UK: Brokers´ liability - assuming the worst

Last month saw the publication of the Commercial Court’s judgment in BP v Aon, one of the most substantial pieces of litigation against an insurance broker in recent years. The case provides important guidance on the broking process and the utilisation of open covers, in both the London and overseas markets, but is perhaps of most interest for its exploration of a broker’s liability to an insured in tort, despite the absence of any contractual relationship.

Last month saw the publication of the Commercial Court’s judgment in BP v Aon, one of the most substantial pieces of litigation against an insurance broker in recent years. The case provides important guidance on the broking process and the utilisation of open covers, in both the London and overseas markets, but is perhaps of most interest for its exploration of a broker’s liability to an insured in tort, despite the absence of any contractual relationship.

The full heading of the case – BP Plc v Aon Limited and Aon Risk Services of Texas Inc [2006] EWHC 424 (Comm) – reveals that there were two Defendants, referred to in the judgment as Aon London and Aon Texas. Aon Texas had contracted to place insurances for a number of construction projects involving BP and its joint venture partners. In practice, the placing of the various insurances was carried out by both Aon Texas and Aon London. The insurances were declared to an open cover which was placed and administered partly by Aon Texas and partly by Aon London. The instructions were initially received by Aon Texas from the Amoco arm of BP Amoco in the US, but subsequently BP in London started to communicate with Aon London direct.

The only written contract in place was a Service Agreement between BP and Aon Texas. That agreement limited Aon Texas’ liability in respect of certain potential claims, but did not mention Aon London. The Aon companies declared risks to the open cover on the basis that only the two leading underwriters needed to be notified of each declaration. In previous litigation, this had been held to be a breach of the terms of the open cover requiring declarations to be notified to following underwriters as well as the two leaders. Those underwriters who had not been notified had been held not to be on risk. BP’s claim against Aon was seeking to make up the shortfall.

Colman J applied a series of cases holding that a sub-agent can be directly liable to the principal in tort, despite the absence of any contract, provided there has been an "assumption of responsibility". On the facts of the case, he found that Aon London had assumed responsibility to BP, even though they were not BP’s direct agent. This concept of assumed responsibility gives rise to considerable uncertainty. In a market were the use of more than one broker is not infrequent, but the relationships between the various parties are rarely documented, it will be very difficult to determine whether a broker has crossed the line of assuming responsibility direct to the insured. A finding of liability in tort in such a case circumvents the need for any contractual or agency relationship, which might reasonably be considered to be a sensible threshold for imposing a duty of care.

The length of the judgment is indicative of the complexity of the factual background, but to understand the effect of the decision it is useful to be aware of some of the history of the relationship.

The open cover giving rise to the dispute was originally placed on behalf of Amoco, which was headquartered in the US and instructed Aon Texas. Amoco was aware that Aon London were involved but there was no real direct contact. The cover was placed by Aon Texas with one of the leads, AIG, and a few of the following markets. Aon London placed with the second lead, Swiss Re, and more of the following markets.

The placing of the open cover had started in late 1998. By the end of 1998, a merger between BP and Amoco had been announced, and with the agreement of all underwriters the cover was extended to include BP projects as well as Amoco projects. Initially, instructions from BP were channelled through the US contacts, but over time there came to be more and more direct communication between BP in London and Aon London.

The key feature of the open cover was that rates were agreed at the time the cover was placed in 1998. The cover was for a period up to 30 June 2000, and towards the end of this period it was becoming clear that the rates at which risks could be attached under the cover was considerably more attractive than the open market rates prevailing at the time. The nature of the cover also made the placing process considerably less time-consuming than an open market placement. This led to a flurry of declarations at the end of the period of the cover, including a number of declarations on 30 June 2000 itself.

Several of these declarations were notified only to the leaders. In earlier proceedings between BP and the following markets, it was held that notification to the leaders was insufficient to bind the following market and that declarations were therefore ineffective against those following underwriters. The leading underwriter clause failed to deal with the declaration process and referred only to the process for agreeing endorsements, and for claims handling. Aon London had inserted this clause in the cover, and loss had been suffered when risks were declared without being notified to the followers.

A strange feature of the litigation is that it was never argued that the following market actually needed to agree any of the declarations. All that would have been necessary for the declarations to have been binding on the following market would have been notification. If notified of a declaration, the following market would not have had the right to decline that risk. Nor would they have needed to take any positive step to consent to a declaration. It seems harsh that Aon London end up picking up the tab for what was essentially an administrative failure when the following underwriters, who were willing to be bound to declarations without any individual rating or assessment of those risks, were able to walk away because of that administrative failing. This does, however, emphasise the need for clarity in drafting such provisions. An open cover is a contract to insure, rather than a contract of insurance, and is not strictly governed by the market’s contract certainty initiative, but need for certainty is, if anything, even more acute where a facility is being granted to which numerous risks may be declared.

The supposed shortcomings which were the subject of the proceedings were in relation to matters handled by Aon London. Aon argued that there was a contractual chain, with only Aon Texas being in a direct contractual relationship with BP, and that there should be no tortious liability outside of that contractual relationship.

In analysing this issue, Colman J undertook an extensive review of the authorities, most notably Henderson v Merrett [1995] 2 AC 145, in which it had been held by the House of Lords that Lloyd’s Managing Agents owed duties in tort to so-called "indirect names" as well as a contractual duty to "direct names". Lord Goff’s speech in Henderson propounded the theory that for a tortious liability to arise there must be an assumption of responsibility.

Curiously, Colman J deals with a case that might be thought of as potentially the most direct relevant – the Court of Appeal’s decision in Punjab v De Boinville [1992] 1 Lloyd’s Rep 7 – almost as an afterthought. In Punjab individual employees of the broker, as well as the corporate entity, were held to owe a duty of care to the insured. Punjab can itself be viewed as a harsh decision, but is of course binding precedent on a lower Court unless distinguishable. Colman J says that he cannot see how that situation can be different from the broker/sub-broker relationship. It may be arguable, however, that there is a clear distinction between the position of employees of the contracting company (the direct agent) and the position of a completely separate company.

Colman J does distinguish the position in the construction industry where there is typically a chain of written contracts down the sub-contracting line, each specifying responsibility for sub-contractors actions. There are a series of cases which state that in that situation only the head contractor is responsible to the client. It may be questionable, however, whether the distinction is as clear-cut as Colman J suggests. The inter-relationships between the different parties in the broking context may in fact be quite similar to those in the construction industry, the difference being that in broking the contractual relationships are not usually committed to paper. That is not to say, however, that the same contractual duties might not arise, if only impliedly. The Judge does not analyse the true contractual position in any great detail.

Having reached the conclusion on the law that there was scope for a duty in tort based on the assumption of responsibility, Colman J analyses the facts in considerable depth to determine whether there was in fact such an assumption of responsibility. Whilst the Judge’s analysis is very thorough and careful, it may be argued that the necessity to undertake such an exercise (with considerable benefit of hindsight) is indicative of the difficulties that are likely to arise if the law continues to develop in the same direction. Both brokers and their clients need to know the potential consequences if errors occur. A simple test that would allow both parties to know where they stand would be preferable to an approach that requires a lengthy Commercial Court trial to know whether a duty has arisen.

Colman J’s judgment also includes lengthy examination of some of the quantum and causation issues. Many of these points were case-specific, but potentially of wider application is Colman J’s approach to the reasonableness of BP’s settlement with certain of the insurers. The Judge essentially holds that the brokers, in the second round of litigation, cannot seek to re-open arguments which they say ought to have been pursued more aggressively in the first round of litigation, provided the overall settlement was broadly reasonable. Brokers will need to be aware that a consequence of their traditional wariness to become directly involved in disputes between (re)insureds and (re)insurers may lead to them being fixed with the outcome of such disputes.

This article was written for Law-Now, CMS Cameron McKenna's free online information service. To register for Law-Now, please go to www.law-now.com/law-now/mondaq

The original publication date for this article was 12/04/2006.