These appeals concern disputes about the construction and effect of the excess layer of a professional indemnity policy issued to Rathbone Brothers plc (“Rathbone plc”) by the members of Lloyd’s Syndicate 2007 and other underwriters (“the excess insurers”) for the 2008 year of account. Rathbone plc is a substantial international group whose trust business included the management of family trusts for wealthy clients. The disputes arise out of litigation before the courts of Jersey brought by certain beneficiaries of a discretionary trust set up by the late Mr Jack Walker, an industrialist and for many years chairman of Blackburn Rovers FC (“the Walker Trust”). These beneficiaries allege that the trustees, who at the material time included Paul Egerton-Vernon (“PEV”), a solicitor who acted as an employee of, and subsequently a consultant to, Rathbone plc’s Jersey subsidiary, made poor investment decisions in breach of their professional and fiduciary duties from the end of 1999.

The excess insurers contend that for various reasons PEV is not covered by the policy either at all or with respect to the particular liability in issue. Furthermore, they submit that even if they are liable to PEV, they are only liable for the excess after other sources of insurance and indemnity have been exhausted. This includes an indemnity under which Rathbone has contractually agreed to indemnify PEV from liability up to a certain level (£40 million per event excluding fraud and wilful misconduct). Furthermore, the insurers submit that they are entitled to be subrogated to PEV’s contractual right to the indemnity. The underwriter of the primary policy, AIG, played no part in these proceedings. It has accepted cover whilst reserving its position depending on the outcome of these proceedings.

Burton J found that PEV could recover under the policy and that the excess insurers were not entitled to rely upon the excess clause. However, he held that they were subrogated to PEV’s right to sue on the indemnity once they had paid out under the policy. The excess insurers appeal the two findings which went against them and Rathbone plc appeals the finding that the insurers have an effective right of subrogation against them.

PEV has been a solicitor since 1971. He moved to Jersey in 1984 and thereafter developed an international practice in trusts. He has been a trustee of a substantial number of trusts, including the Walker Trust. When he first went to Jersey he became a partner of Nigel Harris and Partners. In 1996 that firm became responsible for carrying out the administration of the Walker Trust pursuant to an administration agreement. In November 1998 it transferred its trust business to Nigel Harris Trust Company Limited (“NHTC”) and PEV became a shareholder and director of that company.

On 31 March 2000, NHTC was acquired by Rathbone plc. By an agreement dated 31 March 2000 (“the employment contract”), PEV became an employee of NHTC, which changed its name on 7 May 2002 to Rathbone Trust Company Jersey Limited (“Rathbone Trustees”). On 31 July 2003, Rathbone Trustees and Rathbone plc entered into what was entitled “An Instrument of Release and Indemnity” with PEV (“the Rathbone indemnity”). This indemnified him with respect to certain liabilities arising from the performance of his duties. PEV ceased to work full time on 30 June 2007 and from 1 July he became a consultant of Rathbone Trustees and entered into a number of consecutive agreements on materially the same terms, commencing with an agreement dated 3 August 2007 (“the consultancy agreement”). (On 15 October 2008 Rathbone Trustees was acquired by Hawksford Holdings Limited but nothing turns on that transfer in this appeal because the claim is made under the 2008-2009 policy in relation to the period ending with the acquisition of Rathbone Trustees.)

PEV had been appointed a personal trustee of the Walker Trust on 9 July 1987 and he remained in that office until he retired on 21 July 2009. Lex Nominees International Limited, a company supplied by Nigel Harris & Partners, was at that time appointed a corporate trustee. It is normal in Jersey to have such an arrangement to ensure continuity in the succession of trustees and also the involvement of an individual with personal responsibility towards the Trust. Subsequently in July 1997 another company, Walker Representatives Limited, replaced Lex Nominees as the corporate trustee. Its directors were partners or directors of Rathbone Trustees and included PEV.

After Rathbone plc acquired NHTC, the responsibility for the administration of the Walker Trust pursuant to the 1996 administration agreement fell on Rathbone Trustees. The administration agreement set out the terms on which the Trust was administered. This included procuring the services of its directors to act in their capacity as directors of Walker Representatives Limited, the corporate trustee. Accordingly, Rathbone Trustees had detailed involvement in all aspects of running the Trust. The fees for their services did not distinguish the different functions of personal trustee, corporate trustee, and administration.

Nigel Harris & Partners, and subsequently Rathbone Trustees, submitted invoices at all times from 1996 onwards to the Trust for all services provided to them. The invoice would identify in some detail the time spent by various officers and employees of Rathbone Trustees, including PEV, without differentiation as to whether these services were provided in respect of administration or corporate trusteeship or, in the case of PEV, in his capacity as a personal trustee. This basis of charging was never questioned by the trustees.

As an employee of Rathbone Trustees from 31 March 2000, PEV was remunerated pursuant to his employment contract. This provided for a six-figure salary, together with certain benefits, specifically on the basis that the sum paid was inclusive of any remuneration to which he may be entitled by way of holding office in any other external body. Accordingly, he had to give credit to Rathbone Trustees in respect of the remuneration he received by virtue of being a personal trustee of the Walker Trust. This arrangement continued when he became a consultant. His remuneration was then one half of all time-costed fees recorded and billed to clients and paid, but it was specifically provided by clause 5.2 that “the consultancy fee will not include any fixed trustee or director’s responsibility fees which shall accrue to [Rathbone Trustees] absolutely”.

I have referred to the Rathbone Indemnity made on 31 July 2003 between Rathbone plc, Rathbone Trustees, and PEV. The recitals to it were as follows:

The terms of the Indemnity are that PEV will be indemnified by both Rathbone plc and Rathbone Trustees, whose liability is joint and several, for liabilities arising from the performance of his services, excluding liabilities arising from fraud or wilful misconduct. There is a limit of £40 million “in respect of each event giving rise to any liabilities.”

There is no specific reference in the Rathbone indemnity to any insurance arrangements, although Rathbone Trustees did in fact provide insurance for its staff as required by Jersey financial regulations. However, the consultancy agreement, made in August 2003, the month after the Rathbone indemnity had been entered into, contained an express clause concerning the provision of professional indemnity insurance in the following terms:

Rathbone plc took out insurance for itself and its subsidiaries, including Rathbone Trustees, with AIG and the excess insurers. AIG is responsible for the first layer of £5million, and the excess insurers are responsible for the excess, limited to £45million.

There are three distinct areas of dispute in these appeals. The first is whether the risk is covered by the insurance. The second is whether, by virtue of an excess clause in the policy, PEV must first exhaust any remedies he has otherwise than against the insurers, and in particular, whether he must first seek to cover his loss by relying upon the Rathbone indemnity. The third is whether the insurers, if they are liable to PEV, can be subrogated to his rights so as to enable them to take advantage of the Rathbone indemnity. I will deal with the three issues in turn after considering the policy terms which define the scope of the cover.

The basic risk insured is identified in Clause 1 of the policy:

There are therefore two basic requirements. First, the party claiming under the policy must be an “insured” as defined by the policy. Second, the loss must result from a civil liability.

“The insured” is defined by clause 3.9 as being “any insured company or any insured person”.

By clause 3.10 an insured company is “the policyholder or a subsidiary”. The policyholder is defined by clause 3.18 as “the organisations specified in item 1 of the Schedule” and that in turn refers to everyone, both insured companies and insured persons, for whom cover is provided.

Clause 3.11 defines an insured person as:

Paragraphs (iv) and (v) concern a spouse, executor or administrator who may become liable in certain circumstances. The clause then continues:

In this case Rathbone plc has not suggested that PEV, during the period when he was a consultant, fell within either paragraph (i) or (ii). It submits that he fell within the concept of “a paid employee” in sub-paragraph (iii).

The policy covers “Civil liability” which is defined by clause 3.3 as “a legally enforceable obligation arising from a wrongful act”. In clause 3.30 “wrongful act” is defined as:-

So the definition of “civil liability” and “the insured” has the effect of restricting the cover to liabilities resulting from the performance or failure to perform professional services. That concept is defined by clause 3.21 as:

In order to understand the scope of the policy, it is therefore important to have regard to “the submission”. This is defined in very broad terms by clause 3.27 as:

The insurers’ case emphasises that the alleged wrongful acts all arise out of PEV’s performance as a personal trustee, and furthermore the relevant claim was made during the period when he was a consultant. Accordingly, he was not working under a contract of employment. They submit that, having regard to these features, he does not fall within the scope of the policy for two distinct reasons. First, he was at no stage whist performing his duties as a personal trustee a “paid employee” as defined by clause 3.11(iii) which (see para. 19 above) required that the work be under the direct control or supervision of an insured company. Second, he was not in that capacity performing “professional services” as defined in clause 3.21. There are a number of limbs to each of these arguments. Although these grounds are to some extent interrelated, I will consider them separately.

(a) Not a paid employee when acting as trustee: The first argument is that at no stage, even indeed when PEV was employed under a contract of employment by Rathbone Trust, did he fall within the scope of “paid employee” within the meaning of clause 3.11(iii) when he was acting as a personal trustee. He was never in that capacity acting under “the direct control or supervision” of Rathbone Trustees, as the definition requires. As a personal trustee he had to exercise his own judgment and take personal responsibility for his actions; Rathbone Trustees could not direct him how he should carry out his duties. (A related argument, which I consider below, is that when acting in that capacity he was not, to use the language of clause 3.21, acting “by or on behalf of” Rathbone Trustees so as to bring those activities within the definition of “professional services”).

Second, from 1 July 2007 PEV became a consultant and ceased to be an employee. He could not then be sensibly described as a “paid employee” within the meaning of the policy and therefore was not covered by the policy for any of his professional activities from that date.

The judge rejected both these arguments, and so would I.

(b) No control or supervision: I reject the submission that there was no such direct control or supervision of PEV by an insured company for a number of reasons. First, the logic of the submission is that nobody exercising personal judgment which could not be overridden by the employer would fall within the insurance cover in relation to the exercise of his judgment. This would not just include PEV when acting as a personal trustee but also when acting as a director of a corporate trustee. He would still be obliged to act in what he considered to be the best interests of the beneficiaries of the trust, even if he could then be outvoted by other corporate representatives. It would be an extraordinary limitation on the cover provided by this policy if, when exercising their role as trustees, paid employees were not covered. The provision of trustee services both corporate and personal, is a major part of Rathbone’s business. Indeed, it is one of the principal reasons why it would seek professional indemnity insurance of this kind. Moreover, as one would expect, the submission to the insurers included a list of business activities which made it plain that the professional services included “full UK and international trust and nominee services including personal and corporate trustee services”. It also included “the administration of existing structures … including the provision of trustees”. Another category specified in the submission was expert witnesses who again would have to take personal responsibility for their evidence. In my judgment, there can be no doubt that the insurers were undertaking to cover liabilities resulting from the exercise of these functions, notwithstanding that the employer could not directly control their exercise. The concept of “control or supervision” should be construed accordingly.

In any event I do not think that the concept of “direct control or supervision” is intended to limit the professional services covered by the policy. The purpose is to assist in identifying those who should be treated as paid employees and therefore fall within the cover, as opposed, for example, to those such as independent brokers providing services for the group who are not under such control and who would not. This is why the requirement is found in the definition of “insured person” and not in the definition of “professional services”. “Direct control or supervision” is relevant in defining the person covered by the policy but is not intended to cut back on the activity, provided it is otherwise a professional service. Once someone is defined as a paid employee he is covered for the range of professional services performed in that capacity which are identified in the submission irrespective of the degree of control which the employer may exercise with respect to the exercise of any particular function.

Even if I am wrong about that, it is also pertinent to note that the language of the clause is “control or supervision” and it seems to me that there is quite extensive supervision, even under the consultancy agreement, in relation to the way in which the duties as a personal trustee are performed. The judge pointed out that the role of personal trustee is highly regulated under Jersey law and, as the judge expressly found, Rathbone Trustees exercised detailed supervision over PEV’s activities in accordance with a Code of Practice whose purpose is to ensure that trustees comply with the rules laid down by the Financial Services Commission in Jersey. In addition, both the employment contract and thereafter the consultancy agreements included numerous terms which had a bearing upon the exercise by PEV of his functions as personal trustee. For example, under the latter he had to provide information about his activities; to provide notes or memoranda which were to remain the property of the company relating to those activities; to agree to the monitoring of his activities; and to resign as a trustee when the company by written notice required it. Whilst none of these powers interfere with PEV’s personal duties or responsibilities as trustee, they can sensibly be described as involving the exercise of both control and supervision over his activities in a manner compatible with those duties and responsibilities.

(c) Not a paid employee when a consultant: I turn to the alternative argument that once PEV became a consultant as from 1 July 2007, he could not thereafter be described as a “paid employee” within clause 3.11(iii). The respondents concede that from that time he would not be categorised as an employee at common law. Indeed, the consultancy agreement itself described PEV as working under “a contract for services and not a contract of employment”. However, the respondents contend that the policy has its own definition and that PEV falls within it.

The question is whether the policy was intending to reflect the common law concept or whether it was adopting a more expansive definition. Burton J held that it was the latter. The insurers submit that this was erroneous. First, the natural sense of an employee is someone working under a contract of employment, and this is reinforced by the requirement that the employee should be under the supervision and control of the employer, a phrase redolent of the common law definition. Second, if the broader definition had been intended there would have been no need to separate out paragraphs (i) and (ii) in the definition of an “employed person”.

Clause 3.11 is not entirely happily phrased because it first defines “insured person” by reference to five distinct categories and then does so in a more general way by defining it so as to include “persons employed … in the performance of professional services”. It seems to me that this captures the essence of those intended to be covered, and in my judgment clause 3.11(iii) should be construed compatibly with it. The natural inference is that if a person is employed to act for the company, is subject to control or supervision, and is remunerated by a salary, the definition will apply and he or she will fall within the scope of the policy. It is only if the purported worker is remunerated on a sale or commission basis that he or she will be excluded.

Quite apart from the clause itself, the relevant Jersey Regulations (which we were told are in this respect very similar to the position in the UK) describe a trust company business employee as “a person employed either under a contract of service or a contract for services by the registered person to assist in the provision of trust company business”: see the Trust Company Business (Registration) Jersey Order 2003. The judge held that the insurers should not be taken to have any special knowledge of Jersey legislation, but at the very least this demonstrates that a wide concept of employee is used for regulatory purposes and that it is a perfectly normal commercial use of the term. Indeed Rathbone plc itself described PEV as an employee even after he became a consultant, as the Recitals to the Rathbone indemnity demonstrate.

I have no doubt that the parties would have understood the definition in that way. There was no relevant change in the nature of the relationship between PEV and Rathbone Trustees when he retired save that he went part time and some of his benefits were changed. In substance he carried on his activities in the same way as he had done when employed under a contract of employment. The relationship continued to function as it had when PEV was an employee in the strict sense. In my view the language of the clause 3.11(iii), construed in the context of the commercial purpose of the contract, was apt to cover PEV even when he was employed under the consultancy agreement.

(d) Was PEV acting on behalf of Rathbone Trustees pursuant to a relevant agreement when acting as a personal trustee? For convenience I will repeat the definition of professional services set out in clause 3.21:

It is not disputed that the provision of personal trustee services is declared in the submission, but it is alleged that when acting as a personal trustee, PEV was neither acting “by or on behalf of” Rathbone Trustees, nor was he acting pursuant to a relevant agreement. It is said that the only agreements envisaged in the clause are those where an insured company is itself a party.

The excess insurers submit that the clause further envisages that the liability of the insured person will be co-terminous with the liability of his employer and that the phrase “by or on behalf of” is used in an agency sense. It is not enough that the insured person is acting at the behest of Rathbone Trustees; he must be doing something for the insured company directly so that the liabilities are incurred only when PEV is carrying out professional services for an insured company which that company itself has agreed to provide. Neither Rathbone Trustees nor its predecessors at any stage agreed to provide PEV’s services as a personal trustee. He is acting pursuant to an agreement which he presumably had entered into with the Walker Trust.

This submission is closely related to the argument I have already considered, namely that PEV was not, when acting as a personal trustee, acting under the direction or supervision of Rathbone Trustees. In my judgment, it suffers from the same basic flaw. The provision of personal trustees is, as the judge found, commonplace in Jersey. It is a significant part of Rathbone’s business to provide personal trustees and was (see para. 29 above) identified in the submission. The Rathbone Group is potentially liable for their activities. It was the justifiable understanding of Rathbone plc that employees exercising such functions would be covered by the policy, and in my view it would require very clear words indeed to exclude them. The concept of “by or on behalf of” should, if possible, be construed so as to achieve commercial common sense, and that in my view would mean bringing the activity of personal trustees within the cover if that is compatible with the language.

In my judgment, it is quite natural to speak of someone who is performing services at the request of an employer as doing so “on his behalf”. The words do not merely embrace a strict agency relationship as the insurers contend. So the Rathbone indemnity itself specifically recited that PEV “acts on behalf of [Rathbone Trustees] … as a trustee.” The fact that PEV had to account to Rathbone Trustees for money earned in that capacity further demonstrates, in my view, that in ordinary commercial language he was acting on behalf of Rathbone Trustees rather than on his own account. Moreover, as I have said, he could be removed from his office at their behest. In addition, PEV had entered into a restrictive covenant obliging him, inter alia, not to act as trustee for his clients for a period of two years after termination of his consultancy agreement. This is premised on the parties recognising that Rathbone Trustees have a commercial interest to protect and that the services he provides as a personal trustee are in pursuit of those interests.

This conclusion gains some support from clause 4.14 of the policy which excludes liability with respect to any claim arising out of an act or omission by an insured as trustee of a company’s pension, profit-sharing or employee benefits programme. The natural inference is that personal trustees of such trusts would have been liable save for this exemption.

As to the submission that there must be co-extensive liability, that is not what the policy says and I see no basis for inferring it. In any event, I am not sure what it means. In so far as Rathbone Trustees is now being sued as vicariously liable for the acts of PEV, there could be such co-extensive liability.

For these various reasons, I would reject this submission.

(e) Need the agreement be with an insured company? The other requirement of a professional service is that it should be performed “pursuant to an agreement with a third party”. The insurers say that it should be inferred that an insured company should be a party to the agreement. PEV was not appointed personal trustee of the Walker Trust as a result of any such agreement. He was appointed under the terms of the Trust deed to which neither Rathbone Trustees nor any predecessor was a party.

Like Burton J, I see no reason to read this limitation into the clause. No doubt in the vast majority of cases an agreement will have been entered into by the insured company and the insured person will be carrying out duties assigned to him in the performance of the contract. But I see no reason why clause 3.21 should be so restricted. Here the service is provided for the benefit of Rathbone Trustees and, as I have found, on its behalf. They are potentially liable if there is negligence by the personal trustee and will in any event wish to protect the person acting for their benefit. It is perfectly sensible for the policy to apply in such a situation. This is not a case where PEV was providing the services on his own behalf and for his own benefit. Even if the Trust had understood that to be the case - and the evidence before the judge alone was inconsistent with that inference - that would not be relevant when determining the scope of the policy.

If necessary, I would in any event have been willing to imply an agreement between Rathbone Trustees and the Trust that the former would allow PEV to act as personal trustee. It is true that under the Trust deed PEV could charge for his services and there was no provision to the effect that his employer could. In fact, however, at all times Rathbone Trustees, or its predecessors, received the fees from the Trust to the Trustees’ knowledge. Mr MacDonald Eggers QC, on behalf of the excess insurers, submitted that this was no more than a convenient arrangement for the Walker Trust to pay PEV. But from the beginning it was recognised that PEV was being appointed a personal trustee because of his links with what was then Nigel Harris and Partners. I have no doubt that the Walker Trust would have understood that he could be appointed by his employer. It is not, however, necessary to establish the existence of such a contract in order to conclude that the policy bites.

For all these reasons, therefore, I consider that the policy provides cover for the liabilities arising out of the alleged wrongdoing by PEV as personal trustee of the Walker Trust, assuming the legal claim against him to be successful.

Clause 5.14 of the policy provides as follows:

The insurers submit that this allows them to refuse to pay out on the policy until PEV has exhausted his claims under the Rathbone indemnity. Alternatively, they claim that there are certain other policies, Directors and Officers (“D and O”) policies, which also provide cover and must be exhausted before they are obliged to pay. The respondents submit in response that non-insurance indemnities are not caught by the clause; and if they are, they do not apply to indemnities given by one co-insured to another. As to the D and O policies, they contend that they simply do not apply to professional negligence claims of the kind being pursued against PEV in the Jersey litigation.

(a) Are non-insurance indemnities caught by the clause? The first question is whether this clause applies both to insurance policies and to non-insurance indemnities. As a matter of language it plainly covers both. If it were only intended to cover indemnification provided by insurance then one would not have expected the conjunction ‘and’ (or possibly the phrase ‘and indemnification’) to be found in the clause at all. The phrase naturally refers to two quite distinct categories of rights. Moreover, if the provision were applicable only to insurance policies then the phrase ‘available from any other source’ would appear to be redundant.

The respondents rely upon the fact that the clause is headed ‘Other Insurance’. They submit that this casts light on the proper construction and supports the contention that non-insurance indemnities were not intended to be covered. The difficulty with that submission is that clause 5.20(i) states in terms that “save where the context indicates otherwise, headings are descriptive only and not an aid to interpretation”. Mr Kendrick QC, on behalf of Rathbone plc and PEV, asserted that it is highly unusual to have non- insurance indemnities in an excess clause in a policy of this kind, and that this reflects a context which does indeed suggest otherwise. Even if the assertion was correct - and there was no evidence to sustain it - I do not accept that this would be part of the relevant context, sufficient to engage this sub-clause.

Mr Kendrick also referred to the excess clause in the first section of this policy. This relates to cover in criminal cases. The similar excess clause in that section relates to “valid and collectable insurance”. It is submitted that the two clauses would have been intended to have the same effect. Given the very different language of the two provisions, I do not accept that it is legitimate to draw such an inference. Clause 5.14 must be construed as it stands.

I therefore agree with the excess insurers and with the conclusion of Burton J that the heading cannot be used to cut back on the clear language used in the clause. It follows that the excess clause is in principle capable of applying not just to other insurance policies but also other forms of non-insurance indemnification.

(b) The meaning of “available from any other source”: Mr Kendrick’s other argument is that even if clause 5.14 can apply to indemnities, it is not appropriate that it should include indemnities given by one co-insured to another. The excess insurers submit that there is no justification for reading the language in a restricted way. It naturally includes all other forms of indemnity whether granted by a co-insured or not. The only form of indemnity not caught is that provided by the policy itself.

On this question, I accept the submissions made on behalf of Rathbone plc and PEV. It seems to me that if the insurers can take advantage of an indemnity given by one co-insured to another, this would significantly undermine the protection afforded by the policy. Employers frequently give indemnities to directors and employees for liabilities arising out of their negligent conduct. A major reason for taking out insurance is to protect against the risks of incurring liability as a consequence of such negligence. In my judgment, that would be obvious, both to the insurers and to the insured. It would frustrate the purpose of professional indemnity insurance to interpret the policy so as to exclude the insurers from liability in the very circumstances where that insurance is most likely to be needed. In my judgment, it would require very clear language to treat the indemnity granted by the insured company to be the primary source of cover ahead of the insurance for which the insured company has paid. The commercial understanding would be that the insurers receive the premium to meet precisely that kind of liability.

It would be clear that any other insurance cover would be from a source independent of Rathbone plc and the other insured companies. In my view it is reasonable to construe the clause so that the non-insured indemnification also needs to come from some external source. Again, therefore, I am in agreement with the judge below on this point.

Even if I am wrong about that, it is still necessary to show that the indemnity is available to PEV. That raises the question whether he can seek to enforce it where insurance protection is available. I consider that issue later in the judgment: paras. 82ff.

(c) The D & O policies: There is a separate and quite distinct submission by the excess insurers based upon the availability to PEV of £30 million cover under certain D and O policies. These policies cover PEV in his capacity as a personal trustee. It is accepted that in that capacity he falls within the definition of an “outside entity director”. The question is whether he could make a claim with respect to the loss resulting from the negligent performance of professional services.

The policy provides cover for what is termed “management liability”. Potentially that sounds broad but there is an express exclusion in clause 4.7 in relation to “professional services”. These are defined as including “alleged wrongful act in performance of or failure to perform professional services or related back-office supporting services”. On the face of it the exclusion plainly covers the liability in this case. That is hardly surprising; one would expect that the policies would cover different ground. They are both part of a suite of policies secured by Rathbone to cover relevant risks. The insurer under the D and O policies is AIG which has, in fact, declined cover on the grounds that they are not professional indemnity policies covering this risk.

Mr MacDonald Eggers QC submits that whatever Rathbone plc may have thought, in fact the two policies do cover the same ground, at least with respect to claims of this kind. He argues that whatever the precise scope of this exception, it could not have been intended to cover the performance of professional service by someone acting as a personal trustee because that is the only kind of liability that would in practice arise with respect to such persons.

Even if the premise were right, I do not accept it would be a justification for re-writing the plain words of the policy. But in any event, I do not accept that there would be nothing on which the policy could bite with respect to such persons. The boundary between management and professional services is potentially blurred and it seems to me that some acts connected with the activities of a personal trustee could be said to involve management services rather than professional services.

(d) Are defence costs covered in any event? Finally, there is a separate legal argument to the effect that even if the D & O policies did not cover professional negligence as such, they could nonetheless cover the cost of defending such claims. That seems to me a surprising submission. Liability policies will not habitually give a free-standing coverage for defence costs even where the liability itself is not insured, and in my view there would need to be very clear provision in the policy to that effect in order for the argument to succeed. This is supported by the decision of the Court of Appeal in Astra Zeneca Insurance Co v XL Insurance (Bermuda) Limited [2013] EWCA Civ 1660, reported at [2013] Lloyd’s Rep IR 290 para. 72 where Christopher Clarke LJ observed that in respect of non-marine liability insurance at least, the right to recover defence costs must, absent clear wording to the contrary, depend on some free-standing entitlement under the policy. That was a claim where no liability could be established but precisely the same principle must apply where no liability is covered. As Mr Kendrick pointed out, liability insurers generally have no interest in defeating claims for which they would not be liable under the policy.

Furthermore, clause 3.3(1) of the D and O policy provides that the cover will include “defence costs … resulting from a claim against an insured”. Such a claim must surely mean one in respect of liabilities which are covered under the policy.

Mr MacDonald Eggers relied upon two provisions in the policy, namely clauses 4.2 and 4.6 to support this submission. They provide exclusions for injury and property damage and US claims respectively, and they both include an exception to the exclusion so as to preserve the right to defence costs under the policy. The exclusion for professional services in 4.7 does not contain that exception. Far from assisting the excess insurers, this is a powerful pointer against their submission. Clauses 4.2 and 4.6 demonstrate that there needs to be an exclusion from the exemption to retain any right to defence costs, and since there is no such exclusion in clause 4.7, the inference must be that no such cover is provided.

(a) Introduction: The right of subrogation was described by Lord Bingham of Cornhill as follows in Caledonia North Sea Ltd v British Telecommunications plc (Scotland) and others [2002] UKHL 4; [2002] 1 Lloyd’s Law Reports 533…, para. 11:

A right of subrogation can be excluded in two ways. First, there may be a waiver of the right in the insurance policy itself. Second, even where the right of subrogation subsists under the insurance policy, the terms of an underlying contract between the insured and the third party may denude it of any substance and thus preclude its exercise. In this case the relevant underlying contract is the Rathbone indemnity. The terms of the underlying contract may be such that the insured cannot recover against the third party because on its proper construction, in the particular circumstances, the insured either has no right against the third party, or only has a right to the extent that there is a shortfall following receipt of the insurance moneys. In such cases the underlying contract envisages that the insurer bears the primary liability for the loss in question and the third party is, if liable at all, the secondary source of indemnity. Where that is the case, since the insurer cannot claim any greater rights than the insured, there is nothing to be enforced. This is not defeating the right to subrogation itself; rather it is leaving that right empty of any content in the particular case.

There is in fact a close interrelationship between the two routes, and in some contexts it may be obvious that subrogation should be denied but the precise reason may be difficult to formulate. This is demonstrated by the difficulty which the courts have had when explaining the general principle that insurers who have paid out to an insured for loss or damage cannot bring a subrogated claim against a co-insured who is himself insured in respect of the very same loss or damage. The formulation of this “co-insured” principle by Colman J in National Oilwell v Davy Offshore Ltd [1993] 2 Lloyd’s Rep 582 and Stone Vickers Ltd v Appledore Ferguson Shipbuilders Ltd [1991] 2 Lloyd’s Rep 288 focused on the insurance policy itself; there is an implied term excluding the right of subrogation. The formulation in the judgment of Brooke LJ in this Court in Co-operative Retail Services v Taylor Young Partnership Ltd [2000] EWCA Civ 207; [2001] Lloyd’s IR 122… paras 63-73 focused on an implied term in the underlying contract and rejected an earlier theory that subrogation was denied because it would create circuity of action. In the House of Lords in that case Lord Bingham and Lord Hope were minded to accept Brooke LJ’s analysis as the better explanation of the rule: see [2002] UKHL 17; [2002] 1 WLR 1419 paras. 7 and 63-65.

In this case Rathbone plc and PEV seek to ride both horses. They allege that there has been an implied waiver in the insurance policy, and that, in any event, on a proper construction of the underlying contract, the Rathbone indemnity, it is plain that the parties intended to treat the insurance as the primary indemnity. The Rathbone indemnity was to be only a secondary source of protection for PEV. Burton J rejected both these submissions and concluded that the excess insurers could exercise an effective right of subrogation. Rathbone plc and PEV appeal his ruling and submit that this was a most surprising conclusion which is fundamentally at odds with the commercial purpose underlying the policy and the indemnity arrangements. I will consider the policy and the underlying contract arguments separately.

(b) Is subrogation precluded by the terms of the insurance policy? Starting from the express words of the policy, the answer to this question appears clear. Clause 5.13 explicitly confirms a right of subrogation in the following terms:

Rathbone plc and PEV submit that, notwithstanding the language of the policy, the parties must have understood that the insurers would waive any rights they might otherwise have under the Rathbone indemnity. The obvious purpose of the policy was to insure members of the Rathbone group in respect of their liabilities to the outside world; it was to secure a wholesale transfer of risk from the Rathbone group onto the insurers. That objective would be nullified if the insurers could be subrogated to the rights of one co-insured against another. The policy was taken out for the benefit of Rathbone plc and those whom it insured; the insurers have received the premium and they should now bear the risk and cannot legally seek to transfer it to the very party who paid the premium.

Mr Kendrick accepts that in many circumstances, particularly negligence cases, insurance is properly seen as a last resort and the right of subrogation can be justified on policy grounds. As many authorities have observed, a tortfeasor ought not to be allowed to avoid liability by asserting that there is no loss on the grounds that it has been covered by insurance. Nor should he be allowed take the benefit of insurance for which he had not paid: see e.g Parry v Cleaver [1970] AC 1, 14 per Lord Reid. In that context insurance is properly treated as the secondary liability; it is res inter alios acta so far as the tortfeasor is concerned. There is a similar approach to liabilities incurred in contract, which will (see Caledonia v British Telecommunications plc [2002] UKHL 4; [2002] 1 Lloyd’s Rep 553 at paras. 13, 62, 89, 92 and 95) be primary while the liability of the indemnity insurer will be secondary.

Mr Kendrick contended strongly that this is not such a case. He submitted that Rathbone plc is not seeking to take advantage of insurance arrangements for which it has paid no premium. On the contrary, it has bought insurance to cover precisely the risk which may materialise (if the Jersey case is successful). It is contrary both to common sense and to the commercial realities to treat its indemnity as the primary source of liability and the insurance as the source of last resort.

Rathbone plc and PEV rely upon three features in particular which are said to make this commercial purpose of the insurance policy lucidly clear: the fact that Rathbone plc is the policyholder which acted on behalf of all other insured; that it paid the premium and was alone concerned with the administration of the policy; and that it was not at fault in any respect. It would be bizarre to treat the insurance payments as res inter alios acta as against the very company who sought the insurance and paid the premiums. The court has to stand back and objectively consider where, in all the circumstances, the risk should fall. Having determined that issue, the court can give effect to it by adopting the most appropriate contractual technique.

Mr Kendrick relied upon certain observations of Lord Hope of Craighead in Co-operative Retail Services Ltd v Taylor Young Partnership and others [2002]1 WLR 1419 para.65 where he explicitly endorsed a comment of Mr Recorder Jackson QC (as he then was) in Hopewell Project Management Ltd v Ewbank Preece Ltd [1998] 1 Lloyd’s Rep 448, 458 that the courts will imply a term if that would be necessary to avoid a “nonsensical” result that “could not possibly have been intended” by the parties. He submitted that if necessary this principle should be employed here.

The excess insurers submit that this whole argument displays a misplaced grasp of the underlying principles for construing contracts of this nature. Rathbone plc’s case depends upon an appeal to what it asserts is the purpose of the policy. The excess insurers submit that purpose cannot simply be identified in the abstract. It can only be discerned from a careful consideration of the policy terms as a whole read in their proper context. Insurers will determine premiums in the light of the extent of cover and that in turn depends upon how the parties have decided to allocate risk. Mr MacDonald Eggers relied on the following passage from the judgment of Aikens LJ in BMA Special Opportunity Hub Fund Ltd v African Minerals Finance Ltd [2013] EWCA Civ 416 para. 24 as a statement of the principles which should be applied:

Mr MacDonald Eggers submits that Rathbone plc is falling into the trap identified by Briggs J and Aikens LJ. The court should not conclude that the parties were intending to exclude the right to subrogation with respect to the indemnity if, as the insurers claim, the parties specifically addressed the question of exclusion and chose not to make the exclusion now relied upon. The supposed commercial purpose cannot override the clear and unambiguous terms of the policy.

The excess insurers also say that Rathbone plc and PEV have exaggerated the significance of Rathbone plc in the insurance arrangements. They maintain that it is false to say that Rathbone plc can properly be described as the policyholder quoad them. That is the way Rathbone plc is described in the Primary Policy with AIG - although even then the definition of policyholder embraces not only Rathbone but also its subsidiary companies. But it is not the way Rathbone plc is described in the Excess Policy. That policy refers to the “insured”; it provides in terms that “in consideration of the insured having paid or agreed to pay the premium …… insurers and insured agree as follows …”. The insured includes all persons and organisations identified at Item 1 in the Schedule, which in turn refers to Rathbone plc and its subsidiary companies and all other insured persons and organisations. They are all liable to pay the premium. The excess policy is not, therefore, a composite policy entered into solely by Rathbone plc; rather it is a series of contracts between the insurer and each insured as a principal, with Rathbone plc acting as their agent.

Mr MacDonald Eggers adds that there is in any case no different set of principles which apply depending upon the centrality of the status of the insured. If the policy confers a right of indemnity against a co-insured, there is no justification in principle or authority for conferring upon Rathbone plc some special privilege or protection. Rathbone plc does not fall within the clear language of the exemption created by clause 5.13 and there is no basis for implying any such additional exemption. It would contradict the right of subrogation which is conferred under clause 5.13 with respect to “all insureds’ rights of recovery, contribution and indemnity.”

Furthermore, he relies upon a dictum of Rix LJ in Tyco Fire & Integrated Solutions (UK) Ltd v Rolls-Royce Motor Cars Ltd [2008] EWCA Civ 286; [2008] Lloyd's Rep IR 617 para.77 where Rix LJ expressed the view, obiter, that in circumstances where there is nothing in the underlying contract preventing one co-insured from suing another, a right of subrogation may be lost only by way of an express waiver in the policy itself:

Mr MacDonald Eggers observes that the insurance here is a bundle of composite insurance policies. He further contends that even if there is room in principle for an implied term, there is no basis for any implication here. It would be wholly inconsistent with an express term in clause 5.13, set out in paragraph 70 above, which must be taken as exhaustive of the circumstances in which the insurers have agreed to waive any right to subrogation.

I agree that if one focuses solely upon clause 5.13 of the policy, there is plainly no waiver with respect to the Radcliffe indemnity. The insurers have excluded the right in certain very specific circumstances but only where the potential defendant in the subrogated action is an insured person; there is no exemption for insured companies such as Rathbone plc.

If the obiter observations of Rix LJ in Tyco are correct and of universal application, there is no possibility of any implied term. But there cannot be a rule of law to that effect and Rix LJ was not intending to suggest that there could be since he himself said in Tyco (para. 80) that the scope of the right “is all ultimately a matter of the parties’ intentions found in their contracts.” That must be as true for the insurance policy as for the underlying contract. Moreover, Rix LJ was considering a potential subrogation claim against a party to the underlying contract whose negligence had caused the actual loss and who would be liable under that contract to a co-insured. In such cases it is difficult to see why, absent an express term, the insurer should be taken to have waived the right of subrogation. But this is not an analogous case. Far from causing the loss, Rathbone plc, the potential defendant in the subrogated claim, has simply provided PEV with an indemnity which in large part overlapped with and reinforced the protection which PEV had secured by way of insurance in his consultancy contract.

In determining whether a term should be implied the court must construe the contract in its commercial setting. Lord Hoffmann expressed the position as follows when delivering the judgment of the Privy Council in Attorney General of Belize v Belize Telecom Ltd. [2009] UKPC 10; [2009] 1 WLR 1988 para. 21:

Applying that test, in my judgment, the court should in this case imply a term that the insurers will not seek to be subrogated to PEV’s rights against Rathbone plc under the Rathbone indemnity. In reaching that conclusion I fully recognise that the question how risks have been allocated between insurers and insured has to be construed by reference to the terms which the parties have agreed. It is not for the court to decide how risks ought to have been allocated, only to consider how the parties in fact chose to allocate them. I am also alive to the risk of the court moulding the policy to reflect its own sense of commercial reality. These sound strong notes of caution. Nevertheless, I am satisfied that it could not have been the intention of the parties that the insurers should be able to enforce rights of indemnity against a co-insured where the co-insured was indemnifying the very same risk as the insurers. I believe that implying the term is simply making express what the parties must have intended.

For reasons I have already given when considering the question of excess and the meaning of “from any other source” in paragraph 5.14 of the policy (see paras. 56 -57 above), in my judgment it would seriously undermine the purpose of the policy. It would deny Rathbone plc the very benefit which the policy was intended to confer on Rathbone plc to treat the Rathbone indemnity as the primary source of protection for PEV with the insurance being the secondary source. It would defeat the purpose for which they had paid the premiums. I do not believe that the parties can reasonably be taken to have intended to give priority to the Rathbone indemnity in that way. It was pure happenstance that the indemnity was in place. In my judgment if Rathbone plc had asked in advance whether they could enter into the indemnity with PEV without jeopardising their right to recover under the insurance policy, they would certainly have been told that they could.

In reaching this conclusion I have gained no assistance from two other arguments advanced by Mr Kendrick in support of his submissions on this point. First, he invoked clause 4.7 of the policy, which is in these terms:

Mr Kendrick says that this clause demonstrates that, properly construed, the policy was contemplating that the insurers were not to be concerned with matters internal to the Rathbone group. I do not agree; that clause seems to me to be neutral on the point in issue. In broad terms that clause excludes liability under the policy where the negligence of one insured within the Rathbone group of companies causes loss to another. In so far as there may be loss suffered internally as a result of professional negligence, therefore, the Rathbone group must, through one company or another, bear that loss. It does not follow that where loss is suffered by an external third party the insurers are impliedly undertaking not to exercise a right they may have to hold an insured liable for the loss caused by another co-insured.

Second, Mr Kendrick sought to bring this case within the co-insured principle referred to in paragraph 69 above. In my judgment, that principle is not applicable here since it depends upon the defendant in any subrogated action being insured for the same loss or damage as the claimant in whose name the action is brought. That is true of PEV’s employer, Rathbone Trustees, as the insurers accept. But that is not the position with respect to Rathbone plc who have no liability with respect to any loss suffered by the beneficiaries of the Walker Trust. Nor indeed are they insured with respect to any contractual payments made to PEV since liabilities in contract are specifically excluded by clause 4.3 of the policy. In any event the better view is that the principle is explained by an implied term in the underlying contract rather than the insurance policy itself.

(c) Is the exercise of a right of subrogation precluded by the terms of the underlying contract, the Rathbone Indemnity? Even if I am wrong about the implied policy waiver, and there is a right of subrogation against Rathbone plc with respect to the indemnity, the question arises whether under the terms of the Rathbone indemnity itself it was intended by the parties that the indemnity should be treated as the primary source of protection.

The effect of the Rathbone indemnity and the terms of his consultancy agreement with Rathbone Trustees was that PEV had two quite distinct sources of funding available to meet any liability he might incur as a result of professional negligence. Naturally he cannot seek to recover more than a full indemnity by claiming on both the insurance and the indemnity. If the party who has the primary liability to pay the loss pays him out in full, that discharges the obligation of any other indemnifier. If, on the other hand, the paying party’s liability is secondary then the payment is res inter alios acta and does not discharge the liability of the primary party. Subrogation allows the paying party to recover from the primary source of indemnity.

We were referred to a number of cases in which a tortfeasor has been held entitled to take advantage of insurance cover even where he has not directly paid the premium. Mr Kendrick placed particular emphasis on the decision in Mark Rowlands v Berni Inns Limited [1986] 1 QB 211. A landlord insured premises which were destroyed by the negligence of the tenant resulting in a fire. There was fire insurance and the premiums were paid out of rental income from the tenant. The court held that the insurance policy was intended to enure to the benefit of the tenant and that the parties’ intention - that is, the landlord and tenant - sensibly construed was that any loss suffered by the landlord would be recouped from the insurance monies and that this would discharge the tenant’s obligations. Lord Justice Kerr (with whose judgment Croom Johnson and Glidewell LJJ agreed) said this (page 233):

The insurance money was not res inter alios acta so far as the tenant was concerned. The terms of the tenancy agreement were that the property would be rebuilt from the proceeds of the insurance claim and accordingly there was no claim in negligence which could be brought against the tenant.

Kerr LJ reached this conclusion by an analysis of the contract but he also accepted that it was complemented by considerations of “justice, reasonableness and public policy”. Lord Reid had used this phrase in Parry v Cleaver [1970] AC 1, 13 when identifying considerations which might justify refusing insurers a right of subrogation.

Mark Rowlands is not directly applicable here. First, in that case the subrogated right would have been exercised against a tortfeasor whereas here it would be exercised on behalf of the tortfeasor against an innocent party, Rathbone plc. Second, in that case the party seeking to take advantage of the policy had not paid the premium, whereas in this case it has. However, whilst these differences prevent any direct analogy between the two cases, they are in my view factors which militate in favour of Rathbone plc’s case rather than against it.

We were referred to a number of other cases where because of the particular structure of the underlying contract the court has held that the primary indemnifier was the insurer and there was no right of subrogation: e.g Co-operative Retail Services v Taylor Young Partnership Ltd [2002] UKHL 17; [2002] 1 WLR 1419 and Scottish and Newcastle plc v GD Constructions Ltd. [2003] EWCA Civ 16; [2003] Lloyd’s Rep IR 809. In other cases the court has held that the right of subrogation was not lost, that the wrongdoer should bear the primary liability and that insurance should be treated as the secondary source of indemnity and is therefore res acta inter alios as against the defendant; e.g. “The Yasin” [1979] 2 Lloyd’s Rep 45; Caledonia v British Telecommunications plc [2002] UKHL 4; [2002] 1 Lloyd’s Rep 553; and Trustees of the Tate Gallery v Duffy Construction [2007] EWHC 361 (TCC). However, none of those cases were concerned with the particular circumstances arising in this case; they each turn on the particular contractual arrangements and therefore provide little in the way of assistance. Save for the Caledonia case, they were cases where, as with the Mark Rowlands case, the defendant in any subrogated action was or would have been the party whose negligence caused the loss. None of these cases raises the question whether the insurer can be subrogated to a claim arising out of an indemnity covering the same loss as the policy.

Mr MacDonald Eggers relies upon on two broad principles derived from the authorities which, he submits, point strongly in favour of insurance being the secondary form of indemnity.

First, he submits that this is the general rule. The usual position in a case of this kind involving contractual rights is that insurance will bear only the secondary liability. He relies in particular upon certain observations of Lord Mackay in the Caledonia case at para. 62 when, having considered various authorities, he commented that:

I do not think that this supports the proposition being advanced. In my judgment Lord Mackay had in mind cases where the relevant loss would have been caused by a defendant acting in breach of contract. In such cases there is every reason to make the wrongdoer bear the primary liability and to treat the insurance proceeds as res inter alios acta for reasons already discussed. But Rathbone plc is not a wrongdoer; together with the insurers it is seeking to protect the wrongdoer and thereby to reinforce (and in fact extend) the security which the insurance provides. A mantra that insurance is a last resort does not, in my judgment, assist in this kind of case. In any event, as Lord Reid said in Parry v Cleaver, considerations of “justice, reasonableness and public policy” may require exceptions.

Second, the excess insurers emphasise the fact that there is no link of any kind between the policy and the indemnity. Although Rathbone Trustees have expressly undertaken in the consultancy agreements (although not the earlier employment contracts) that they will insure PEV against liability for professional negligence, Rathbone plc itself has not done so. Nor does the Rathbone indemnity state that PEV should first look to his insurance before having recourse to the indemnity. The insurers refer in this context to the judgment of Longmore LJ in Scottish and Newcastle case. The Court of Appeal held that as a matter of construction of a contract between the main contractor and a sub contractor involved in refurbishing a country house pub, the main contractor was obliged to take out insurance to protect the sub-contractor from liability for causing loss by fire, even where the fire was negligently caused. His Lordship said this (para. 57):

I do not think that this observation materially assists the insurers. No doubt where there is an express clause linking the exclusion of liability with an obligation to insure, it can readily be seen that insurance is intended to be the primary source of indemnity. But as the passage from Longmore LJ’s judgment also makes plain, this might still be the appropriate inference to draw from the facts even where there is no explicit link.

In my judgment, the relevant question is to ask whether PEV would naturally have understood that his claim under the indemnity had been exhausted once his liability had been fully met under the insurance policy. Was the insurance the primary source of indemnification thereby discharging any obligation which Rathbone plc might otherwise have under the Rathbone indemnity?

In my judgment, it was. It seems to me that in substance Rathbone plc has made available two ways in which any professional negligence liability of PEV may be met. It is a matter of indifference to PEV how he is protected, provided that one way or the other there is the necessary indemnity. They are not co-extensive and there may be circumstances where the insurance cover does not meet the full liability and recourse would then be had to the indemnity. But to the extent that the liability is discharged by the insurance monies, in my view Rathbone plc can take advantage of that payment and treat it as discharging pro tanto its own obligations.

In my view, one can readily imply a term into the Rathbone indemnity contract to the effect that it is intended to provide supplemental protection only once the claim against the insurance company had been exhausted. It makes little sense to treat the insurance payments secured by a premium paid by Rathbone plc itself as being res inter alios acta so far as Rathbone plc is concerned. I should add that even if, as Mr MacDonald Eggers submits, it may be said that PEV is indirectly paying for the premium so far as the policy relates to him, I would reach the same conclusion. It is still not a policy which he voluntarily obtained for himself, in which case one could see that it ought to have no bearing on his relationship with Rathbone as was the position in the Caledonia case.

I do not believe that the fact that there is no extrinsic link between the insurance and the indemnity should preclude implying the term. All the parties were fully aware that insurance protection had been provided and this was the context in which the Rathbone indemnity had been agreed. A term in the indemnity to the effect that insurance was intended to be the primary liability would have been conclusive, but I do not consider that the lack of it demonstrates that there was no such intention.

Burton J found against the respondents on this ground because he thought that the logic of the argument was that “there can never be a subrogated claim brought against a policyholder who has paid the premium in respect of coverage of a loss for which the policyholder was not insured.” I respectfully disagree. It is not the mere fact that Rathbone plc has paid the premium which justifies treating insurance as the primary liability. It is the fact that Rathbone plc is not at fault and that the subrogated right relates to an indemnity which is providing the same protection as the insurance itself.

Accordingly, even if there is a right of subrogation in principle which is not waived by the policy itself, in my view that right has no substance in the circumstances of this case. There is no right to which the insurers can be subrogated.

It follows that I would uphold the appeal with respect to subrogation but dismiss the other grounds.

There are two further points in dispute. First, the insurers contended that in accordance with the terms of the policy it is not necessary for them to have paid out under the policy before exercising the right of subrogation. In view of my conclusion that there is no effective right of subrogation, the point does not arise. In the circumstances suffice it to say that I agree with conclusion and reasoning of Burton J on the point.

The second point did not arise before the judge. It crystallised as a result of his conclusion that subrogation was available to the insurers. It is concerned with a situation - in practice highly unlikely to arise - where PEV’s liability is greater than the sums indemnified in the Rathbone indemnity. The issue is whether the insurers can only claim to be reimbursed once PEV’s loss has been fully indemnified or whether they cover the same layer of loss as the indemnity and can therefore seek to recover under the policy even if PEV has an outstanding liability. Given my finding on subrogation, the point no longer arises and in the circumstances, I propose to say no more about it.

I would therefore dismiss the appeals on all matters save the issue of subrogation. In my judgment, the insurers cannot exercise any right of subrogation so as to be indemnified out of the Rathbone indemnity.

When a draft of this judgment had been almost completed, the court was told that there was a proposed settlement which might render this judgment irrelevant as far as the parties are concerned. We have since been told that the matter has in fact been settled. Nonetheless, we have decided that it is appropriate to give judgment. The respondents seek a judgment and they say - and we agree - that the case raises issues of some general interest in the insurance world.

I am grateful to Lord Justice Elias for his comprehensive description of the facts and the issues on this appeal. I agree with his conclusion and his reasons for concluding that PEV is able to recover under the policy and that the excess insurers are not entitled to rely upon the excess clause. Accordingly, I agree that the excess insurers’ appeal against Burton J’s order to this effect should be dismissed.

I also agree that, in the circumstances of this case, the excess insurers do not have an effective right to be subrogated to PEV’s contractual right to the indemnity he has against Rathbone plc and Rathbone Trustees under the “Rathbone indemnity”. I therefore agree with Elias LJ that the appeal by Rathbone plc and PEV against Burton J’s order that they do have such a right should be allowed.

Elias LJ identified two routes to his conclusion about subrogation. The first is based on an implied exclusion of the right to subrogation in the contract of insurance itself, here the professional indemnity policy issued to Rathbone plc by the excess insurers. The second is based on the terms of the underlying contract, in this case the “Rathbone indemnity” made in July 2003 between Rathbone plc, Rathbone Trustees, and PEV. I respectfully agree with the second route to this conclusion. In the present case, for the reasons given by my Lord, any right of subrogation under the insurance policy is denuded of any substance by the terms of the “Rathbone indemnity”, which has the effect that PEV will have no claim under it in respect of sums received from the excess insurers. The result is that, while the excess insurers will, in principle, have a right to subrogation, the terms of the underlying contract mean there is no claim to which the excess insurers’ right to subrogation can attach. That right is thus left empty of any content.

As to the first of my Lord’s routes, notwithstanding the force of Mr Kendrick’s submissions and of my Lord’s analysis, including what he has said about the relationship between the two routes to the exclusion of subrogation, in the circumstances of this case I have difficulty in regarding the insurance policy itself as excluding the excess insurers’ right to subrogation.

I agree that, in Tyco’s case, Rix LJ was not intending to suggest that there is a rule of law that only an express exclusion of subrogation in the insurance policy will suffice. But a number of factors suggest that, in the light of the nature of an insurance contract and the nature of the right which it is sought to exclude, his words may reflect what should be the position in practice in all but an exceptional case. There is some analogy to the approach that has been taken to the effect of breach of a “fundamental term” in a contract or the consequences of a “fundamental breach” since the House of Lords, in Photo Production Ltd v Securicor Transport Ltd [1980] AC 827, laid to rest the view that no exemption clause could prevail against such breaches and replaced it with a rule of construction. It remains the case that parties to a contract are less likely to be taken to have agreed that one of them shall be excused in the case of a total non-performance or a performance which is wholly at variance with the object of the contract, but there is no rule to this effect. It is possible to exclude such liability but very clear language is needed.

In the case of a contract of insurance the purpose of the contract is to provide indemnity, and one of the means by which effect is given to the principle of indemnity embodied in it is the doctrine of subrogation. In Lord Napier and Ettrick v RF Kershaw Ltd (No. 1) [1993] AC 713 at 743 – 744 Lord Goff stated that the right of “subrogation arises from the very nature of the contract itself”, and see Lord Templeman at 738B. In Banque Financière de la Cité SA v Parc (Battersea) Ltd [1999] 1 AC 221 at 223, Lord Hoffmann stated that “the doctrine of subrogation in insurance…gives effect to the principle of indemnity embodied in the contract”. Subrogation is thus an important means of ensuring that no more than indemnity is provided to the assured.

There are many statements in the cases that clear words are needed to exclude a right of subrogation: see, for example, Liberty Mutual Insurance Co UK v HSBC Bank plc [2002] EWCA Civ 691 at [49] – [50] and [56] per Rix LJ, albeit in the context of suretyship, not insurance. The cases cited by Rix LJ for that proposition are, however, from a number of contexts. In this case, the policy itself did not contain clear words of exclusion. Indeed, clause 5.13 expressly provides that the insurer “shall be subrogated to all insureds’ rights of recovery, contribution and indemnity …” and the words in the latter part of the clause carving out an exception do not cover the circumstances in this case. The contract of insurance was entered into by experienced commercial parties who are, or must be regarded as, being entirely familiar with the concepts at issue. It is true that sometimes courts can conclude that words used by such parties do not have their ordinary meaning. For a particularly striking example concerning the meaning of the word “condition”, see LG Schuler AG v Wickman Machine Tool Sales Ltd [1974] AC 235. But great care must be taken in doing this.

Where, as in this case, experienced commercial parties have addressed the issue of subrogation and provided for it in a particular way, but did not choose to exclude the right to subrogation in the circumstances which have arisen, the court should be very slow to conclude that they were intending to do so. As Lord Hoffmann stated in Chartbrook Ltd v Persimmon Homes Ltd [2009] UKHL 38, reported at [2009] 1 AC 1101 at [15], “it clearly requires a strong case to persuade the court that something must have gone wrong with the language” in order to justify a meaning which departs from the words actually used. This is a very different case from National Oil Well (UK) Ltd v Davy Offshore Ltd [1993] 2 Lloyd’s Rep 582, where Colman J was considering an insurance policy with a clause which provided for waiver against “any assured and any person …whose interests are covered by this policy”. Those are very different words to those in clause 5.13.

The majority of the recent cases in this area now proceed on the basis of the construction of the underlying contract as having the effect that there is no claim to which the insurer’s right to subrogation can attach: see Co-operative Retail Services v Taylor Young Partnership [2000] EWCA Civ 207, reported at [2001] Lloyd’s Rep IR 122 at [69] per Brooke LJ and [2002] UKHL 17, reported at [2002] 1 WLR 1419 at [65] per Lord Hope. In Tyco’s case, Rix LJ ([2008] EWCA Civ 286, reported at [2008] Lloyd’s Rep IR 617 at [75]) described this as “a difficult and controversial area” and adopted Lord Hope’s view in the Co-operative case that “the true basis of the rule is to be found in the contract between the parties”, i.e. the underlying contract. In the circumstances of this case, because of the very clear position in the underlying contract it does not appear necessary to imply additional provisions into the insurance policy itself to ensure what Elias LJ has stated is the commercial purpose of the policy.

I agree with the conclusion of my Lords as to the outcome of this appeal.

As Lord Justice Beatson has said, the difference between my Lords is whether there are two routes or only one to the conclusion about subrogation. I agree with Lord Justice Elias that there are the two routes he identifies. The proposition that it was part of the commercial bargain between the excess insurers and Rathbone, that the excess insurers could receive a premium from Rathbone for insuring PEV’s excess, could pay the excess to PEV, but could then put themselves into PEV’s shoes to claim the excess from Rathbone under the Rathbone indemnity struck me as a surprising one during the course of argument and it still does. The obvious purpose of the policy was to transfer risk outside the Rathbones “group” and on to insurers, who received a premium for accepting that risk. I would add that this conclusion seems to me to sit more comfortably with the twin objectives of subrogation which are to ensure that the insured receives no more than a fair indemnity and that the cost of the loss should be borne by the responsible person (or more probably their insurers).

I agree with the cautionary observations made by both Lord Justice Elias and Lord Justice Beatson about the implication of terms into a contract between experienced commercial parties. Nonetheless, applying the test set out by Lord Hoffmann at paragraph 21 in Attorney General of Belize, for the reasons given by Lord Justice Elias, with which I agree, in my view, the court should imply a term into the policy that the insurers will not seek to be subrogated to PEV’s rights against Rathbone plc under the Rathbone indemnity.