging work in Iraqi waters when Iraq invaded Kuwait in August 1990. As a riposte to UN sanctions the Iraqi Revolutionary Command Council promulgated a law to the effect that the assets of companies incorporated in those countries which had enacted sanctions legislation against Iraq should be seized. The Netherlands was such a country. To all intents and purposes both the fleet and its personnel both aboard and on shore managing the joint venture in which it was workingwere detained in Iraq. The Dutch owners of the fleet negotiated an agreement with the Iraqi authorities pursuant to which both the fleet and its personnel left Iraq in safety a month or so before the outbreak of war between Iraq and the USA and its allies in January 1991. The agreement involved the insured waiving and forgoing substantial claims which the joint venture would otherwise have had against the employer, the General Establishment of Iraqi Port, an arm of the Iraqi Ministry of Transport and Communications. The insured claimed to recover the value of the forgone claims as sue and labour expenses.

Rix J found that the expenditure had been incurred both to avoid or minimise the loss of the fleet, which was insured, and to avoid or minimise the continued detention of the personnel, who were not insured, at any rate under this policy on goods. Rix J concluded that the expenses were recoverable as sue and labour expenses but would be apportioned as to 50% incurred for the purpose of avoiding or minimising loss of the insured property and as to 50% incurred for the purpose of avoiding or minimising the continued detention of the uninsured personnel. It was, he said, “one package” with a “dual purpose”. Since no financial value could be put on the safety of the personnel, Rix J ascribed, anomalously, an equal value to the two interests preserved by the inextricable dual purpose. Rix J regarded it as legitimate to extend the principle of averaging to expenses incurred for the dual purpose of protecting insured property and protecting some other interest. Rix J first referred to the classic exposition of averaging in the context of underinsurance set out by Walton J in Cunard Steamship Co Ltd v Marten [1902] 2 KB 624. A pointin issue in that case was whether the policy sued upon by the shipowners was a policy on goods or a policy affording to the shipowners an indemnity in a limited amount against liability to the owners of the goods carried. Having determined that it was the latter and that the sue and labour clause formed no part of the policy, Walton J said this, at page 629:-

“Again, the suing and labouring clause undoubtedly contemplates and implies that, whilst the underwriters are to bear their share of any suing and labouring expenses, they are to bear such share only in the proportion of the amount underwritten to the whole value of the property or interest insured. If the assured has insured himself of goods to the extent of one-half only of the value of his property or interest in the goods insured, he, in respect of each and every item of suing and labouring expense, recovers one-half and bears one-half himself. This is the perfectly well-established basis of every adjustment of suing and labouring expenses.”

After setting out this passage, Rix J continued, in Royal Boskalis, at page 602:-

“The case went to the Court of Appeal [1903] 2 KB 511 where Lord Justice Vaughan Williams adopted the reasoning of Mr Justice Walton in full (at p. 515). Mr Aikens submits that this passage is of no relevance merely applying the principle of “averaging”, as when an insured is underinsured. In my view, however, the reasoning is directly related to the concept of suing and labouring: if half the goods must be treated as uninsured (the averaging point), then the sue and labour expenditure must be apportioned between the goods insured and uninsured. This, in my judgment, makes good sense. If a merchant chooses to insure only half the goods in a vessel, and then sues and labours to save his goods as a whole, he should receive only an apportioned share of his sue and labour expenses from his insurer. If a shipowner insures only one of his two vessels, and then sues and labours to save the pair of them, a similar apportionment should apply. It may be different if an uninsured loss is prevented only incidentally by sue and labour expenditure aimed at the protection of some other insured property; or if, again a somewhat different case, the assured’s purposes were manifold, but one was subsidiary to another.”

The significance of the observations of Rix J, and their application, is, obviously, that Rix J envisaged that averaging or apportionment might be appropriate not just in a case of underinsurance but also in a case where the relevant sue and labour expenditure is incurred both for the purpose of protecting and preserving insured property and for the purpose of preserving or protecting some other uninsured interest. For so long as that other interest is property capable of insurance, the suggested extension is a small one, although as I shall show nonetheless unprincipled. If however the uninsured interest is different in kind from insurable property or incapable of valuation, as in that case it in fact was, the suggested extension is unworkable, as pointed out when the Royal Boskalis case reached the Court of Appeal.

In the Royal Boskalis case insurers appealed to the Court of Appeal. In the Court of Appeal it was held that the agreement reached with the Iraqis was unenforceable as a result of which the insured would have been able to maintain against the General Establishment of Iraqi Port the claims apparently forgone. The insured could not therefore demonstrate that entering into the agreement resulted in a loss recoverable as sue and labour expenditure. The observations of Rix J were thereby rendered technically obiter, although they are of course entitled to the greatest respect. Phillips LJ in the Court of Appeal also indicated that had the point arisen he would not for his part have apportioned the expenditure. At page 647 of the report, (also reported at [1999] QB 674, 738-9) after again setting out the passage from Walton J’s judgment in Cunard Steamship v Marten, which I have already reproduced above, he said this:-

“I do not believe there to be any doubt that where ship or cargo is under-insured, sue and labour expenses will only be recoverable in the same proportion that insured value bears to actual value. In such circumstances, it is possible arithmetically to apportion the expenses and thus identify, with only a modest degree of artificially (sic) [artificiality], that portion of the expenses incurred for the benefit of the insured, as opposed to the uninsured, property.

It is also true that lives can be the subject of insurance, and that it is possible to insure against liability to pay life salvage. Those who are interested in ship and cargo do not usually, however, have insurable interests in the lives of crew or passengers. It must frequently be the case that, just as in the case of salvage and general average, sue and labour expenses are incurred, in part, for the benefit of lives which are also at risk as a result of the insured peril. Salvors who save lives as well as property have their award against ship and cargo enhanced to reflect that fact, and underwriters of ship and cargo between them bear the whole cost – The Bosworth No 3 [1962] 1 Lloyd’s Rep 483. Never before has it been suggested that liability under the sue and labour clause should be reduced to reflect the fact that the exertions in question have been motivated in part by a desire to save lives. In such circumstances, as the Judge recognized, it is impossible to carry out an arithmetical apportionment between property and lives at risk. The reality is that the entirety of the expenditure is directed to two objectives which are different in kind. Preservation of life cannot be equated with preservation of property. Provided that the expenses can reasonably be said to have been incurred for the preservation of the property, it does not seem to me either sound in principle or desirable that the assured should be penalised if they were sufficiently concerned for lives at risk to have been concerned to save not only their property but those lives.

For these reasons I consider that the Judge should have held the joint venture entitled to recover the full cost of entering into the finalization agreement rather than only half that cost. As the joint venture failed to establish that there was any cost at all, the success of the Cross Appeal can do no more than rub salt in the wound.”

I would accept that the thrust of these remarks of Phillips LJ is directed to the point that preservation of life cannot be equated with preservation of property. I would also accept that Phillips LJ was not intending to discuss the broad question whether averaging should be extended beyond the confines of marine property insurance. On the other hand there are two aspects of his remarks which deserve emphasis. One is the reaffirmation that averaging or apportionment is a concept having its origin in underinsurance of ship or cargo. The other is the observation that an arithmetical apportionment is likely to be difficult if not impossible if the sue and labour expenditure is directed to two objectives which are different in kind.

Within three days of delivering his monumental judgment in Royal Boskalis Rix J also delivered a long judgment in Kuwait Airways Corporation v Kuwait Insurance Company [1996] 1 Lloyd’s Rep 664. That was a case of aviation insurance, not marine insurance. KAC’s aircraft were insured up to a limit of US$ 300M in respect of war risks to which they might be subject whilst on the ground. Aircraft to a value of US$ 692M were seized by invading Iraqi forces and flown out of Kuwait. Spare parts to a value of over US$ 300M were likewise seized and removed. War risks underwriters paid an indemnity of US$ 300M. There was a sue and labour clause, which the judge regarded as highly unusual in a non-marine policy, although he was told that such clauses are found in aviation policies. KAC incurred sue and labour expenditure in effecting recovery of some of the seized aircraft. A question arose whether sue and labour expenses were recoverable in addition to the overall limit of liability, US$ 300M and, if so, whether underwriters were responsible for all of the expenses or only responsible in the proportion 300/692 as representing the insured interest. Rix J held that sue and labour expenses were payable over and above the insured limit, a conclusion which was ultimately overturned in the House of Lords rendering obiter the remainder of Rix J’s observations on this topic. Rix J adopted the same approach to apportionment as he had adopted in Royal Boskalis. At page 697 he said this:-

“It seems to me to be clear that, by analogy with the sue and labour principles developed in the marine context, an assured is only entitled and an insurer only liable for a due proportion of expenses, in accordance with their respective interests at stake at the time of the incurring of the expense in question . . . .”

Then, after citing s.78(3) of the Marine Insurance Act 1906, which he regarded as stating a rule derived from the common law, and setting out once more the relevant passage of Walton J in Cunard v Marten, he continued, at page 698:-

“It seems to me to follow that sue and labour expenses incurred to avert or diminish a loss will have to be apportioned in accordance with assured’s and insurer’s prospective interests and potential liabilities in respect of the loss; and sue and labour expenses incurred to recover insured property will have to be apportioned in accordance with assured’s and insurer’s respective interest in potential benefit from the recovery . . .”

Mr Schaff relies upon the fact that Rix J rejected a submission that averaging or apportionment applies only in the field of marine insurance. Nonetheless, the extension was only to another class of property insurance. However, Mr Schaff can point to the fact that in the Court of Appeal, where the point did not arise for a different reason from that ultimately adopted in the House of Lords, Staughton LJ saw no objection in principle to “apportionment in accordance with the assured’s and the insurer’s respective interests and potential liability in respect of that loss”.

The final step in the argument was for Mr Schaff to ask, rhetorically, why what he described as the default position should not equally apply in liability insurance. In cases of dual purpose, one object being insured and one uninsured, it should he submitted make no difference whether the insured interest is to avoid property damage or to avoid insured liabilities. In both situations, the expenditure is being incurred partly to benefit the insured alone and partly to benefit the insured and insurer jointly. Such expenditure should not fall on the insurer alone. Indeed, Mr Schaff goes further. He asked what would happen if the uninsured interests had in fact been the subject of insurance cover through a different insurance programme; and he hypothesises that one could imagine a situation where an insured takes out two liability insurances, covering different liabilities, or a situation where an insured takes out one liability insurance and one insurance covering its commercial interests. Why, asks Mr Schaff, should mitigating expense designed to avoid liability or loss under both policies fall exclusively to the account of one insurer alone?

Attractively though this argument was presented, it is I think misconceived, essentially for the reasons put forward by Mr Leggatt in his elegant refutation of it. There is no default position applicable outside the field of marine property insurance. The approach in that field derives from the practice of adjusting marine losses of ship or cargo upon the assumption that the subject matter assured is fully covered by insurance. That assumption is inapposite and in any event unworkable in the field of liability insurance to which the principle cannot therefore be transferred.

Mr Leggatt first drew attention to s.67 of the Marine Insurance Act 1906 and the note thereon by Sir Mackenzie Chalmers himself in the first edition of his Commentary on the Act. Section 67 of the Act under the sub-title of Measure of Indemnity provides as follows:-

“67.

(1) The sum which the assured can recover in respect of a loss on a policy by which he is insured, in the case of an unvalued policy, to the full extent of the insurable value, or, in the case of a valued policy, to the full extent of the value fixed by the policy, is called the measure of indemnity.

(2)

Where there is a loss recoverable under the policy, the insurer, or each insurer if there be more than one, is liable for such proportion of the measure of indemnity as the amount of his subscription bears to the value fixed by the policy, in the case of a valued policy, or to the insurable value, in the case of an unvalued policy.”

Chalmers’ note to this section reads as follows:-

“Insurance is a contract of indemnity, but in marine insurance the indemnity is conventional, and the following sections supply the standard or measure for ascertaining it. The adjustment of marine losses proceeds upon the hypothesis that the subject-matter insured is fully covered by insurance. Suppose a ship valued at £10,000 is insured for £1000 only. The shipowner is said to be “his own insurer” for £9000, and any loss which occurs must be adjusted on this basis, see section 81. The following cases may be put in illustration of this principle:-

1.

A cargo valued at £10,000 is insured for £1000 by ten under-writers, who each subscribe for £100. It is damaged by sea perils to the extent of £1000. Each underwriter is liable for £10 only.

2.

A ship valued at £5000 is insured for £1000. The ship is stranded, and the owner spends £1000 in trying to get her off, but eventually she is totally lost. The insurer must pay £1000 on the policy, and £200 (i.e. one-fifth) under the suing and labouring clause. It is immaterial whether the real value of the ship be £4500 or £5500.

As to the suing and labouring clause, which is a distinct engagement in the policy, see section 79; and for a quasi exception, see section 74.”

Section 81 of the Act, to which the commentary cross refers, provides:-

“Where the assured is insured for an amount less than the insurable value, or, in the case of a valued policy, for an amount less than the policy valuation, he is deemed to be his own insurer in respect of the uninsured balance.”

I have already referred at paragraph 31 above to the issue which arose in Cunard v Marten. It concerned a policy to protect shipowners against “liability of any kind” up to £20,000 to the owners of mules shipped on board their vessel arising from the omission from the contract of affreightment pursuant to which the mules were carried of a standard form clause exempting shipowners from liability for loss of the mules through the negligence of the shipowners’ servants. The mules exceeded £20,000 in value. During the voyage the vessel stranded through the negligence of the shipowners’ servants and expenses were incurred by the shipowners in saving some of the mules and in attempting to save others which were lost. The shipowners sought to recover these expenses, not as a direct loss under the policy, but under the suing and labouring clause as expenses incurred to avert or reduce the amount of the loss. In the course of his judgment Walton J said this, at page 629:-

“If, however, the policy is not to be treated as a policy “on goods”, but as a contract of indemnity against a certain kind of liability up to a certain limited amount, it is very difficult to apply the suing and labouring clause to such a contract. That clause applies when there is a suing or labour for the safeguard and recovery of “the said goods”; that is to say the goods insured. As I have said, this is not an insurance on goods.”

There then follows the passage cited by Rix J in Royal Boskalis and Kuwait Airways Corporation, and indeed by Phillips LJ on appeal in the Royal Boskalis case, followed by this:-

“But how can this be applied in the case of a contract of indemnity against liability to a limited amount such as is here sued upon? There might be liabilities covered by the present policy which did not depend upon the safety of the goods. But assuming that the liability can be measured by the value of the goods – that is to say, the mules, and depends upon their safety, how is the proportion of suing and labour expenses to be borne by underwriters to be arrived at? In the one case, where all the mules are on board the ship and are all in equal danger of total loss, and an expense is incurred to avert such loss – as, for instance, by towing the ship when sinking and placing her in safety on the beach – there would be no difficulty. But, assuming again that the total value of the mules is 40,000l., and the liability covered is up to 20,000l., as in the present case, and expense is incurred in saving the mules by getting them ashore one by one, and mules to the value of 20,000l. are thus saved and the rest lost, how is the expense to be apportioned? In such a case it might very well be said that the whole expense was for the benefit of the assured, and not for the benefit of the under-writers at all. Intermediate cases would present even greater difficulties.”

In the course of his judgment Walton J also referred to Joyce v Kennard (1871) LR 7 QB 78 as being, like Cunard v Marten, concerned with a policy which was “not an ordinary marine policy,” i.e. not a traditional policy on ship or goods on the traditional Lloyd’s SG Form. That case again concerned a liability policy, whereunder lightermen insured their liability in respect of goods carried in their vessels. The policy was subscribed by different underwriters in an amount totalling £2000. The defendant underwriter subscribed a line of £100. The lightermen incurred a liability of £1100. At the time, the total value of goods at risk on the water in their barges was £20,000. The defendant contended that his liability was limited to such proportion of £1100 as his line of £100 bore to the total value of goods at risk, viz, £20,000. Sir George Honyman QC in his argument invoked the practice in marine property insurance contending that by parity of reasoning the lightermen were their own insurers to the extent of the full value of the goods in excess of £2000. The argument was rejected for the simple reason that the policy was not a policy on goods and not therefore “strictly a marine insurance”. It was, rather, a contract by which the insurers agreed to indemnify the lightermen against any liability which they might incur as carriers to the owners of the goods entrusted to them. The defendant underwriter’s liability was therefore £55, being his 5% share of the limit of indemnity offered, £2000, applied to the loss, £1100. Lush J said this, at pages 82-82:-

“I am of the same opinion. This is an exceptional policy, and we have only to construe the language used; and when I look at the position of the plaintiffs and find that they are carriers upon the river, I cannot doubt the intention of parties. The object of the plaintiffs was to secure an indemnity against any loss in whole or in part which they might sustain as carriers, and it is not a mere policy on goods. A case may be supposed in which the goods have perished and yet the underwriters might not be liable on this policy. The subject-matter insured against is the liability which the plaintiffs would sustain in respect of the goods by reason of their accepting them as carriers. I cannot interpret the words of the policy in any other sense than as importing that the under-writers undertook to be responsible to the extent of their subscriptions for all the losses, which the plaintiffs might sustain in respect of those goods, and for which they would be liable to the owners. The language has that meaning, and I do not entertain a doubt that that is what the parties intended. It is not an ordinary marine policy, but a policy of a mixed nature, the object of which was to secure to the plaintiffs an indemnity to the extent of the sum subscribed for, for any loss during the year which they might sustain by reason of their being responsible as carriers for the loss of the goods.”

Mellor and Hannen JJ agreed with him.

In the light of the origin of the practice as to averaging or apportionment in marine property insurance, and the failure of the attempt to extend that concept to liability insurance in Joyce v Kennard, it is clear, in my judgment, as Mr Leggatt submitted, that Walton J’s remarks in Cunard v Marten went no further than to enunciate the classic approach to underinsurance in marine property insurance. Even within the confines of marine property insurance, there is no warrant for bringing into the averaging calculation interests which are not the subject of the insurance under which the claim is made. Averaging is concerned only with the relationship between the value of property insured, whether their actual value or their conventional value, as in a valued policy, and the proportion of the risk insured assumed by individual subscribers to the insurance. Rix J was, I think, with great respect, in error in thinking that Walton J’s approach was equally applicable to the apportionment of expenditure directed to saving both an insured and an uninsured vessel. If the expenditure was wholly directed towards saving insured vessel A, then, subject to any question of underinsurance of that vessel, there is no support in the principle enunciated by Walton J for apportionment of the expenditure if, fortuitously, it happens also to have been wholly directed towards saving uninsured vessel B. If, as is perhaps a more likely example, vessel A and vessel B are both insured but with different insurers, the underwriters of vessel A may of course be entitled to pursue the underwriters of vessel B for a contribution, but that does not detract from the entitlement of the owners of vessel A to recover the expenditure in full.

The judgment of Phillips LJ in Royal Boskalis likewise identifies the rationale of the apportionment principle as being underinsurance of the property insured and furthermore identifies the impossibility of apportionment where the uninsured interest sought to be preserved is incapable of precise valuation. Not only did the court in Joyce v Kennard reject the proposed extension of the principle from liability insurance but s.74 of the Marine Insurance Act itself prescribes that the measure of indemnity in liability insurance is that the claim under the policy is paid for the full amount of the liability up to or subject to any policy limit or restriction. Thus s.74 provides:-

“Where the assured has effected an insurance in express terms against any liability to a third party, the measure of indemnity, subject to any express provision in the policy, is the amount paid or payable by him to such third party in respect of such liability.”

It may be that aviation property losses are traditionally adjusted in the same manner as marine property losses, but there is no finding to that effect in the Kuwait Airways case, and thus no immediately discernible rationale for the extension of the rule from marine property insurance to aviation property insurance. Since the House of Lords decided that sue and labour expenses were subject to the same ground limit as aircraft and spares, the question of apportionment of sue and labour expense did not ultimately arise and Lord Hobhouse, who delivered the leading speech, expressed no view about it. I too need say nothing further save that the observations of Rix J and of Staughton LJ to which I have already referred lend no support to the extension of the principle to liability insurance.

Finally, I agree with Mr Leggatt that not only is it clear from the authorities that the rationale underlying the principle of apportionment has no place in liability insurance, but also that it would in fact be irrational and unprincipled to attempt to introduce it. In liability insurance, the assured recovers his loss up to the policy limit. Thus where a liability policy has a limit of £100M and liabilities are incurred of £200M, it is not open to insurers to say that the insured is “his own insurer” in excess of £100M with the consequence that insurers’ liability in respect of the £200M loss is limited to £50M. The insured recovers the full £100M. As Mr Leggatt submitted, if that is the position in respect of the recovery of losses, then it is only logical that the same principle should apply where costs are incurred in order to avoid a loss. In the property case the logic is that if the loss, if recoverable, is going to be apportioned, on account of underinsurance, so too should the costs of suing and labouring to prevent that loss. The equal and opposite logic is, submits Mr Leggatt, that if under a liability insurance the loss if it eventuates will not fall to be apportioned, then neither should there be apportioned the costs of sue and labour expenses incurred in order to avoid or minimise such liability.

As Mr Leggatt also pointed out, the concept of underinsurance makes no sense in the context of liability insurance where the extent of the liabilities to be incurred is unknown when the policy is agreed. An insured may prove to be inadequately insured in the light of the liabilities to which he is potentially exposed and which ultimately eventuate, but that does not come about as a result of deliberately underestimating the extent of any known liability. By definition the liability will not have been incurred at the inception of the insurance.

Finally, as this case itself demonstrates, extension of the principle of apportionment to the liability context would be productive of great uncertainty because the liability sought to be avoided would very often if not usually be difficult if not impossible to quantify. Here the interest sought to be protected, at any rate in part, was the brand, or, to put it another way, the loss sought to be avoided was brand damage. Here SLAL put an estimate on that damage but no-one could claim that that was an accurate figure and one can readily envisage circumstances in which it would be simply impossible to produce any sort of an estimate as to the likely value of damage to reputation and commercial interests. Furthermore, it could not be appropriate to attribute a value ex post facto. It is necessary to judge whether the proposed mitigation action is reasonable from the perspective of the insured at the time he is called upon to make the relevant decision. In such circumstances the insured may very plausibly say that he had no idea what the extent of the loss might be save only that it would be very large, in which event apportionment is impossible for the very reason given by Phillips LJ in Royal Boskalis.

The foregoing is I hope sufficient to demonstrate why apportionment is here unavailable whether in relation to expenditure directed to the avoidance of brand damage or to expenditure directed to the avoidance of claims from investors in excess of the policy limit of £100M. In relation to the latter however it should be borne in mind that the extent of such claims was likewise unknown, albeit there were parameters by which the level of claims necessarily fell to be bounded, of which perhaps the upper bound of £190M is the more reliable. As I have already mentioned the lower bound figure of £124M was thought by SLAL to be unreliable, or of spurious accuracy, both because not all those with potentially valid claims would actually make a claim and because the cost of defending claims would not be limited to defending claims from those who had received misleading literature. Thus the arithmetic exercise is equally intractable in the case of avoiding claims from investors in the Fund as it is in relation to avoiding brand damage.

In relation to the suggestion that costs incurred in avoiding or reducing the cost of claims in excess of £100M should be apportioned because such claims were “uninsured”, I should record Mr Leggatt’s further argument that the definition of mitigation costs refers to the costs of taking action to avoid claims “of a type which would have been covered under this policy”. I agree with Mr Leggatt that it is not open to the insurers to say of some of the claims that they may not have been covered by the policy because the relevant test is not whether they would have been covered but whether they are of a type which would have been covered. I also agree with Mr Leggatt that the existence of that provision would make it all the more irrational that avoidance of brand damage should lead to apportionment, bearing in mind that brand damage falls completely outside the scope of the policy whereas claims in excess of £100M are at least claims which would, but for the limit, have ranked for indemnity. Given that the necessity to apportion derives from underinsurance, it would seem strange if protection of an interest with which the policy is not concerned should lead to apportionment whereas the avoidance of claims of a type with which the policy is concerned but which fall outside the policy limit does not.

For all these reasons therefore I believe that apportionment is unavailable both because inconsistent with the clear language of the policy and because it is not a principle which has any application in liability insurance.

I can deal much more shortly with the subsidiary ground of appeal which I have identified at paragraph 8 above. The argument is that to the extent that the Cash Injection bestowed a windfall upon investors in the Fund who would have had no claim because no misrepresentation had been made to them as to the manner in which the Fund was invested, so the Cash Injection is pro tanto irrecoverable as not reasonably and necessarily incurred in taking action to avoid or reduce third party claims. 36% of the Cash Injection, it is said, was paid to customers who were not perceived as having valid claims.

I am afraid that I regard this argument as quite hopeless, and to be fair to Mr Schaff he did not, once pressed, pursue it without regard to the realities. The judge found that the only alternative to Option 1 was Option 2. Once Option 1 was rejected, the only way in which SLAL could take action to avoid third party claims in a manner thought likely to be less expensive than pursuing Option 1 was by reinstatement of the Fund value. That necessarily involved paying the entire 4.8% lost value into the Fund, whether directly or indirectly by making good the loss to those who had disinvested since the one-off reduction. Option 2 could not be performed in any other way. The judge found that the payment was both reasonably and necessarily incurred in taking action to avoid or reduce third party claims of a type which would have been covered under the policy. The payment was indivisible. The fact that it produced a windfall to some investors seems to me quite irrelevant to its recoverability. The payment was indivisible and could not have been made in a reduced amount if it was to achieve its purpose. There is no basis upon which insurers can resist indemnifying SLAL in respect of the entirety of the Cash Injection.

I would dismiss this appeal.

I agree.

I also agree.