Speech by Julian Adams, Head of Wholesale Insurance Firms, FSA
The Geneva Association International Insurance and Finance Seminar
7 December 2006

Good morning. I’d like to start by thanking the Geneva Association for giving me the opportunity today to speak on what is an increasingly important topic in the insurance arena.

The subject of how firms can use their capital in the most efficient manner is of course one of the most pressing concerns facing the insurance industry as a whole after decades of generating relatively poor returns for shareholders. In light of this, it’s no surprise that much management time and effort is being spent addressing the question of how capital is allocated and employed to maximise returns within a pre-defined risk appetite.

For any financial group, and particularly for an insurance group, key questions within any capital management strategy relate to how much risk should be retained, and the methods that should be used to transfer that which is not. Traditionally, of course, the main means of risk transfer for insurers has been through the purchase of reinsurance, but perhaps that route will become more constrained. As the number of highly-rated reinsurers contracts, whilst the quantum and nature of risk to be transferred continues to grow, insurance companies are naturally starting to explore ways in which they can tap directly and more readily into the capital markets.

Against this backdrop, what I’d like to do with my time with you this morning is to cover two main topics. In both cases, as well as looking at what the situation is now, I’d like to look ahead to how the environment is changing in a European context, both in terms of the Reinsurance Directive and developments in Solvency II.

First, I want to explore what capital efficiency when pursued at a group-wide level may mean for national regulators – and indeed for the entities they supervise – offering a possible solution but also looking at some of the practical problems which would need to be addressed in order to bring it about.

Second, I want to reflect on the undoubted expansion of interest in insurance risk from the capital markets. As part of this I’d like to say what we’ve already done in the UK to accommodate some of this, but also outline a number of questions that the market needs to address from a regulatory perspective if this is to continue to develop.

At the end, I’d also like to touch on some other ways in which we’re trying to make life a bit easier for the insurance market in the short term, with the hope of consolidating London’s appeal as an insurance centre.

Supervision of groups

There are essentially two different ways of looking at insurance groups. The first, and possibly more progressive, regards an insurance group as a single economic entity in which risk can be pooled and diversified. This is true of insurance groups in general, and arguably even more so of reinsurance groups, whose business models and value proposition are based on diversification of risk, both globally and sectorally.

The second, and more traditional, approach sees the group as a collection of legal entities with segregated risks. From a management perspective, this may seem a little antiquated but the reality is that each insurance policy is the liability of a specific legal entity. Hence, most regulation still bites at the legal entity level and bankruptcy and insolvency law tends to work on this basis too. If regulation of insurance groups continues to focus on greater levels of capital being held locally, there is then a risk that the diversification benefit that legitimately exists at group level may be ignored or at best unravelled. This would force a fragmentation of a group's capital which in turn could translate into higher prices for policyholders and a constraint on capacity without necessarily providing greater protection for policyholders.

Last month the UK Treasury and the FSA jointly published a discussion paper which reflected on how group supervision might be undertaken in the context of Solvency II. In doing so, we were attempting to stimulate a debate and a wider consideration of the issues posed by group supervision in the insurance sector. We were also seeking to set out the UK’s stall on this issue, bearing in mind that our views are not necessarily shared by other supervisors across the EU.

There are two elements to what is proposed in the paper: first, that groups should be allowed greater freedom to allocate their capital resources between different subsidiaries within the EU; and second, there is a discussion on the relative responsibilities of the supervisors of groups and their subsidiaries.

The essence of the capital point is this: we are proposing that subsidiaries within a group should not be required locally to hold capital in excess of their Minimum Capital Requirement. Rather, capital in excess of the MCR and any Pillar 2 requirements may be held at group level, for the benefit of subsidiaries. The SCR and Pillar 2 requirements would be set at the group level with the group being able to claim a capital benefit to the extent that the risk is effectively diversified across the group. The group supervisor would be responsible for ensuring the group’s compliance with its obligations to support its subsidiaries, as well as for meeting its SCR and Pillar 2 requirements. It is important to recognise that none of this compromises the responsibilities of local supervisors in respect of the MCR, governance issues relating to the subsidiary and conduct of business. Furthermore, the success of the whole approach would hinge upon the ability of the group supervisor to co-ordinate supervisory activities and agree group-wide approaches to capital and supervision more generally. However, the success of this proposal will depend upon its ability to remove some of the duplicative effects of group and entity supervision arrangements.

We recognise that this approach is a significant departure from current supervisory practice, but we believe that there are powerful arguments for change, and significant benefits that could accrue if we were to do so. For insurance groups there would be a significant reduction in the regulatory burden, as duplication between entity and group supervisors would be reduced. More efficient cross-border co-operation between supervisors should lead to harmonisation around more sophisticated risk management techniques and systems and controls. This in turn would, we believe, promote a deeper and more efficient single market in insurance services, supporting the international competitiveness of EU insurance groups. Policyholders, meanwhile, would benefit from this reduced burden and improvement in risk management through more competitively priced insurance products whilst at the same time still enjoying the same levels of policyholder protection.

Notwithstanding the many attractions of this approach, there are, however, a number of serious issues which need to be addressed to enable it to work effectively. These cannot be dismissed as the predictable complaints of ‘flat earthers’; they raise legitimate questions of public policy and policyholder protection. The FSA does not believe that these need be insuperable barriers but it is vital that groups such as those represented here at the Geneva Association work together with regulators to find answers to these legitimate concerns. Whilst there are a number of issues to consider, I am going to highlight here what I see as three of the most pressing for industry to participate in.

The first relates to how capital flows to support subsidiaries can be ensured including in extreme circumstances. We would be looking to groups to explain when they claim diversification credit the extent to which this credit is constrained by their ability to move funds around their groups in practice, including taking into account, amongst other issues, tax implications, legal restrictions and levels of capital tied up in life funds. The UK paper seeks specifically to address this issue by proposing that a group must be able to provide support to subsidiaries up to the difference between the MCR and a solo company’s undiversified SCR. Any supervisory review of this would necessarily focus on the extent to which groups undertook stress and scenario testing on their positions.

The second issue relates to the failure of a subsidiary, possibly leading to a failure of the group. The issue here relates to how the transferability of funds can be ensured on insolvency (of a subsidiary or the group) in order to maintain equal protection for all policyholders in the group. There are a number of ways in which this could be addressed, including the potential for the creation of a central EU guarantee fund. We are considering how we might best address this issue and the input of groups, such as your own, would be especially welcome in this area.

Finally, whilst the paper relates to the activities of groups within the EU, there is the issue of if - and how - the framework to allow group diversification benefits could be extended to cover third countries. This is possibly the most complex issue but one which is of fundamental importance to many EU insurance groups with global operations or ambitions.

As I said at the start, this is an area where we are seeking the market’s input into an initiative which could be of real benefit to the market, and I therefore urge you first of all to read the paper, and second to become involved in what is certain to become an interesting debate.

Risk transfer to the capital markets

I want to turn now to my second subject – the interface between the insurance and capital markets. The appetite of the capital markets for insurance risk was amply demonstrated in the aftermath of Hurricane Katrina. Reductions in capacity in traditional reinsurance markets and an anticipation of significant rate increases in, particularly, catastrophe-exposed business led to increased capital market participation in two different ways. First, we saw a great appetite for financing and re-financing traditional insurance vehicles, but second, and perhaps more significantly, we saw greater moves towards the direct assumption of certain types of risk, through vehicles such as sidecars and cat bonds.

In the months following Hurricane Katrina it is estimated that just under $20 bn flowed into existing or start-up reinsurance companies, many of them domiciled in Bermuda. A further $5.2 bn flowed into alternative investment vehicles such as sidecars and insurance-linked securities. There has been a particular and growing interest in catastrophe bonds. Total issuances between the mid-1990s and the end of 2004 came to around $10bn, whereas it is estimated that $3.35 bn of such bonds has been issued so far in 2006 alone, which may be indicative of broader interest developing across the sector. This seems borne out by the fact that there has been a record number of transactions undertaken (17) with 7 first-time sponsors. Similarly, a recent survey of senior executives in the life industry showed that, whilst only 2% of responding firms had already entered into securitisation transactions, 50% expected to do so within the next two years.

Whilst the appetite for entering into capital markets transactions clearly exists on both sides of the bargain, there remain a number of constraints on the growth of this market in insurance risk. These were summarised earlier this year by the Group of Thirty in their wider report on Reinsurance and International Financial Markets, and included such issues as data and modelling, risk disclosure, transaction costs (especially in cases where there is a ‘novelty premium’ for innovative structures), and a lack of widely accepted indices for the analysis of risk.

Many of these are issues for the industry to address, and no doubt this will happen as the market continues to mature, but one of the points raised by the G30 is that of the regulatory treatment of securitisations. Whilst the economic effect of securitisation is often the same as that of a reinsurance contract, the effect on regulatory capital calculations can often differ considerably. This could lead to a situation whereby risk was transferred with little or no solvency relief, which might bring into question the whole economic purpose of the transaction in the first place. The issue here is not necessarily one of inherent conservatism on the part of regulators, but rather that many regulatory rules were drafted without contemplation of securitisation as a form of risk transfer, meaning that the regulatory framework is simply not clear.

So, what are we as a regulator in a leading insurance market doing to address this? As part of our implementation of the Reinsurance Directive, we have taken the opportunity to facilitate the creation of a market in the UK for insurance special purpose vehicles, or ISPVs. Currently, an ISPV in the UK would have to be authorised, supervised, and indeed taxed, as a full reinsurance company, and would need to maintain a regulatory capital surplus in the same way. The regime we will introduce is a pared-down set of requirements which recognises the relatively lower levels of risk associated with the structure of SPV transactions.

The authorisation requirements will be minimal, with little documentation required that would not already be produced as part of setting up the SPV. Supervision will take place through the oversight of the ceding company, rather than separate supervision of the ISPV. We already have a tool available to us in the form of ICAS which allows us to do this effectively, and I will describe this in a little more detail shortly. Finally, the solvency requirement for an ISPV will be a straightforward one, namely that its assets will simply need to be greater than its liabilities. In short what we aim to deliver is a regime which balances the desire to facilitate transactions quickly and economically with the need to maintain appropriate levels of policyholder protection.

As I have mentioned, the tool that we will use to ensure that the risks associated with ISPVs are properly mitigated is the ICA of the ceding company. This once again is entirely consistent with our principle of senior management responsibility and allows us to assess ISPV arrangements on a case-by-case basis and give appropriate credit depending on how risks are managed.

Whilst the treatment of specific risks will be an area of focus for us, our starting point when looking at all such transactions will be to assess whether there is sufficient and genuine risk transfer. We are conscious that for intra group ISPVs risk transfer may not be the only driver – achieving a particular tax outcome or a desired corporate structure may be equally powerful factors. But we will not entertain attempts to use ISPVs as a means merely to achieve a form of regulatory arbitrage and, specifically, will look at how the proposed risk transfer takes effect both at an entity and group level.

Whilst looking at the ICA we will expect cedants to have modelled the effects of the ISPV on all of the relevant risk areas in the ICA. Whilst credit risk may be substantially eliminated through collateralisation, some may remain, depending on the nature of the collateral or the structure of the arrangements. Perhaps more importantly, we will expect firms to consider the insurance risk associated both with exceeding limits of coverage and with any forced commutation clauses which may be present. The latter is not to be underestimated, and clearly we will have discussions with firms when ISPV transactions are unwound in exactly the same way as we will when they are established.

Another major consideration for us will be basis risk. The market often seeks clarity from us on what the regulatory treatment of SPV transactions will be, and this is an area where we feel there is a legitimate challenge back to the market to explain what the scope and extent of risk transfer will be. Basis risk exists in almost all alternative forms of risk transfer, and we will expect to see an analysis, amongst other things, of how triggers will function in stressed circumstances, and how the indemnity will operate in response to differing events. Of course, the precise nature of the contracts involved will influence what we will expect to see, and this is another reason for the non-prescriptive, case-by-case approach to this issue that I have already described.

It is often said that reinsurance is insurance between consenting adults, and we believe that our approach to ISPVs is an extension of this general principle. By placing the onus for risk identification and mitigation on firms’ senior management, and deliberately not prescribing in advance structures that we will and will not accept, we hope to encourage innovation in the market, and we believe that we are creating the opportunity for a significant new market here in the UK.

As part of this, we urge firms to come and talk to us as soon as they are considering any new forms of securitisation or any other special purpose vehicles. The earlier we are able to engage in dialogue with firms the more likely it is that we will be able to resolve issues with them and help them to achieve their objectives without compromising our own.

Some other initiatives

As I said at the start, we have taken some steps recently to help the establishment of new insurance companies in London. It is often perceived that authorisation is a very lengthy process here in the UK, and whilst that was arguably a fair accusation to make in the days before regulation was fully integrated in the FSA, today such views have little bearing on reality. However, perceptions – grounded in reality or otherwise – matter and so we felt it important to take another look at our processes around authorising new entities to see if further improvements could be made. As a result we have made changes to our authorisation processes to make them more attuned to the needs of the market, whilst safeguarding the standards which we are – rightly – bound by statute to maintain.

Insurance company applications are currently taking an average of sixteen weeks from receipt to authorisation and significantly less where they relate to a group which is already known to us. This is not merely an aspiration – we have authorised a number of insurers within these timescales already this year. In addition, we have also said that, in times of market stress, we will deal with applications even more quickly, potentially within four weeks of receipt from a group that is known to us already and ten where they are not. It should be added that all of these timescales are subject to receipt of sufficient and robust information with an application for us to meet our statutory obligations.

So in closing, all that I have said to you this morning has been designed to map out some of the ways in which the regulatory landscape is changing – we hope for the better – both in respect of the realistic treatment of complex groups and also a more innovative market in risk transfer. Much of what is happening is still very much open to debate, and we at the FSA hope and trust that the industry will engage in the debate, and potentially reap significant rewards from doing so.