Motto & Ors v Trafigura Ltd & Anor

ATE Premium

Should Leigh Day & Co have obtained a staged policy, to reflect the change in risk over the course of the litigation?

What is the reasonable and proportionate premium to have paid for ATE insurance for the claims?

Mr Bacon for the Defendants raised the further question: “Is the premium claimed a premium within the meaning of s29 of the Access to Justice Act 1999?” I answer that question at the end of this judgment.

First Assist is an insurance intermediary which acts under a Managing General Agency Agreement (MGA) with Great Lakes Re-Insurance UK Plc, which is part of the Munich Re Group, which provides the insurance capacity for First Assist’s business.

Great Lakes was the insurer of the risk and was liable to meet claims. Premium for all insurance business is collected by First Assist on Great Lakes’ behalf, and banked into Great Lakes’ client account. Claims and intermediary commissions are charged against this account. Great Lakes were entitled to deduct the agreed fee for providing the insurance capacity and their costs in providing the facility. The balance belongs to First Assist, and is its dominant source of income with which to run its business. The premium sought by the Claimants in this case is £9,677,554.

The policy states that it is administered by First Assist Insurance Services Ltd, and is underwritten by Great Lakes Reinsurance (UK) Plc, the insurer. The preamble states:

“The Policy, the Schedule and any Endorsement are to be read as one document. Any word or expression used with a specific meaning in any one of them has the same meaning wherever it appears.”

I set out below relevant parts of the policy:

“The Insurer will provide the insurance described in this Policy in consideration of the Insured’s promise to pay the Premium.

insurer

Great Lakes Reinsurance (UK) Plc

we/us/our

First Assist Insurance Services Ltd which administers the insurance on the insurer’s behalf and to which any correspondence must be addressed …

…

insured

The individual, business or other organisation named as the Insured on the Schedule.

…

solicitor

The Solicitor representing the Insured under the Conditional Fee Agreement and named on the Schedule.

limit of indemnity

The Insurer’s liability in respect of Expenses shall not exceed the amount stated in the Schedule.

territorial limits

England & Wales.

court

A court, tribunal or other competent authority.

the opponent

The party or parties from whom the insured is claiming money or damages in the Legal Proceedings.

adverse costs

The net costs of the Opponent in the Legal Proceedings to the extent that the Insured is legally liable to discharge them, after taking account of any costs awarded against the Opponent or agreed to be paid by the Opponent. No cover is provided in respect of any success fee to which the Opponent or Opponent’s solicitor or Opponent’s barrister may be entitled.

normal fees

The agreed or assessed professional fees of the Solicitor acting for the Insured under the Conditional Fee Agreement and which form the basis for the calculation of the Success Fee.

expenses

Expenses and other disbursements paid by the Solicitor to other parties which are reasonably and properly incurred by the Solicitor

a)

in connection with the Legal Proceedings

b)

in appealing or resisting an appeal against the judgment of a Court in connection with the Legal Proceedings provided Our prior written consent has been obtained to the appeal.

The Insurer shall not be liable for Counsel’s fees where Counsel is retained to act in the Legal Proceedings under a conditional fee agreement.

legal proceedings

The legal action described in the Proposal and brought by the Insured to pursue money or damages in compensation and which is the subject of the Conditional Fee Agreement.

period of insurance

Cover commences at the later of the inception date shown on the Schedule and the date of signing of the Conditional Fee Agreement.

Cover ceases when

a)

the Legal Proceedings are concluded by a judgment of the Court of first instance or following an appeal to which We have given Our prior written consent or

b)

the Legal Proceedings are settled or discontinued with Our prior written consent or

c)

the Conditional Fee Agreement is terminated

d)

the Legal Proceedings are transferred outside the Territorial Limits

whichever is the earliest.

premium

The amount calculated using the Premium Rate as defined below multiplied by the Normal Fees of the Solicitor.

The Premium becomes payable when the outcome of the Legal Proceedings is a Success.

Where the Insured receives an offer to settle or a payment into Court which equals or exceeds the definition of Success the Insured may at that time immediately pay the Premium based on the Normal Fees at that date. The Insured may elect by written notice to Us to continue with the action and defer payment of the Premium to the conclusion of the Legal Proceedings. The Premium will then be based on the Normal Fees at the conclusion of the Legal Proceedings. If, having made this election, the Insured ultimately fails to achieve a Successful outcome to the Legal Proceedings the Premium payable at the date when the Insured made the election shall remain payable.

premium rate

The Premium Rate is calculated by

adding the costs of the Opponent and the Expenses,

dividing the total of these by the Normal Fees of the Solicitor

multiplying the resultant number by the Multiplier shown in the Schedule.

The figures used for the calculation of the Premium Rate and Premium in respect of Normal Fees and Expenses shall be the actual values assessed as due in the case or agreed as due between the parties to the case. The figures used for the calculation of the Premium Rate and Premium in respect of Opponent’s costs shall be the actual costs of the Opponent (including the Opponent’s Solicitor’s costs and the Opponent’s Expenses) as certified by the Opponent’s Solicitor.

In the event that the Opponent refuses to provide Us with the value of the Opponent’s costs, then for the purposes of the calculation of Premium, the value of Opponent’s costs will be deemed to be equal to the sum of the Normal Fees and Expenses as defined herein.

success fee

The amount specified in the Conditional Fee Agreement as the percentage uplift to the Solicitor’s Normal Fees payable on the successful conclusion of the Legal Proceedings.

…

success/successful

Legal Proceedings will be deemed successful if the Insured is offered or obtains at any time a net entitlement to money and/or damages and/or costs which, taking into account any counter - or cross-claim in the Legal Proceedings, equals or exceeds the definition of Success as shown in the Schedule.

…

conditional fee agreement

The Conditional Fee Agreement relating to the Legal Proceedings and entered into by the Insured and the Solicitor pursuant to Section 58 of the Courts and Legal Services Act 1990

cover

Where the outcome of the Legal Proceedings is not a Success the Insurer will, subject to the Limit of Indemnity, indemnify the Insured in respect of

a)

Adverse Costs

provided that

1.

the Court makes an award of Adverse Costs against the Insured or

2.

the Legal Proceedings are settled or discontinued with the prior written agreement of the Insured, the Solicitor and Us

3.

the Insurer shall not be liable to pay any Adverse Costs until the Legal Proceedings are finally concluded.

b)

Expenses

provided that

1.

the Insurer shall only be liable for Expenses to the extent that the Insured is not entitled to recover them from the Opponent or any other party

2.

the Insurer shall not be liable to pay any Expenses until the Legal Proceedings are finally concluded.

…

conditions

…

4.

Minimising Claims or Legal Proceedings

The Insured must take all reasonable measures to comply with the overriding objectives of the Civil Procedure Rules and to minimise the cost of Legal Proceedings.

…”

The Schedule to the policy sets out the policy number; the name of the administrator: First Assist Insurance Services Limited; the insureds’ solicitor; and the insureds “The Ivory Coast Action Group”; the names of the opponents; the date of the proposal form: 25/01/2007; the limit of indemnity: £250,000; the inception date: 09/02/2007; and definition of success:

“The Claim for Damages is finally decided in favour of the Insured Group whether by a Court Judgment in its favour or where an offer is received which the Insureds’ Solicitor recommends should be accepted or any other offer is accepted.”

The multiplier: 61.923% (plus IPT at the prevailing rate).

“Premium Rate:

(Opponents’ costs + Expenses) divided by Normal Fees of own Solicitor.

All these figures will be the actual values as determined at the conclusion of the case.

This figure is then multiplied by the above Multiplier.”

The premium:

“The premium will be:-

The Premium Rate calculation shown above, multiplied by the Normal Fees of own Solicitor

OR

The sum of £100,000 (+ IPT at the prevailing rate)

whichever shall be the greater.”

The Schedule is dated 4 April 2007. The schedule is dated 4 April 2007 and states:

“This insurance policy is administered by First Assist Insurance Services Ltd and underwritten by Great Lakes Reinsurance (UK) Plc.”

Attached to the schedule is an endorsement as follows:

“IT IS NOTED AND AGREED

1.

The Limit of Indemnity applies to Expenses only and excludes Counsels’ Fees.

2.

No cover is provided for the Insureds’ Solicitor’s Normal Fees

3.

The Premium as defined in the Policy Wording is hereby amended to show that this is based upon the Insureds’ Solicitor’s total base costs in the case as opposed to only those under the Conditional Fee Agreements.

4.

The Insureds are each a Party to a Multi Party Action (the Action) comprising Claimants each of whom is acting under a Conditional Fee Agreement. Each of the Insureds will be a Party to this Insurance subject to:-

(1)

Signing a separate Conditional Fee Agreement

(2)

Agreeing to be liable, in the event of the definition of success being achieved, for an equal share (to be determined when all of those eligible and wishing to join the Action as Claimants have done so) of the Premium save under Paragraph 5 below.

(3)

Agreeing to be liable, in the event of the definition of success not being achieved, for an equal share (to be determined when all of those eligible and wishing to join the Action as a Claimant have done so) of Adverse Costs save under Paragraph 5 below.

5.

In the event that 1 or more of the Insureds discontinues or settles his/her Claim prior to any Hearing or Settlement of the Action and thereby ceases to be a Party to it, those remaining Insureds in the Action to its conclusion by way of a Hearing or an agreed Settlement will be equally liable (amongst themselves) for the Premium where the definition of success has been achieved and equally liable (amongst themselves) for Adverse Costs where the definition of success has not been achieved.”

Mr Peter Smith of First Assist Insurance Services Ltd initially made two witness statements, the first dated 27 October 2010 in which he explained how the fee for the ATE premium was arrived at, and his second dated 6 January 2011, in which he dealt with various issues raised by Mr Nurney in his eighth witness statement, and with the report dated 10 December 2010 from the Defendants’ expert, Mr Trevor Clegg.

Mr Smith sets out his career in legal expenses insurance, starting in 1992 when he was appointed an underwriting claims manager. He also sets out the origins of the Pursuit Insurance Scheme, and states that he has been responsible for the development of First Assist ATE business at all times from its origins in 1999. The purpose of this policy was to support the use of CFAs in areas outside the mass personal injury market. The premium for such a policy would have to be much larger than in run of the mill personal injury cases. First Assist devised a policy with a deferred premium, which was conditional upon success. Thus, those who won their cases would have the means to pay the premium, those who lost would not need to do so. Payment of the premium in won cases would be deferred until the case was won. The policy was intended to provide unlimited adverse costs cover. The Pursuit Scheme was launched in March 1999.

Mr Smith explains:

“21.

Every insurance policy is necessarily based upon a set of underwriting assumptions. These assumptions are the reference points for the rating mechanism. The essential rating components are the likelihood of a claim (risk factor) and the likely cost of such a claim (exposure factor). …

22.

The rationale we applied was that, because the legal fees and disbursements which are the subject matter (exposure factor) of the insurance inevitably increase the further the litigation progresses (and the exposure to the insurer therefore increases likewise), it was necessary for the premium to reflect this as well. It followed that the premium needed to be calculated in proportion to the level of insured risk at any given time. In effect, it needed to be infinitely staged with the premium for each risk reflecting the merits of the specific case (risk factor) and the costs of the case (exposure factor).

23.

We then had to consider how to quantify the value of the premium to be charged at the conclusion of the case. This clearly had to be based upon those costs and disbursements that we were able to know about at the time the risk was presented to us.

24.

It would therefore be ideal to calculate the value of the premium by reference to the claimants’ disbursements, defendants’ costs and defendants’ disbursements in exact equivalence to what is being insured.”

Because it was not possible in 1999 to know accurate details of the Defendants’ costs and disbursements, the company took the view that the costs of each party are often similar. They, therefore, decided that the exposure factor of the premium should be calculated by reference to own solicitors’ costs, as with a success fee under a CFA. This produced a sliding scale premium calculated by reference to own solicitors’ fees. Each case was rated on the estimated costs of that particular case. In order to calculate how the policy should be rated in terms of risk factor, ie, the multiplier to be applied to the solicitors’ fees:

“… it was necessary to form a view as to:

(i)

the likelihood of the insured’s case succeeding in percentage terms; and

(ii)

if the policy was called upon, how much would have to be paid out, in order to work out the break even position, or risk premium for the scheme, where the costs of claims paid exactly matches the value of the risk premium income (excluding expenses and commission) received.

Given the product was an entirely new concept, the absence of previous underwriting data, and the considerable volatility in rates in the ATE market, we resolved to apply basic principles to arrive at an appropriate rating model.”

This method of calculating the premium was challenged in 2004 by a number of defendant liability insurers, leading to the RSA Pursuit Test Cases in 2005, which was argued before me. I found (Order dated 27 May 2005) that the contract of insurance was not void for uncertainty, nor was the arrangement between the client and/or the solicitors and the insurer unlawful on the grounds of champerty. In my judgment I found certain flaws in the way in which the premium had been calculated in the Test Cases, largely because the estimated costs which formed the basis of the premium calculation were extremely inaccurate. As a result of this, the policy wording was amended by First Assist, thus allowing the premium rate to be applied directly to the actual exposure faced by the company, ie, the actual costs of the opponent at the time the case was settled in favour of the insured (plus the own disbursements to the extent that they were insured). It is this methodology which has been adopted in the present case.

Mr Smith sets out the formula used for calculating the premium:

“● an exposure multiplicand is calculated, being the sum of own disbursements and opponent’s costs (subject to any limit of indemnity on the former);

● a risk multiplier is applied to the exposure multiplicand;

● a basic risk multiplier is ascertained in the same way as a success fee is ascertained, namely by dividing the chance of losing by the chance of winning;

● this basic figure therefore represents the pure “burning cost” of the insurance (that is the “break even” cost disregarding the need for the underwriter to fund its business overheads and make a profit);

● the basic figure is then adjusted to include an allowance for overheads, marketing and brokerage costs and profit;

● this provides the overall risk multiplier;

● the overall risk multiplier is applied to the exposure multiplicand, to derive the net premium;

● insurance premium tax is then applied to produce the overall figure.”

Mr Smith says that this method of calculating the risk multiplier was approved by the Court of Appeal in Rogers v Merthyr Tydfil CBC [2006] EWCA Civ 1132; [2007] 1 WLR 808.

With regard to underwriting the Abidjan case, Mr Smith states that First Assist were first approached in November 2006. The matter was handled initially by Mike Fallen, who left the company at an early stage, and the conduct was taken over by Ian Coleman, with Mr Smith acting as supervisor. Leigh Day had supplied papers setting out the background to the case, but “unusually for such a case” there was no advice from counsel.

Ian Coleman asked for additional information, which was provided by Leigh Day towards the end of January 2007. One of the documents supplied to First Assist was Leigh Day’s risk assessment, which I have dealt with in some detail under Issue 13 Success Fee. Mr Smith states:

“42.

This set out succinctly the risk factors which they had identified, all of which appeared to correspond with the picture we had formed of the risk from the other material provided to us.”

Ian Coleman prepared a risk assessment, but did not include an assessment of the merits in percentage terms. He did, however, express the view that the case was “in its early stages, but appears to be a good case”. Mr Smith then considered the file, and noted his view that the merits were “65% … on the facts as known”. This formed the basis of First Assist’s quotation. Terms were put to Leigh Day in February 2007, and eventually agreed as follows (paragraph 48):

“Cover for the group with the right for new claimants to be added to the group.

The premium was the joint responsibility of each and every member of the group, such that the full premium was recoverable in the event of success.

Unlimited cover for opponents’ costs (but a limit of £250,000 for own disbursements).

Lien on damages to the extent necessary to cover the premium.

Definition of success to be finalised; the premium rate was to be 62% (actually 61.923%).”

The premium rate was agreed, and cover confirmed on 9 February 2007. The definition of success was the subject of some discussion, but following agreement the policy documentation was prepared and issued at the start of April 2007.

In arriving at the figure of 65%, the factors taken into account by Mr Smith and his colleagues were (paragraph 53):

“The complexity of group actions;

whether the case would be run under a GLO and the range of possible costs orders which could be made if some claimants were successful and others not, or discontinued their actions;

the need for LD to have the ability to add more claimants (which could in some circumstances dilute the prospects of the whole group succeeding);

the evidential difficulties given the location of the incident and of the claimants, together with foreseeable issues around language and identification of individual claimants;

issues around liability and causation coupled with the known denial of liability by Trafigura, and an assessment of Trafigura’s ability to fund a robust defence of the case;

the possibility of Trafigura seeking and obtaining a security for costs order, notwithstanding the existence of the ATE policy;

the possibility of Tommy’s involvement “breaking the liability chain”;

the policy coverage, which gave the insurer no unilateral right to come off cover if merits fell;

the difficulty in obtaining payment of the premium by the claimants and the risks that the full premium may not be recovered from either the claimants or the defendant.”

Mr Smith points out that the company was sensitive to the likelihood that the assessment of merits could be challenged, and they were therefore keen to adopt a reasonable and conservative figure. The formula used to produce the premium rate was as follows:

“57.

Merits of 65% gives a risk of losing of 35%.

35/65 x 100 = 53.8% risk premium rate.

Loading of 15% for admin and profit (53.8 x 1.15) = 61.923% premium rate per the policy schedule.”

First Assist did make efforts to obtain co-insurance, ultimately without success. Mr Smith is not aware of any ATE insurer which had the capacity to write a risk of £14 million in 2007.

The actual premium claimed has been calculated by applying the premium rate to the sum of Trafigura’s costs and the Claimants’ disbursements. Trafigura’s costs were said to total £14,634,160 exclusive of VAT.

At the time Leigh Day arranged the after the event cover proceedings had been issued, and the Claimants were therefore exposed to the risk of being liable for the Defendants’ costs if the case failed. They also had a liability for their own disbursements. It was clear to First Assist that the Claimants had an insurable interest. Had there been no after the event insurance in place, the Defendants could have sought to stifle the claim by making a security for costs application.

With regard to the suggestion that the premium rate should have been adjusted as the merits of the case changed, Mr Smith points out that the premium rate for the policy is a contract term agreed at the outset. This gives the insured certainty, and passes the risk of fluctuations in the merits to the insurer. It would not be acceptable to a policy holder if the rate increased when the merits deteriorated, nor would it be acceptable to the insurer were it to reduce if the merits improved.

In his second witness statement Mr Smith asserts that the ATE premium rate was set in accordance with the judgment in the RSA Pursuit Test Cases, and was based on the merits of the case assessed by the insurer at the time the application for insurance was made. The contractually agreed premium rate was applied to the costs exposure which the insurer faced when the case failed. He states the reason the premium was at the level claimed is due to the level of costs incurred by the Defendants.

As to the timing of the insurance, he points out that it generally becomes less available and/or more expensive the closer to trial an application for insurance is made.

At the time the application by Leigh Day was made First Assist believed that it had sufficient understanding of the case to offer insurance. Mr Smith states (paragraph 10):

“There was a substantial body of information available, including United Nations Report confirming the existence of toxic waste, and the fact that some 100,000 inhabitants of Abidjan had sought medical attention at the time immediately following the dumping of the material. Despite Trafigura’s denial of liability, no other plausible explanation for the events has been provided.”

At the time the insurance was incepted First Assist had no knowledge of the payment by Trafigura to the Government of the Ivory Coast. Mr Smith is unable to say what effect knowledge of that payment would have had on the risk assessment, since it is still not clear to him the basis on which the funds were paid. He concludes that the payment did not diminish the risk to the Claimants.

With regard to Mr Nurney’s suggestion that First Assist should re-assess the merits of the case as it progressed, Mr Smith says that this “is a seriously mistaken view”:

“15.

The terms are agreed before cover incepts and are then unchanged unless altered by mutual agreement. If the merits were to deteriorate the insurer cannot increase the premium. Likewise it will not reduce the premium if merits improve. Nor is this approach peculiar to ATE insurance. Most insurance, whether it covers lives, cars, ships or buildings, involves a one-off, ab initio, assessment of risk.”

He goes on to suggest that the Pursuit Model is better than any other ATE premium model on the market in that early settlement has a direct effect on the premium, as it is linked to the level of the Defendants’ costs.

With regard to the limit of cover, Mr Smith states that he personally drafted the Pursuit Policy wording, and that the reality is that the policy was unlimited in respect of adverse costs.

At the outset Mr Smith had sought co-insurance or re-insurance, and had approached QBE, an established and respected ATE provider, to see if they would be interested in sharing the risk. This was ultimately declined. In early 2008, with the costs escalating and no offer of settlement, Mr Smith made further efforts to share the risk using a specialist broker: Universal Legal Protection. However, no other insurer or re-insurer was willing to participate in the risk on the terms of First Assist’s policy.

Dealing with Mr Clegg’s report, and the suggestion that the exact form of wording was under negotiation until October 2007, Mr Smith states (paragraph 35):

“The fact is that the wording was issued in early April 2007 in accordance with the basis of cover agreed in early February. After the policy documentation was issued two typographical errors on the schedule were corrected in May 2008. It is, therefore, incorrect to suggest that the wording was not finalised between April and October.”

Mr Smith takes issue with Mr Clegg’s assertion that the limit of indemnity of the policy was £2 million. The cover was unlimited in respect of opponents’ costs, and only Leigh Day’s disbursements were limited by agreement to £250,000. It was only in January 2008, after this particular policy had been issued, that First Assist introduced a £2 million long-stop to the standard Pursuit cover.

With regard to Mr Clegg’s suggestion of a discount or profit commission, Mr Smith says this would eat into the risk premium, and provide an unacceptable premium for the insurer.

Mr Myrtle is the managing director of Universal Legal Protection Ltd, and is a broker who has worked exclusively in insurance for over 30 years. He has dealt directly with legal expenses insurance for some 20 years, and after the event insurance for 15 years. He has been asked by Leigh Day to comment on and answer the following questions:

Dealing with question (i), Mr Myrtle deals with the set up of the ATE market, and the experience of the various insurers involved. Following the collapse of The Accident Group, Lloyds syndicates and NIG withdrew from new business and ran off their existing business at very substantial underwriting losses. He suggests that other insurers who avoided case management companies such as Claims Direct and TAG quickly learned how to underwrite successfully, and in his opinion ATE insurers have been making reasonable underwriting profits since the middle of the past decade. He suggests that “this is now a competitive and saturated market”.

The number of ATE insurers prepared to consider underwriting one-off litigation cases has always been much smaller than for run of the mill personal injury litigation. The bigger the limit of indemnity required, the smaller the number of providers becomes. Over the past few years certain new insurers have joined the market as retail insurers, but others have fallen away, either due to lack of spread of risk, or adverse underwriting results. First Assist was the first to enter the market, and is seen as being the market leader.

When Leigh Day first approached First Assist in 2006 they sought cover for their own disbursements up to £250,000, and estimated the exposure to adverse costs at £750,000 because they hoped for an early settlement. Mr Myrtle suggests that a limit of indemnity of £750,000 “would have been an exceptionally large limit”, which would overbalance any ATE insurer’s book of business. He concludes that:

“I do not consider it likely that any other ATE insurer could or would have given unlimited cover in this case … I doubt very much that another insurer’s terms would have been better than the terms offered by First Assist.”

Question (ii) asked Mr Myrtle to consider the levels of indemnity alternative providers might have provided. He does not think that in late 2006 any of the other ATE insurers would have been prepared to provide the unlimited cover given by First Assist. Had the limit of indemnity remained at £750,000, it would have been of interest to a small number of ATE insurers. He suggests other providers who might have been approached included LAMP. LAMP did not reply to Leigh Day’s approach, and Mr Myrtle very much doubts that any of the insurers could have offered that amount of cover in reality.

He goes on to suggest that even if another insurer had been minded to provide the cover, it would probably have taken the view that their limit of indemnity could prove inadequate, and that the prospect of obtaining any sort of top-up would have been problematic. The fear would be that Leigh Day might be forced to discontinue the case through lack of funds, leaving the ATE insurer facing a huge bill, even if the underlying claim was still reasonably good.

If Leigh Day had sought a higher indemnity limit, Mr Myrtle suggests that QBE and Brit might have been prepared to offer a limit of £2 million, and Templeton £1 million. It might have been possible to persuade QBE and Templeton to co-insure the risk, and increase the level of cover to an absolute maximum of £3 or £4 million. They would have insisted on their own policy wording terms and conditions, so that the solicitors could have ended up with two very different contracts, with all the difficulties that that might have entailed. The most obvious problem being an offer of settlement, which the solicitor and one insurer wish to reject, but the other insurer wishes to accept. Mr Myrtle stated that he has tried, from time to time during the last ten years, to persuade insurers to subscribe to a risk placement by way of co-insurance using a single policy wording, but has never succeeded. He tried to do this in 2008 when First Assist asked him to investigate co-insurance or re-insurance, he was unable to do so.

The Claimants needed ATE insurance in order to avoid the risk of the Defendants seeking an order for security for costs. The ATE policy had to be with a company which was clearly good for the money.

Mr Myrtle does not think that there would have been a market for open-ended liability in 2006, any insurer would have been extremely nervous as to the risks, and in his view no insurer, other than First Assist, would have offered a firm quotation.

With regard to question (iii) regarding the risk factor applied, Mr Myrtle is not qualified to answer the question directly, but expresses surprise that First Assist did not apply a higher risk factor.

He suggests that an administration fee of 15% is normal across the spectrum of insurances.

Question (iv) deals with Mr Clegg’s suggestion that the Claimants should have cancelled and re-negotiated the policy once it was known that the Defendants had paid money to the Ivorian Government. Mr Myrtle recalls discussing with Mr Smith what was known about the payment in 2008, when he was evaluating how best to go about seeking co-insurance for First Assist. One view was that the payment might have been intended by Trafigura to draw a line under the affair, and might have diminished rather than have increased the prospects of the Defendants settling with the Claimants.

Had the Claimants cancelled the policy, and then asked First Assist for a new policy, the Claimants would have faced the very real risk that First Assist would either refuse to issue a replacement, or that the replacement would have been on less advantageous terms for the Claimants. Had he been approached by the Claimants to advise on whether the policy should be cancelled and re-negotiated, he would have told Leigh Day that they would be taking a massive risk.

Question (v) dealt with the possibility of a staged premium. Mr Myrtle suggests that the First Assist premium was staged, in that had the litigation settled early, the premium would have been significantly less. The First Assist premium increases as the exposure to First Assist increases.

As to question (vi), and the alternatives that were open to the Claimants in relation to the ATE market, Mr Myrtle is of the view that even if a different insurer had been prepared to quote (which he doubts), it seems highly unlikely that the policy on offer would have proved satisfactory to Leigh Day compared to First Assist’s Pursuit product.

In answer to question (vii), how likely would it have been that Leigh Day would have obtained a cheaper premium using the services of a broker, Mr Myrtle states that the answer is: very unlikely indeed. Additional costs would have been incurred of between 10% and 15% of the gross premium. Using a broker would only have saved the Claimants money in terms of a lower premium if a different insurer had offered better terms than First Assist, but, as he has already explained, this was extremely unlikely to come about.

Mr Myrtle explains his difficulties in trying to arrange co-insurance and re-insurance, and agrees with Mr Clegg that in the worldwide insurance market there are those who will insure unusual risks, but such insurers would demand the premium at the outset, which was quite simply not a possibility in this case. Such providers would never agree to the premium being refunded should the case be lost.

The Defendants’ expert, Mr Clegg, raised a number of questions on 13 January 2011 prior to a meeting of the experts. At that meeting the experts agreed that Leigh Day could have sought alternative quotes in late 2006, or early 2007, and that there were markets other than First Assist that would have been likely to offer terms, depending on the limit of indemnity sought or required. Mr Myrtle states that those terms would not have been on the same premium computation basis, or on an unlimited indemnity for adverse costs basis.

The experts agreed that, at the time of the inception of the policy in February 2007, the anticipated own costs and disbursements submitted by Leigh Day & Co were £750,000.

The experts could not agree on whether Leigh Day should have cancelled the policy, and renegotiated it once the payment by Trafigura to the Ivorian Government became known, nor could they agree whether Leigh Day could, or should, have bought an alternative policy from another insurer that could have been supplemented by an excess policy.

It was Mr Myrtle’s view that a broker, working for a fee rather than a commission, would almost certainly have made it clear to Leigh Day & Co the value of the unlimited policy offered by First Assist.

At the first hearing before me, Mr Bacon raised the issue that the premium was not in fact a premium within the meaning of Section 29 of the Access to Justice Act 1999. Mr Smith accordingly made a third witness statement dated 26 January 2011. He states that the relationship at the centre of the present costs claim is that between the Claimants, who were the insured under the policy and Great Lakes which was the insurer. The premium is due to Great Lakes and First Assist are simply fulfilling their contractual obligations in seeking to collect the premium which is due. He continues:

“6.

In the event of success, as defined, the insured contracted to pay the premium to Great Lakes. In the event the outcome of the case was not a success, the insurer contracted to indemnify the insured in respect of their liability to meet adverse costs and expenses as defined.

7.

First Assist was not a party to this contract, has no entitlement to the premium and no liability to indemnify the insured.

8.

First Assist was, however, an agent of Great Lakes under the policy and was authorised to administer the policy on behalf of Great Lakes. These functions included the risk assessment process, the provision of a quotation, agreement of the contract terms and all other policy related tasks until the policy concluded, including the collection of any premium or the payment of any claim …”

He comments that this type of MGA has been common in the UK insurance market for many years.

Mr Smith sets out the way in which the business was transferred from Royal Sun Alliance to First Assist. Consideration was given to acquiring an insurance licence, but eventually it was decided the most effective solution was to obtain the right to use another insurer’s capital and licence. He goes on to explain:

“15.

We entered into an agreement with Munich Re that met our needs. For a period, RSA permitted us to continue to use their licence and their name. After this period, all the insurance business was transferred by a court process known as a Part VII transfer, to Great Lakes …

16.

At all times the Insurance Regulator, the Financial Services Authority, was fully involved by RSA, Munich Re, Great Lakes and First Assist …”

The arrangements now in force have not changed materially since the Part VII transfer.

Mr Myrtle had been unable to deal with the questions raised by Mr Clegg. Eight questions were asked, and Mr Smith supplied answers to each. It is only necessary to refer to two questions for the purpose of this decision. In answer to the first question, relating to the status of First Assist, Mr Smith explained:

“The heritage and culture of the business was therefore one of insurer. However, First Assist determined not to become an insurer but to “outsource” the risk carrying function to Great Lakes. This arrangement required Great Lakes to provide insurance capacity in like manner to that previously provided by RSA; First Assist remunerated Great Lakes for providing the capacity.

It was important for First Assist to arrange insurance capacity with a “blue chip” insurer, as First Assist’s reputation would be affected by the quality of the security it offered to its customers and corporate clients. Great Lakes and Munich Re fulfilled this requirement very well.”

The second question queried the necessity for a bank guarantee to protect First Assist’s balance sheet. To this, Mr Smith replied:

“First Assist was not the insurer of the risk; Great Lakes was. It is Great Lakes which is liable to meet claims.

The bank guarantee was arranged because a £14 million claim on the insurance account would have affected First Assist’s revenue.

In essence, the model works as follows. Premium for all its insurance businesses is collected by First Assist on Great Lakes behalf and banked into a Great Lakes client account. Claims and intermediaries’ commission (and a variety of other industry standard costs such as IPT and the FSCS levy) are charged against this account. Great Lakes are entitled to deduct the agreed fee for providing the insurance capacity and their costs in providing the facility. The balance belongs to First Assist and is its dominant source of income with which to run its business.

Should the insurance result be negative (ie, claims and other expenses exceed premiums), then there is no income for First Assist.

The bank guarantee was obtained to ensure that First Assist could continue to meet its obligations should its income from the insurance account be disrupted by a loss on this unusually large case.”

At the hearing before me on 27 January 2011 Mr Bacon sought disclosure from the Claimants of the details relating to the ATE insurance policy. Having heard argument I directed disclosure limited to the point raised by Mr Bacon, namely: was the £9 million truly premium? The disclosure was to be to the Defendants’ legal advisers only and their expert. The Defendants undertaking that the information would not be used other than in this litigation.

In his witness statement dated 7 February 2011 Mr Smith explains that the relationship between Great Lakes Reinsurance (UK) Plc and First Assist is governed by two agreements, a Managing General Agency Agreement (MGA) and a Remuneration Agreement, both of which were disclosed in full, subject only to redaction of certain commercially sensitive information. Also disclosed were all the Change Control Notices issued since inception of the MGA. Taken together they provide a complete description of the relationship between Great Lakes and First Assist. He explains in relation to the MGA:

“9.

… Subsequent clauses provide considerable detail as to the respective roles and responsibilities of the parties, the primary features of which are that it is the role of First Assist to be out in the market obtaining business, liaising and entering agreements with those requiring insurance, monitoring the operation of the issued policies, keeping Great Lakes informed and it is the role of Great Lakes to provide the actual cover as the insurer. The schedules provide evidence, amongst other matters, as to the nature of this work which First Assist is required to perform on behalf of Great Lakes.

10.

MGAs are a well established commercial model in the UK insurance market. Indeed a Managing General Agent Association is currently being launched to represent the 200 MGAs which together administer around 10% of the UK’s general insurance premium income.”

Mr Smith confirms that the terms of the MGA have been fully disclosed to the Financial Services Authority, the Regulatory Body of both insurers and insurance intermediaries.

With regard to the Remuneration Agreement, Mr Smith states:

“23.

This agreement is between First Assist, Great Lakes and the latter’s parent company Munich Re. Munich Re it will be noted, reinsures 100% of Great Lakes’ interest in the MGA. The existence of this reinsurance arrangement affirms that the insurer is Great Lakes (and no other) and that the business to be reinsured is “bona fide” insurance.

24.

…. However, it will be noted that the profit share payable to First Assist depends on the aggregate result of business written by Great Lakes under both the RSA and Great Lakes MGAs.

…

27.

Clauses 2 and 3 are designed to aggregate the profit share calculation under both the RSA and the Great Lakes MGAs, this is to allow a loss under one to be set off against a surplus in the other and ensures Great Lakes do not pay profit share unless there is an aggregate surplus.

28.

The profits withheld account described at Clause 4 is intended to ensure that Great Lakes retains control over sufficient funds due to First Assist to deal with any services which First Assist may be unable to deliver. This clause is making it clear that Great Lakes is the sole party with responsibility for meeting claims payments.

29.

The risk carrying charge described at Clause 5 is a fee related to the cost of providing capital (and at a normal cost built into pricing by any insurer).”

Under the agreement First Assist is required to retain enough flexibility in its financing structure to deal with unbudgeted surges in the loss ratio. A £14 million loss would have had a material impact on First Assist’s aggregate loss ratio. Mr Smith continues:

“32.

It will be clear from the Remuneration Agreement that all premiums on all First Assist business (of which the ATE premiums amount to around 10% of the whole) are take into Great Lakes’ books and are recorded as premium income. The normal costs of an insurer are then deducted (claims, commission, etc). To the extent that there is a surplus after making a series of prudent assessments of the insurer’s liabilities, this surplus is released to First Assist to run its business.

…

35.

First, the premium on any policy contributes to an overall pot, held by the insurer, from which the insurer meets claims and other liabilities. That pot involves several hundred thousand policies. Premiums are being paid in and claims are being paid out all the time. The premium in this case, albeit large, is simply one of those policies. The insurer and we as agent do not calculate the profitability of any individual policy. Indeed the very notion has no rationale within the insurance world. Clearly any policy where there is no payout is a gain and any claim paid out is a loss. What all insurers do is look at the spread of cover. Our role, as agents, is to obtain business that across its range provides an overall profit for the insurer and for us. In these terms it can be seen that there is no sense in selecting a single policy and arguing that its premium is somehow different from those on other policies. It cannot be reasonable to ignore the risk that any policy introduces into the pool of insured risks and to contend that the premium on any specific policy is pure profit if that policy does not have a claim.”

Trevor Clegg, the Defendants’ expert, has worked for 40 years in the insurance industry, and has held several senior positions, including non executive director of a Lloyd’s Managing Agency in charge of Underwriting Peer Review. During his career he has underwritten both liability and property insurance and reinsurance risks, and speciality insurance risk. He is familiar with the principles of ATE insurance, underwriting and claims handling, having carried out due diligence at The Accident Group when asked by brokers to underwrite the risk, an invitation which he declined.

He describes the situation in this case as a unique one-off situation “that has no real precedent in the ATE market which is a very restricted market.”

In his report, dated 10 December 2010, Mr Clegg sets out the questions on which he has been asked to give his expert opinion:

“2.1

…

(a)

whether the underwriting approach utilised by First Assist in this case is typical of the underwriting principles normally applied to ATE insurance?

(b)

whether the method of premium calculation used is the only method that in all the circumstances of this particular case was reasonable, or are there alternative methods of underwriting the risk and calculating the premium?

(c)

whether the method of premium calculation applied by First Assist was the only method that could have been applied in all the circumstances of this particular case?

(d)

whether any insurer would have written the risk on a sliding scale premium basis that would reduce the final premium charged if the chances of success increased after inception of the policy?

(e)

how likely would it have been for Leigh Day & Co to have obtained a reduced premium and/or better terms if an insurance intermediary had been employed?

(f)

was this risk written on a commercially acceptable basis and at the time prior to or subsequent to its acceptance by First Assist could or should the terms have been amended to reduce the premium?”

Mr Clegg expresses surprise that in this case, which was clearly exceptional, First Assist did not carry out a far more detailed analysis of risk, or request counsel’s opinion as to the merits.

In February 2007 First Assist advised Leigh Day that its risk estimate was 65%, and that it intended to arrange co-insurance or re-insurance. So far as Mr Clegg is aware, the company was working on the basis of estimated exposure of £750,000. In his view the premium at that stage was within normal limits. Whilst Mr Smith had asserted that he thought his risk assessment was fairly generous, Mr Clegg regards it as “relatively pessimistic”.

On 2 February 2007 First Assist indicated in an email that the cover would be unlimited:

“but with estimated exposure for adverse costs and own disbursements of £750,000

● we propose a minimum (conditional and deferred) premium of £100,000 (this is because this is a large case which may either settle or run to trial);

● the insurance will be arranged on a co-insurance basis.”

On that basis Mr Clegg says that the expected premium would have been £465,000, exclusive of IPT. He accepts that the additional 15% for expenses and profits is not unreasonable.

Mr Clegg accepts as reasonable the suggested minimum premium of £100,000, that was based on 13.33% of the expected adverse costs and disbursements. Had the ultimate figure for Macfarlanes’s costs been known at the time, the minimum premium would have been £1, 984,037.

Mr Clegg suggests that the issuing of the policy was rushed. In his opinion Leigh Day should have sought “a more equitable premium”, either by direct negotiation with First Assist, or by employing a professional broker. He suggests that given the competitive nature of the ATE market there was a good chance that alterative insurers could have been found, and alternative methods of premium calculation utilised. He suggests that Leigh Day should have given notice of cancellation to First Assist by 23 February 2007, and renegotiated a more favourable premium in the light of the fact that Trafigura were in negotiation with the Ivorian Government with a view to setting up a compensation fund.

In answer to the question:

“Whether the underwriting approach utilised by First Assist in this case is typical of the underwriting principles normally applied by ATE insurers?”

Mr Clegg states:

“In the vast majority of single claimant cases, and sometimes perhaps in multiple claimant actions, the underwriting approach and premium charge described by Mr Smith will produce a result that is likely to be within a reasonable range based on the quantum risk and the level of litigation work undertaken by the parties. It does have the ability though to produce an inequitable result if the costs are very high, but the litigation risk is not.”

In answer to the question:

“Whether the method of premium calculation used was the only method that in all the circumstances of this particular case was reasonable, or are there alternative methods of underwriting the risk and of calculating the premium?”

Mr Clegg states:

“9.9

A simple method that can be employed by insurers when the original estimate of loss is too variable to estimate with any accuracy is to set, by whatever rationale it chooses, what is considered a premium that will match or exceed the risk of loss – an overly adequate premium. But the insurance contract will allow for a return of premium, in the form of a profit commission, to the party paying the premium in the event that certain pre-agreed margins are made when the outcome of the risk or event insured against becomes known.”

He suggests that it might be possible for the premium to be discounted with a 50% profit commission back for the benefit of the party paying the premium, ie, the premium is discounted retrospectively if the insurer suffers no loss under the policy.

Mr Clegg then turns to the question:

“How likely would it have been for Leigh Day & Co to have obtained a reduced premium and/or better terms if an insurance intermediary had been employed?”

He does not think at the time when the risk was first being placed that a broker could have achieved better than the anticipated premium of £465,000. The likely brokerage charge would have been between 10% and 15%.

In response to the question whether the method of premium calculation was the only method that could have been applied in this case, Mr Clegg points out that this case is unique, the method of calculation used has been applied in the past, but not in cases comparable to this one. He relies on his suggestion for a return premium or profit commission in the event of success which would still leave First Assist with a very significant recovery, but mitigate the sum being claimed as premium.

Finally, in answer to the question:

“Whether any insurer would have written the risk on a sliding scale premium basis, that would reduce the final premium charges if the chances of success increased after inception of the policy?”

Mr Clegg states that whilst it is possible to envisage a sliding scale approach, that would increase the percentage chance of success after inception contingent on certain criteria that would have to be pre-agreed, which he suggests would not be easy, although not impossible. He suggests that had Leigh Day invoked the 14 day cancellation window, they could have negotiated an improved premium based on improved chances of success, and a professional broker could have assisted in this regard.

He expresses the view that had the costs and disbursements remained within the initial estimate provided by Leigh Day of £750,000, the premium of £465,000 would have been commercially acceptable in all the circumstances. He continues:

“Had the true chances of success been only 50/50 or even worse, then there is a commercial argument to be made that a premium of £9,677,554 could be justified, even though this is an extraordinarily high amount.”

Mr Bacon raises three points in relation to this issue; firstly whether the sums claimed from the Defendants are “premium” within Section 29 of the Access to Justice Act 1999, and thus whether that premium should be deconstructed to identify the sums paid to insurers and intermediaries. Secondly, the premium calculation: where the Defendants argue that the premium rating formula set out in the schedule to the policy should be applied to the costs at risk, rather than to Leigh Day & Co’s costs. Thirdly, the question of “adverse costs”: in respect of which it is argued that the premium should be calculated by applying the risk multiplier to the actual costs at risk, not to the sum that the Defendants had actually incurred by way of legal costs, but the sum that the Claimants would be liable to pay if Trafigura’s legal costs were to be subject to detailed assessment, which is now said to be £10,393,031.70.

Is the Premium All Premium Within The Meaning of Section 29 of the Access to Justice Act 1999?

Mr Bacon argues that the premium should be deconstructed because the disclosure provided supports the Defendants’ case that only a part of the premium claimed will actually go to Great Lakes the insurer. He makes the point that First Assist, whilst possessing the skills required to be an insurer, has chosen not to be the risk carrier, but has out-sourced this. In his submission it is only the risk carrying element which has been passed to Great Lakes, and he suggests that it is this cost of out-sourcing the insurance element to Great Lakes that should be the starting point for the calculation of the premium. He suggests that all other benefits and services offered by First Assist under the Pursuit product are provided by First Assist. He suggests that First Assist is therefore acting as a “quasi insurer”. He suggests that the Defendants’ obligation to pay an ATE premium should not be “contaminated or affected” by First Assist’s overall arrangements with Great Lakes.

He raises the following queries: How much is the Great Lakes Risk Carrying Charge? What does First Assist charge? How much is the First Assist profit share? In order to answer these questions he seeks further disclosure, particularly of the figure which best represents the risk carrying charge received by Great Lakes from the premium. This, he says, is the starting point for building a reasonable premium. Without that figure he says, the court is not in a position to assess whether the First Assist charges are reasonable and proportionate, as against the actual cost of providing the insurance.

Mr Bacon’s argument is based in part on his assertion that First Assist shares in the carrying of the risk, and also shares in the profit. He suggests that it is First Assist which is actually paying claims out of its profit, and that there is no burning cost for Great Lakes until the losses exceed the level of First Assist’s experience account.

Mr Bacon poses two “central questions”:

In summary, Mr Bacon argues that First Assist acts like an insurer and shares the risk with Great Lakes. It cannot recover the cost of doing so, since it is not premium. He submits that the court must excise the costs of First Assist providing shared risk within the policy, and the Claimants should be directed to give further disclosure to enable the premium to be deconstructed and then built up again.

At paragraph 57 of his first witness statement Mr Smith states that there was a loading of 15% for administration and profit added to the risk premium rate of 53.8%, giving a premium rate of 61.923% (62%). At paragraphs 23 to 26 of his second witness statement Mr Smith points out that this level of 15% for administration costs and profit was accepted as reasonable in the RSA Pursuit Test Cases, and was he suggests “a transparent mechanism to charge a reasonable sum to meet the costs of running the business”. The percentage also “reflects the work that has to be done by the insurer to keep on top of the case”. Finally, it also represents a profit element.

Relying on the evidence of Mr Clegg, Mr Bacon argues that the 15% administration/ profit charge leads to an “extreme windfall profit element”. He suggests that the administration/ profit element should be allowed at a fixed fee of no more than £100,000, or in the alternative should be charged as a straight percentage at a lower rate (eg, 10%), not as a percentage of the adverse costs but of the final premium.

Following disclosure of the MGA and Remuneration Agreement, Mr Bacon argues that there is no such charge within the make-up of the premium. First Assist does not in fact charge a 15% commission, or any set percentage. This, he says, is a further reason (in addition to those mentioned above) for the premium to be deconstructed so that it may be built up from the figure that Great Lakes receives by way of premium income.

Mr Bacon points out that in his first witness statement Mr Smith sets out the formula for calculating the premium. I have quoted this at paragraph 15 above. This appears to differ from the policy schedule, quoted at paragraph 5 above, in that the policy schedule requires the premium rate to be ascertained by dividing the sum of the opponent’s costs plus expenses by the normal fees of own solicitor. The premium is calculated by taking the premium rate calculation, and multiplying it by the normal fees of own solicitors.

Mr Bacon’s argument is, that if one follows the formula set out in the policy schedule, one would come to a multiplier of 0.20 or 0.30 (based on the exposure multiplicand of £14,884,160, divided by own solicitor costs of £30 million). He says that the policy is then incorrect to require that the 0.20 or 0.30 should be applied to own solicitor’s costs, since those costs are not the costs to which the insurer is exposed. If the figure for own solicitor’s costs is utilised, the premium comes out at £9.2 million.

Mr Bacon also argues that the prospects of success utilised by First Assist were too low, and he bases this on two factors: firstly, that in my decision on success fee, in which I found that a 58% success fee was appropriate, I had taken too low a starting point, and thus arrived at an incorrect success fee. Mr Bacon’s submission is that the appropriate success fee should have been 25%. Secondly, he argues that, since I found that the risk had varied as the litigation progressed, so First Assist should have taken into account the varying risk when assessing the prospects of success. In particular, he suggests that First Assist were aware, or ought to have been aware, of the settlement between Trafigura and the Ivory Coast Government, which, in my success fee decision, I had indicated improved the prospects of success by some 5%. He therefore argues that First Assist should have used a 70% prospect of success rather than 65%. He also suggests that instead of having one after the event insurance policy, a series of policies could have been entered into as more claimants joined the group, and thus the later premiums would have been lower than the earlier ones.

With regard to the timing of the policy, the policy was incepted on 9 February 2007. Leigh Day informed First Assist of rumours of settlement on 12 February 2007. The settlement itself was signed on 13 February, and the BBC reported the settlement on 15 February. He points out that in March 2007, the definition of “success” had still to be agreed. Martin Day and First Assist were still discussing risk, and First Assist should therefore have reflected the change in risk in the policy, the schedule of which is dated 4 April 2007. Since I had found that the Ivorian settlement had brought about a 5% change in risk, in Mr Bacon’s submission, First Assist’s assessment should also move up to a 70% prospect of success. He suggests that since the wording of the policy was still being negotiated in March 2007, it may not have been necessary to cease the policy and issue a new one. The policy does not include any contractual term enabling the insurer, or the insureds, to alter the risk. Applying a test of reasonableness, he argues that this is what should have happened.

Mr Bacon also argues that there should have been separate ATE policies, in the same way that there were separate CFAs. Thus, the later policies should accommodate late joiners, and reflect increasing prospects of success.

Mr Bacon further argues that the term “opponent’s costs”, which appears in the policy schedule, is not defined, and “Adverse Costs”, which does not appear in the schedule, is defined. His case is that the premium should be based on the insurer’s liability to pay adverse costs, ie, only those costs which the opponent could legitimately recover on detailed assessment. In the example which he gave in his skeleton argument, he suggested that rather than the figure of £14,634,160 (£14,884,160 minus £250,000 expenses), a lower figure should be used, and he illustrated his argument utilising a figure of £10,243,912, which, together with the £250,000 expenses times 0.2, produced a premium figure of £2,098,782.

Since the policy under the heading “premium rate” provides:

“The figures used for the calculation of the premium rate and premium in respect of opponent’s costs shall be the actual costs of the opponent (including the opponents’ solicitors costs and the opponents’ expenses) as certified by the opponent’s solicitor”

I invited the Defendants to state what their actual costs (ie, adverse costs) should be, and to indicate the reasons why that figure differed from the original figure given. It was common ground between the parties that the reference to those costs being “certified” did not require a formal certificate to be given.

In response to that invitation, the Defendants have prepared a schedule of their costs, which they say would be properly recoverable as a matter of principle between the parties had the Claimants lost the action, that figure is £10,393,031.70. Using Mr Bacon’s calculation this would produce a premium of £2.8 million.

Explaining the alteration in the figure for Defendants’ costs, Mr Bacon submits that this is not a deliberate attempt to reduce the costs. The original figure of £14.6 million was the figure for the costs actually billed to Trafigura. The client had been billed for other costs outside the events in Abidjan. Macfarlanes and their costs draftsmen have gone through their bills to work out what could be claimed. He suggested that the Defendants had been asked for the information about costs before they had seen the policy wording. There is disagreement between the parties as to exactly when the policy wording was sent to the Defendants. I do not have sufficient evidence to enable me to decide that issue. Mr Bacon argues that if the £14.6 million figure is not reduced, the risk assessment must be altered, since the Claimants were never in danger of having to pay that figure.

Mr Butcher QC dealt, in admirably clear submissions, with Mr Bacon’s deconstruction arguments. He points out that the Defendants accept that the insurer is Great Lakes, and that the agent is First Assist. He refers to Section 29 of the Access to Justice Act 1999:

“Where in any proceedings a costs order is made in favour of any party who has taken out an insurance policy against the risk of incurring a liability in those proceedings, the costs payable to him may, subject in the case of court proceedings to rules of court, include costs in respect of the premium of the policy.”

In order to identify “the premium” Mr Butcher relies on MacGillivray on Insurance Law 11^th Edition at paragraph 7-002:

“Premium Defined

The premium is the consideration required of the assured in return for which the insurer undertakes his obligations under the contract of insurance. (Footnote: ¹)

… The amount or adequacy of the premium in relation to the risks run is a matter for the insurer rather than a court …”

Relying on the definitions which I have quoted, Mr Butcher argues that the premium is what the insureds have to pay in exchange for the insurer’s obligations to indemnify the insureds. He submits that the court should only look at the contract of insurance, ie, the First Assist policy, part of which I have set out at paragraph 4 ff above. The definitions make it clear that the insurer is Great Lakes, and that First Assist administers the insurance on the insurer’s behalf. The policy also defines the premium. This, Mr Butcher says, is what the insureds have promised to pay. It does not mean part of it, nor any money paid to someone else. It is a unitary promise to pay Great Lakes for indemnifying the insureds. There is therefore, he says, no basis for saying this is not “the premium of the policy”.

The definition of Cover (adverse costs and expenses) is simply an insurance promise. The policy schedule makes it clear that the policy is administered by First Assist Insurance Services Limited, and underwritten by Great Lakes Reinsurance (UK) Plc, both of which are regulated. In Mr Butcher’s submission there is no part of the policy which is not insurance. The only thing which the insureds’ promise to pay premium buys, is insurance.

In my decision in the RSA Pursuit Test Cases [2005] EWHC 90003 (Costs) I stated:

“I should say at the outset that I am satisfied that the Pursuit policy is the product of an honest attempt by a reputable insurer to produce an ATE policy to fill a specific gap in the market. Unlike the packages offered by Claims Direct and The Accident Group ([2003] EWCA Civ 136 and [2004] EWCA Civ 575 respectively), the premium payable in respect of the Pursuit policy is exclusively referable to that policy and does not cover the cost of any ancillary benefits for the insured.”

Mr Butcher argues that the position is still the same today, and there is no reason for me to change my original view. None of Mr Bacon’s points were argued in that case, and in his submission they are not arguable now. This case has no similarity to Claims Direct or The Accident Group, where the premium was buying benefits which were not insurance at all. In the Claims Direct Test Cases [2003] EWCA Civ 136, Lord Justice Brooke, with whom the other two members of the court agreed, stated:

“87.

In my judgment, in this quite exceptional case, it was inevitable that the Master should adopt this approach in order to identify what should truly be treated as the premium.”

Mr Butcher submits that what Great Lakes provides is exclusively insurance; enquiry into what the insurer does with the premium is irrelevant, and does not stop the money paid being premium.

The Defendants rely on the fact that Great Lakes has agreed that part of the premium should go to its agent, which carries out certain tasks. In this connection Mr Butcher relies on Callery v Gray (No.2) [2001] EWCA Civ 1246, where the court stated:

“12.

It is important in this context to draw a distinction between two separate matters. The first is the nature of the benefits to which the litigant is contractually entitled in exchange for the payment of the premium. This falls to be determined from the terms of the contract under which the premium is paid. Section 29 permits the recovery of a premium where this is payment for insurance against the risk of liability for costs. If payment of a so-called premium buys a contractual entitlement to other benefits it is, to say the least, arguable that the premium cannot, to that extent, be recovered under section 29. Thus the Court has to consider the terms of the contract under which the premium is paid to see whether it is simply a contract of insurance against liability for costs or whether it is something other than, or additional to, that.

13.

The contractual benefits purchased by the premium must be distinguished from the use made by the insurer of the premium. An insurer will necessarily look to premium income to meet the costs of the business. The primary costs are likely to be those of meeting claims, but the costs will also include matters such as commissions, advertising and, indeed, refurbishing the insurer’s premises. The Court will not be directly concerned with how, or on what, the insurer spends the premium income. The Court will, however, be concerned with the question of whether the premium is a reasonable price to pay for the benefits that it purchases. Ultimately, this should be a question to be considered having regard to experience, or evidence, of the market. If an insurer is conducting his business in a manner which incurs extravagant, extraneous or otherwise unnecessary expenditure, which has to be covered by the premiums, those premiums are likely to be uncompetitive. To pay such a premium where other more reasonable premiums are available may disentitle the litigant from making a full recovery of the costs of the premium.”

On that basis Mr Butcher argues that Great Lakes is entitled to, and receives the premium out of which claims are first paid. Lord Phillips in the judgment of the court in Callery, was saying that the market would decide the level of the premium.

It is ordinary and prevalent insurance practice for insurers to agree with agents that they should do some of the tasks connected with the insurance. Agents may underwrite and administer, but it is the insurer who is solely liable for losses incurred under the policy. In support of this he gave the example of a Lloyds Syndicate where the names bear the risk, but have nothing to do with insurance and cannot interfere with the activities of the managing agents. It has never been argued that premium is not premium because of the way the money is split up. Such arrangements are common in ATE insurance. Mr Butcher refers to the judgment of the Court of Appeal in Callery v Gray (No.1) [2001] EWCA Civ 1117, where Lord Woolf gave the judgment of the court:

“16.

Written evidence was given by Mr Christopher Ward, the managing director of the specialist legal expenses underwriting agency which has run the Law Society’s Accident Line Protect conditional fee insurance scheme since its inception in 1995. During the last six years, this agency has issued over 85,000 ATE policies. The insurance was always issued in conjunction with personal injury claims conducted under a CFA.

17.

This agency is not itself the risk carrier in insurance terms. It manages the insurance on behalf of the underwriter who carries the risk. For the first five years, this was an American company. The Law Society told us that the ATE premiums for this type of cover were originally very modest, because the success rate of personal injury litigation was thought to be so high. Adverse claims experience, however, drove the premium up sharply, and in 2000 the original underwriters withdrew from the market after suffering major losses.”

Mr Butcher seeks to demonstrate that the arrangement between Great Lakes and First Assist is neither irregular nor unusual, by reference to Temple Legal Protection v QBE Insurance [2009] EWCA Civ 453. In that case QBE was the authorised insurer, and the sole party to the insurance contracts with the policy holders, while Temple was the managing agent. Significant insurance functions in respect of the writing and administration of the insurance policies were delegated to Temple, to the extent that it was the “Temple” brand which was dominant, rather than that of its principal QBE. The Court of Appeal explained:

“3.

Although it did not carry the insurance risk itself, Temple had developed the business and was responsible for its administration. It was therefore able to market the cover as its own product and had a developed a large following among solicitors who regularly acted for clients under conditional fee arrangements. … on 1st December 2005 Temple entered into a binder with QBE which came into effect on 1st January 2006. QBE had no previous experience in the writing of legal expenses insurance.

4.

The binder is a lengthy document … By Section 1 QBE appointed Temple its agent to write various classes of legal expenses insurance on its behalf and to issue certificates of insurance to policyholders evidencing cover in respect of insurances bound under the agreement. Temple was also authorised to receive and hold premiums and premium refunds, to receive and hold claims money and to manage and settle claims in accordance with the terms of the binder. …”

Thus the risk under the policy was QBE’s, Temple was the agent and owed all the fiduciary duties expected of an agent.

In his third witness statement Mr Smith confirms that the relationship between First Assist and Great Lakes is that of insurer and managing general agent. He sets out at paragraph 19 to 28 of that statement the arrangement between the two companies. Mr Smith deals with the MGA at paragraphs 9 to 22 of his fourth witness statement. The arrangements described are, in Mr Butcher’s submission, just another example of the normal insurance arrangements between insurer and agent. The MGA is in standard form, and provides at Clause 2:

“2.1

Subject to the terms of this agreement the agent shall perform the services in accordance with the service levels.

2.2

Notwithstanding the generality of Clause 2.1, and subject to the terms of this agreement, the direct insurer appoints the agent from the commencement date as its agent for the purpose of performing such of the services (in accordance with the service levels) as are required to be performed on behalf of the direct insurer in order to create valid and binding legal obligations between the direct insurer and the relevant counter party, or in order to satisfy the direct insurer’s regulatory obligations. The agent hereby accepts this appointment.”

Although at Clause 18 of the MGA it was recognised that First Assist might itself become an insurer, this never happened.

Clause 22 of the MGA gives the insurer the right to appoint a management team to step in, to manage or provide the services or any part of the services, if the agent is in material breach of any of its obligations, or if the direct insurer is required by a regulatory authority to exercise its rights under the clause. This is because Great Lakes is ultimately responsible for the insurance obligations.

The Remuneration Agreement is a function of the respective position of the parties, but it does not alter the status of the insurer. It has nothing to do with the insureds, and does not alter the character of the premium.

As Mr Smith has stated, the Financial Services Authority is familiar with these arrangements, and has not queried them.

Dealing with Mr Bacon’s point about the 15% administration charge, Mr Butcher relies on the explanation given by Mr Smith in his first witness statement, which I have dealt with at paragraphs 10 to 26 above. The addition of 15% to the burning cost explains how the burning cost is loaded. Burning cost is the best estimate of premium that will cover a pay out on loss.

Mr Butcher argues that the split of premium between First Assist and Great Lakes is merely an agreed split between principal and agent. Mr Bacon’s suggestion that if there is a negative balance, and First Assist “is unable to meet its obligations” Great Lakes would have to step back in and pay the claim is incorrect. Great Lakes alone is under the obligation to the insureds, First Assist is under no such obligation. Mr Bacon’s suggestion that First Assist acts as a “quasi insurer” is meaningless. First Assist provides insurance services, it is an intermediary and almost all agents act in this way.

With regard to Mr Bacon’s query: “what was the premium, for the purposes of Section 29 of the 1999 Act?” Mr Butcher argues that the premium is that set out in the policy, ie, the sum paid by the insureds to secure the insurance. Great Lakes make use of all the premiums received to pay claims, and Great Lakes’ premium is the full amount. Great Lakes receive a risk carrying charge, because First Assist carries out the various tasks required under the MGA. If there is a profit, First Assist retains it. Great Lakes receive the risk carrying charge. None of these however, it is submitted, affect the reasonableness of the premium. If all the services were provided by a unitary insurance company, the position would be no different than if the premium were split with an agent. It is therefore not right to ask what part of the premium is retained by the insurer. Great Lakes would not have taken on this risk on any other terms. It would only do so on the basis of the First Assist Scheme.

Mr Butcher submits that Mr Bacon’s submission that what Great Lakes is providing is a form of financial guarantee is misconceived. The position is as set out in the policy.

With regard to the questions raised by Mr Bacon in his skeleton (what does Great Lakes charge? how much is the Great Lakes risk carrying charge? what does FA charge? how much is the FA profit share?), these questions start at the wrong place. The appropriate starting place is the policy document. The arrangements between First Assist and Great Lakes are not relevant and will not help with arriving at a reasonable premium. The risk carrying charge is not the true basis of the premium. The premium must cover the burning cost, together with administration and profit. All the premiums received go to pay the claims, any surplus being distributed as profit.

Mr Williams submits that Mr Smith’s simplified premium calculation in his statement is correct. He argues that the proposition that the terms of the policy should be disregarded, so that the premium is calculated by applying the premium rate multiplier to “exposure”, is misconceived. The multiplier to be applied must represent risk, and, in his submission, this is not the premium rate multiplier, it is the 61.923%. He says that the only argument should be about the correct figures for exposure and risk.

With regard to the timing of the policy, Mr Williams submits that the correct starting point is the state of knowledge of First Assist. In his submission a 65% prospect of success was reasonable, even if the Ivory Coast settlement deal had been known to them. To support this proposition he relies on paragraph 440 of my judgment of 15 February 2011:

“… the first CFA in November 2006 correctly indicated a 50% chance of winning. Once the acknowledgment of service had been served on 4 December 2006, the risk in respect of forum would have gone. The chance of winning therefore increased to 61.2%. On 21 March 2007 the Defendants informed the Claimants of the settlement agreement between Trafigura and the Ivorian Government. This would, in my judgment, have reduced the breach of duty risk by 5%, increasing the chance of winning to 65%.”

The inception date of the policy was 9 February 2007, and the deal was notified to First Assist on 21 March 2007.

On 9 February 2007, Leigh Day had written to Macfarlanes:

“The author has just returned from a trip Abidjan where there were many rumours circulating regarding negotiations between Trafigura and the Cote d’Ivoire Government. We understand that those negotiations have been reported in the local press this morning.

It may well be that the negotiations are matters that do not concern our clients, such as the release of the two executives in prison and the losses incurred by the Government as a result of the dumping of the waste. We simply write to ask whether there is any intention of your client’s part to enter into early negotiations with us on behalf of our clients, to also see the resolution of the health related claims. While these are still relatively early days regarding the bringing in of the individual claims, if your clients are keen to see a resolution of the whole issue there may be sense in having preliminary discussions soon rather than later.”

Macfarlanes replied on 21 March 2007:

“As you are aware our clients have reached an agreement with the state of the Ivory Coast which – without admissions of liability on either side – has sought to respond to the needs of the individuals in the Ivory Coast by establishing effective means whereby, irrespective of fault or recrimination, the innocent victims of the events in Abidjan could be compensated.”

The letter goes on to explain the compensation fund and how it works. The letter continues:

“The individuals will only be required to demonstrate a causal link between exposure to the materials and their medical symptoms or losses – only one element of the many that they would have to plead and prove in order to succeed on a claim in tort in the English High Court. Further, claiming under the Fund will obviously be a far more straightforward, cost effective and efficient process by which the individuals concerned will obtain compensation.

In addition, the payment of compensation to your clients and other affected individuals will have a material effect on the level of any damages recoverable by such Ivorian individuals in the English High Court.”

The letter then goes on to invite Leigh Day to agree to stay all claims commenced on behalf of the Claimants, until such time as those Claimants have made applications to the Compensation Fund and payments have been made from the Fund.

Mr Williams points out that Mr Smith, at paragraph 13 of his second witness statement, states:

“Had First Assist known of the payment prior to offering terms, it would have factored that knowledge in to the risk assessment. What effect that would have had on the terms offered is impossible to say, since to this day I have no real understanding of the basis on which funds were paid, save that it has been reported as amounting to a payment of around £100 million, which coincided with the release of Claude Dauphin, the Founder of Trafigura, plus another senior executive, from prison in the Ivory Coast, and was followed by Trafigura and its staff being granted immunity from suit in that country. The somewhat impenetrable arrangements, even if known to us, would therefore have thrown little or no light on the material question of whether Trafigura was liable in principle to compensate the Claimants for a civil wrong recognised in the English Courts. …”

Mr Williams accordingly submits that a 65% prospect of success was still appropriate. With regard to the proposition that the Claimant should have novated the policy on better terms, he says that the Claimants would have been foolish to do so, since there was no evidence that better terms were available. The solicitors’ duty to the Claimants was to act reasonably in their interests, and it was not in the interests of the Claimants to cancel a good policy where there could be no expectation that First Assist would improve the terms, and always the risk that it would offer less favourable ones.

Since the figure of £14,634,160 was the figure originally put forward by the Defendants as “opponent’s costs”, that is the figure which should be utilised. The Defendants’ “adverse costs” are now said to be £10.4 million. Mr Williams suggests that this represents a reduction of some 29%, and makes the point that the Defendants’ submission is not supported by any disclosure. Thus he argues that the revised figure should be disregarded. He suggests that it is implausible to suppose that the costs were overstated by such a margin.

Mr Williams points out that the original figure of £14.6 million supplied by the Defendants’ solicitors remained unchanged, through the exchange of Points of Dispute and Replies over the course of 12 months, and was relied on by the Defendants to compare with the Claimants’ costs claim. The Claimants’ request that the Defendants should waive their privilege in respect of the solicitor and own client bills was refused. In his submission, the bulk of the costs would have been incurred in respect of counsels’ fees and experts’ fees. The Claimants had insured against the maximum exposure, ie, £14.6 million, not the minimum.

He suggests that the costs in respect of the Bou injunction should not be reduced, and there would be a dispute as to whether or not the costs of experts not relied on were recoverable. Also the 15% reduction was, he says, double counting.

Relying on my own judgment in the RSA Pursuit Test Cases, and the decision of the Court of Appeal in Rogers v Merthyr Tydfil, Mr Williams submits that the court should uphold the policy as fair, and, since the premium was the only one in the market, the Claimants, he suggests, had no choice but to accept it.

With regard to the risk multiplier, Mr Williams points out that in my success fee judgment I had assessed the prospects of success at 61.2%. He suggests that First Assist’s assessment of 65% was within an acceptable bracket, but goes on to point out that the risks to First Assist were different from those facing the solicitors, who might have lost individual cases, and were faced with issues of breach of duty, medical causation and other causation. The question for the court is, he says, was the insurer’s view of 65% reasonable? It cannot be said that a 35% risk of losing is overstated, particularly in the light of Mr Nurney’s views expressed in his witness statement. In any event, he suggests that my assessment of the prospects of success are not necessarily relevant to the insurer’s assessment of risk. First Assist used a different method to that used by the solicitors, ie, by induction. He argues that the starting point for the premium is a multiplier of 0.54, which is increased by 15% for administration and profit, giving an overall multiplier of 0.62. This is, he suggests, well within market norms.

With regard to Mr Bacon’s suggestion that First Assist “pretended” that the 15% addition was for administration, Mr Williams points out that the 15% was the allowance in the insurance contract price, being the uplift above the pure burning cost, giving a multiplier of 62% (61.923%). Mr Williams submits that 15% is well within market norms, and may even be modest in a volatile market. He points out that in the RSA Pursuit Test Cases, Mr Christopher Wait of Temple Legal Protection Ltd would have added, in the circumstances of that case, a percentage for profit and insurance administration costs of 60%. In Rogers v Merthyr Tydfil CBC, DAS applied an uplift of 25% for overheads. In the same case Temple and Keystone took the view that 25% was a minimum allowance for overheads, and that a more realistic allowance would be closer to 50%. DAS was in fact making a loss at 25%.

Mr Myrtle expressed the view that 15% was well within market norms, and Mr Clegg accepted that 15% was not an unreasonable percentage in what he called “the normal course of business with normal premiums”. He goes on to express the view that as the costs escalated, so the additional percentage for administration and profit became “an extreme windfall profit element”. Mr Williams argues that the more an insurer has to use capital in one case, the less there is for another case. Furthermore, it is not open to the court to say what an insurance company’s profit should be. In his submission the premium and its calculation is reasonable, it was a transparent transaction with very high exposure. Exactly the same process would have been used if the exposure had been £14,000, rather than £14 million.

Finally, he points out that the Defendants were offered one way costs shifting, an offer which was declined.

In his submission, the Claimants have acted reasonably in utilising the First Assist policy, which insured the full claims which might be made against the Claimants, ie, the costs notified by the Defendants. He suggests that paying parties will never over declare their costs, and, in some cases, might under declare. He does not suggest that this is happening in this case. He does, however, argue that further deduction from the figure originally given by the Defendants would be unsafe, and would not yield an adequate premium.

He points out that First Assist is exposed in unsuccessful cases, not only to the substantive costs, but also to the costs of the subrogated assessment proceedings and interest, in addition they would lose the premium.

He submits that the contractual formula is a genuine attempt to reach a reasonable outcome, notwithstanding the obvious dangers of the system to First Assist itself (ie, under declaration). Given that this is a reasonable methodology, the court should not interfere with it.

Mr Williams argues that if Mr Bacon’s figure of 0.2% is correct, this would represent an 83% chance of success, and thus only a 17% exposure for the insurer. He points out that in the submissions on the success fee issue, Mr Bacon had argued for a success fee of 25%. In addition to this, in Mr Nurney’s witness statement it was contended that the Claimants had no arguable case. First Assist’s view that the prospects were 65% were slightly higher than the view which I had taken, that the prospects were in the region of 63%, a figure which, if utilised, would produce a higher premium.

In relation to Mr Bacon’s submissions in respect of deconstruction of the premium and the 15% addition for administration and profit, I accept Mr Butcher’s submissions, which are clearly correct. In both the Claims Direct and Accident Group cases I was persuaded, exceptionally, to deconstruct the premiums, because it was clear from the evidence in each of those cases that the so called premium, whilst containing an element for insurance, contained numerous other elements which were clearly not insurance. Thus, the only way to arrive at the figure for premium was to carry out a deconstruction, and make an allowance for that element which was truly premium. In this case, as Mr Butcher has pointed out, there is one premium, one insurer and one agent. There is no basis whatsoever upon which this court could or should deconstruct the premium. The question for the court is whether the premium sought is reasonable and proportionate. The two areas to be considered are the exposure and the risk multiplier.

I have referred, at paragraph 112 above, to the judgment in Callery v Gray (No.2). In the Claims Direct Test Cases [2002] EWCA Civ 428 Arden LJ, referred to the decision of the Court of Appeal in Callery v Gray (No.2) in these terms:

“46.

… The court specifically added “satellite litigation involving such an exercise [ie, examining evidence of insurance cover] is however unsatisfactory. The Judge can only be expected to give broad consideration to such evidence. It is not part of a function of a Judge assessing costs to carry out an audit of the insurance business.

…

The court may wish to check the overall result which it reaches by reference to the alternative methods of obtaining access to justice. This might involve looking at alternative rates of cover, or the costs which would be involved if litigation were to be funded in some other way ….”

It is evident from the judgment of Lord Woolf in Callery v Gray (No.1) (paragraph 114 above), and the judgment of the Court of Appeal in Temple Legal Protection v QBE Insurance (paragraph 115 above) that the arrangement between Great Lakes and First Assist is a normal business arrangement between insurer and intermediary. Mr Bacon’s submission under this head fails.

Mr Williams was able to demonstrate, by means of worked examples, that dividing the relevant figures by the normal fees of own solicitor, and then multiplying the resultant figure by those same fees, makes no difference whatsoever to the overall calculation, since the two exercises cancel each other out.

The flaw in Mr Bacon’s argument is that if one does away with the requirement in the policy schedule to multiply by solicitors’ normal costs, one also has to do away with the requirement to divide by normal costs as well. The correct calculation is, therefore, as set out in Mr Smith’s witness statement. The crucial question, as Mr Williams argues, is whether the 61.923% risk multiplier is correct in all the circumstances.

Had the case proceeded to trial, and the Claimants been successful, the ATE premium would have been even higher, since the insurer’s exposure to costs would have been significantly more than the £14 million Defendants’ costs incurred up to the date of settlement.

With regard to Mr Bacon’s suggestion that a series of policies should have been entered into, this is quite clearly not a viable proposition. Leigh Day & Co needed to have adequate after the event insurance in place in order to stave off the Defendants’ security application, which would inevitably have stifled the litigation. With regard to the varying risk prospects, this is more than adequately dealt with by Mr Smith in his evidence, which I accept. Leigh Day’s risk was at all times greater than that of First Assist, and the prospect of success identified by First Assist was calculated on a different basis from the solicitor’s risk assessment.

With regard to the suggestion that, following the settlement between Trafigura and the Government of the Ivory Coast, the policy should somehow have been renegotiated, or even cancelled and a fresh policy entered into, I again accept the evidence of Mr Smith and Mr Myrtle. Even if First Assist had been aware of the settlement at an earlier stage, and had taken into account the possible alteration in risk, there is no evidence that their final assessment of the prospects of success would have altered. First Assist may well have taken the view that having paid out $200,000,000 Trafigura would resist even more strongly having to pay out any further compensation. I accept that cancelling the policy, and attempting to obtain a fresh policy, carried with it the unacceptable risk to the Claimants that no insurer would be prepared to accept the case.

Mr Bacon argues that my decision on success fee, and the appropriate risk factor, was flawed, and he puts forward a number of points why this was so. Since my decision has already been given, I do not propose to explain it further under this head. I merely record the fact that it is Mr Bacon’s case that my risk assessment was flawed. One of the matters which he put forward (which was not argued on the earlier occasion) was that the risk in any case can be no greater than the weakest risk factor. His authority for this being “Risk Assessment in Litigation, Conditional Fee Agreements, Insurance and Funding” by David Chalk, who was at that time Principal Lecturer at Anglia Polytechnic University, published by Butterworths in 2001. Mr Williams argues that this theory is itself based on a book by Dr Cohen in 1977, which was a hypothesis not made in relation to CFAs. He goes on to argue that the proposition that all risks must be subsumed against the greatest risk is untenable, and gives the example of a policy covering fire, flood and theft, all of which are assessed as having a 10% risk. He asks rhetorically if the overall risk is 10%? Clearly the answer has to be that the risk must be higher than 10%. I do not propose to comment further on risk assessment.

With regard to Mr Bacon’s suggestion that the risk should have been adjusted as the case proceeded, this is, Mr Williams submits, not the way group insurance works. I accept that submission. Had there been no insurance policy in place, the Defendants would have applied for security. First Assist regretted giving unlimited cover, and subsequently sought to ameliorate the position, but without success. The risk to the insurer could, of course, have increased as well as decreased.

The Defendants were offered the option of one way costs shifting by the Claimants, but declined it.

Mr Bacon’s suggestion that the administration/profit element should be allowed at a fixed fee of no more than £100,000, or should be charged at a lower rate, for example 10%, is without foundation or evidence. Save that £100,000 was quoted as a minimum premium at the outset. The suggestion of a 10% administration/profit element is contrary to normal insurance practice, and is an attempt to reduce the premium to counteract the “extreme windfall profit element”. There is, however, nothing to support this proposition as being appropriate, and I reject it. Similarly, for the reasons which I have given, I reject his suggestion that First Assist should have used the 70% prospect of success, rather than 65%.

Whilst there is no denying that the premium claimed in this case is extremely high, it is not open to me to reduce that premium, purely on account of the fact that it is so large. In order to reduce the premium I would need expert evidence of policies providing similar cover available in the market at lower premiums. There is no such evidence in this case. Mr Clegg, in his evidence, accepted that an additional 15% for administration and profit was not unreasonable. He accepted that a minimum premium of £100,000 was reasonable. He did not think that, at the time when the risk was first being placed, a broker could have achieved better than the anticipated premium of £465,000. The likely brokerage charge would have been between 10% and 15%, and had the Defendants’ costs and disbursements remained within the initial estimate, the premium would have been commercially acceptable in all the circumstances. He concluded:

“Had the true chances of success been only 50/50, or even worse, then there is a commercial argument to be made that the premium of £9,677,554 could be justified, even though this is an extraordinarily high amount.”

No criticism is made of the actual mechanism used by First Assist in order to calculate the premium, save that it is too high.

Reductions have been made by the Defendants under a number of heads, including ancillary costs. The Defendants’ submission states that this reduction is principally made up of costs which are in fact:

“in relation to advice beyond and unrelated to the Abidjan Personal Injury Group Litigation – such as consideration of separate civil and criminal claims in the Cote d’Ivoire, criminal investigation proceedings in other jurisdictions, international investigations and reports, as well as regulatory advice”.

It is suggested that these expenses were:

“simply billed under the matter number for the Abidjan Personal Injury Group Litigation as a matter of administrative convenience.”

The reduction under this head is £2.02 million.

Under the heading “Injunctions” a reduction of £225,000 is made in relation to the Bou injunction in April/May 2009. It was suggested that these were not recoverable on an inter partes basis. However, the costs in those proceedings were reserved, and swept up in the final agreement.

Under the heading “solicitor and client”, almost £600,000 has been deducted, being the costs of:

“various legal advisers around the world in respect of the Probo Koala incident, resulting in the need for those lawyers to keep one another appraised. Such liaison, and the associated cost did not relate, however, to the Abidjan Personal Injury Group Litigation.”

The Defendants’ paper suggests that there was also a significantly higher than normal element of solicitor and own client costs, which would not be recoverable inter partes, eg, regular reporting to the Board, meetings in respect of which extraneous matters were considered, and advising on requests for information from third parties.

Further reductions of just under £57,000 have been made under each of the heads “Administrative” and “Overheads”, the overheads are said to relate principally to secondment charges, library, post and courier charges.

A reduction of £125,000 has been made in respect of experts who were not relied on, and a further reduction of £63,600 in respect of other Trafigura lawyers. Although these fees have been billed to the Defendants by Macfarlanes, it is submitted that they cannot be said to be recoverable on an inter partes basis, as they relate to matters beyond and separate from the Abidjan Personal Injury Group Litigation. These reductions reduce the figure which the Defendants say would have been recoverable by them to approximately £11.5 million. A further reduction of 15% is then made “to balance for unreasonableness”. The figure deducted is £1.093 million, this appears to be a reduction of approximately 9.5%, rather than 15%. The overall result is that the figure for Defendants’ costs is reduced from £14,634,160 to £10,393,031, to which must be added £250,000 Claimants’ disbursements.

The policy provides, under the heading “Premium Rate”:

“The figures used for the calculation of the premium rate and premium in respect of opponent’s costs shall be the actual cost of the opponent (including the opponents’ solicitors’ costs and the opponents’ expenses) as certified by the opponents’ solicitors.”

Neither the original figure of £14.6 million, nor the revised figure of £10.4 million have been certified by the Defendants’ solicitors, but nothing turns on that. The original figure put forward, plus the Claimants’ expenses of £250,000, produces a premium of £9.23 million. Using the Defendants’ amended figure for costs, produces a premium of £6.51 million. In my judgment, however, there are difficulties with regard to the Defendants’ revised figures. Given that the Defendants originally put forward a very much higher figure, there needs to be evidence to support the reductions, particularly in regard to ancillary, solicitor and client and other Trafigura lawyers. I was told that the Defendants have extreme sensitivity about their files, and are accordingly unwilling to give disclosure, but when the court is faced with reductions of this magnitude (which are not accepted by the Claimants) it is not open to me to calculate the premium on the revised figure without some evidence to justify those figures.

Not having heard detailed argument about the figures, I express a preliminary view that the reduction, in respect of the Bou injunction, should not be made, nor should the reduction in respect of experts not relied on. With regard to the 15% reduction for unreasonableness, this appears, in fact, to be a reduction of approximately 9.5%, but nonetheless, there is no reason to suppose, if other appropriate reductions had been made elsewhere, that a further 15% (or 9.5%) should be deducted in addition.

Accordingly, the options for the Defendants are either to accept that the premium should be calculated on the original figures submitted, or to produce evidence justifying a lower figure. It is, of course, open to the parties to agree the appropriate figure if possible. Should the Defendants decide that they wish to submit evidence in respect of the reduction in their costs, I will hear submissions as to the appropriate directions to be given.

I answer the questions raised in Key Issue 14 as follows:

The answer to the question: “Is the premium claimed a premium within the meaning of s29 of the Access to Justice Act 1999? is: yes.