Investment - Articles - Risk transfer – and the risks it creates

 It is a great pleasure to speak to you today. May I firstly congratulate the conference organisers today on their choice of subject topic – namely bulk purchase annuities and longevity risk transfers – both of which are extremely topical and timely and give me a welcome opportunity to give the regulator’s perspective.

 What I plan to cover today essentially follows on from a speech I gave in February at the Investment and Life Assurance Group (ILAG) Conference in London. You may have seen that this got some press coverage, complete with, surprisingly, a photo of myself in one august journal – I had not realised the subject would be so popular! At that event, I introduced the Bank of England’s perspective on the topic of longevity risk transfer and I would like today to give that topic some further consideration as well as discussing the Bulk Purchase Annuities (BPA) market and current trends we are seeing here. These two topics are not of course mutually exclusive: bulk purchase deals, usually involving transfer of large blocks of liabilities, often rely on the availability to the insurer of risk transfer mechanisms. It is, moreover, important to recognise that “risk management” in this context in practice needs to embrace many discrete and challenging areas of risk: in particular, pricing, asset allocation, longevity, the practical machinery of risk transfer, and documentation and operational risks generally all come fully into play in this market.

 The backcloth to all these matters of course is Solvency II. For some 119 days now, the insurance sector has been operating under the regulatory framework of Solvency II. Getting to this point has been no mean feat and has taken practitioners and regulators alike many years of concerted effort to implement. However, whilst Solvency II is still in its infancy, at least as an operative regime, if not in design, all the signs are that it is bedding down well as “the new normal”.

 Solvency II is often and rightly described as a “once in a generation” reform of all aspects of insurance regulation. It is, moreover, a maximum harmonizing Directive. But given the context – 28 jurisdictions and national regimes, each with discrete macroeconomic and insurance product and market structures – it would be unrealistic to expect that the Directive would, overnight, impose a complete uniformity of prudential regulation. There are transitional measures in place, some extending out by some 16 years in order to make the shift from Solvency I to Solvency II smoother for individual jurisdictions. These differences mean that complete harmonisation is not yet fully in place and so comparisons across markets must still be qualified. Nevertheless, we have taken a massive step towards a genuinely harmonised system for capital requirements, governance and risk management standards, group supervision, data gathering and reporting, and public disclosure.

 As the new Solvency II rules fully bed-down and become the new ‘business as usual’, wrinkles in the regime will need to be ironed out. Meanwhile, Solvency II will undoubtedly have an impact on our insurers, impacts that are already manifesting themselves through changes to business models and strategies.

 Meanwhile, the reforms announced in the 2014 Budget have, somewhat unusually for the industry, made pensions regular front page news and have of course led many insurers to rethink their strategies and approach to the retirement market. Since the reforms came into effect in April last year, there has been a sharp decline in individual annuity sales, with only 17,800 sold in the first three months of the new regime (April-June 2015)[1] – this represents a fall in sales to about a quarter of the level that prevailed before the reforms were announced.

 The significant fall in volume of the individual at-retirement market has led insurers to look instead to new product designs to replace individual annuities business. It has also encouraged more insurers to enter the bulk annuity market as a way of deploying their existing skills and experience in longevity risk management to offset the decline in sales of individual annuities. With a large number of insurers looking to achieve a market share, the bulk annuity market is likely to become increasingly competitive and it would seem reasonable to expect this unleashing of competitive forces in turn to lead to some erosion in margins.

 As I mentioned in my February speech to ILAG, this increasingly competitive business environment is also being affected by the prevailing low interest rate environment and this is placing additional pressure on profit margins. Here discussion inevitably turns to the risk margin component of Solvency II, which is perhaps the single biggest design difference in the new regime compared to the ICAS/P2/realistic balance sheet regime we introduced in the UK a decade ago under the umbrella known as the Tiner reforms.

 Let me say a few words as to our view of the risk margin. We at the Bank recognise that the current design of the risk margin makes it highly sensitive to interest rate conditions to a degree that was probably not foreseen by the system’s architects and is undesirable. In the Bank’s recent and public response to the European Commission’s Call for Evidence on the topic of EU Financial services regulation, which is available on the Bank’s web site, we noted that this level of interest rate sensitivity was making the risk margin excessively volatile and that this was likely in turn to have significant absolute and hedging costs for firms. We also noted that this degree of volatility was likely to be undesirable from both a macro and microprudential point of view because of its potential to promote pro-cyclical investment behavior. Consequently, whilst we support the concept and purpose of the risk margin, our view is that its design needs to deliver more stable balance sheet outcomes for insurers and hence support their key role as long term investors in the real economy.

 Bearing in mind this confluence of competitive pressure and the mechanical operation of the new regime, Solvency II could provide firms with a powerful incentive to carry out trades to transfer longevity risk by way of reinsurance. That is particularly the case with new business which does not have the benefit of transitional measures. Transferring risk in this way could ease the net financial impact of having to hold a risk margin in respect of the reinsured liabilities; but in turn any such transfer also brings into play a number of substantial risk management and regulatory considerations and so can hardly be considered to be risk free.

 Now we recognise of course that reinsurance is an inherent and important part of the risk management arsenal available to insurers and that it has long established and entirely legitimate uses. Nevertheless, we have also made clear that the Bank will be monitoring closely if firms become active in longevity reinsurance regularly and exclusively for motivations other than seeking a genuine transfer of risk or in any way that calls into question the adequacy and effectiveness of policyholder protection. The importance of ensuring policy holder protection is enshrined in the wording of the Directive and lies at the heart of it as a construct – and as the regulator, we will seek, through our supervisory judgments and approach, to uphold what for us is this cardinal objective.

 Bulk Purchase Annuities (BPA)

 I would like to turn briefly to recent trends and developments in the Bulk Purchase Annuities (BPA) market.

 Defined Benefit (DB) pension schemes are increasingly looking for ways to reduce their risk exposures. Insurers are well placed in principle to provide solutions and this has resulted in an increase in pension de-risking transactions in the form of buy-ins, buy-outs and longevity swaps in recent years. Indeed many commentators have argued that insurers are the natural home for the long term risks associated with pension schemes and so have welcomed this development. As the prudential insurance regulator we are not blind to the potential business opportunities and competitive advantage of insurers in this area. However, there are areas in which it would be appropriate for me as a prudential regulator – and you would expect nothing less of me here (!) - to strike some cautionary notes, particularly for those firms that are new to this market.

 Let me start with market size. The market has grown substantially in recent years, with bulk annuity deal flows increasing from £4bn in 2012 to just over £13bn in 2014; 2015 also saw over £10bn liabilities transferred to the insurance market through buy-in and buy-outs.

 However these figures pale in comparison to the potential size of the market. Total liabilities of UK DB pension schemes currently stand at £2 trillion and to date only £150bn of these have been transferred to insurers. One analysis I have seen suggests that a transfer of just a quarter of these DB liabilities would consume all of the present excess capacity of the UK insurance market.

 This all has the makings of a potentially massive market opportunity. There is also an argument that a wider social good is done. By taking out pension liabilities, insurers can enable corporates in other sectors either to focus more on their core business propositions and/or reduce the risk of disorderly wind down. However, in the past we have seen instances in the BPA market where:

 Firstly, in order to gain market presence, pressure has been put on firms with shorter track records to use price as a differentiating factor in order to win initial deals; and in turn, price has been actively seen as a good lever with which to attract the necessary volume of bulk contracts, particularly where firms are seeking urgently to replace ‘lost’ volumes of other business.

 For me, as a non- actuary I hasten to add, one particular fact stands out crucially in all this: whilst pricing errors can cause problems in the individual market, in the bulk market the effects of such errors are inherently magnified, indeed one might even say supercharged, to a degree that in extremis could be highly damaging for the firm concerned. Taking on large volumes of under-priced annuities in one fell swoop carries a high likelihood of pressure on capital availability for many years to come as that business runs off. It is also likely to make any attempts to transfer risks associated with the business look relatively expensive and hence unattractive or unviable, which in turn creates a further regulatory problem.

 On a related note, to become, or indeed remain, competitive in a low interest rate environment, firms may succumb to the temptation to venture into asset classes where they have no, or limited, experience. We saw some instances of this in recent years where new firms tried to enter the individual enhanced annuity market but struggled to price competitively with a corporate-bond based asset strategy. We see potential parallels today in the bulk purchase space. Solvency II places strong emphasis on good risk management and the prudent person principles. I take this opportunity to remind firms that these requirements need to be met.

 As a final point on market size, we observe that many firms enter into the bulk market with stated limits on minimum and maximum deal sizes as part of their risk management. But we also occasionally observe that these limits can be overridden because they are viewed only as a “soft” or “indicative” limit; or indeed they may simply and conveniently be forgotten when larger or smaller schemes become available. Where we observe such behaviour, we are concerned because it suggests that these limits were not intended to be adhered to in the first place and therefore calls into question the effectiveness of firms’ risk management, culture and governance.

 Moving on to the more operational aspects, our over-arching concern is to ensure that a firm’s governance, risk management and operational processes are fit for the task of taking on pension liabilities in bulk, even where the capital strain is manageable.

 I emphasise that it is no trivial matter to manage the transfer of pensions to the insurance sector. Pension scheme liabilities often require a highly granular data cleansing exercise to be undertaken with scheme rules and trustee-level contractual clauses potentially further adding to the complexity. Care is also required around the legal basis of the transfer and the definition of risks that are and are not to be transferred. Firms therefore cannot approach this market in the same way as they do individual business. Systems and processes will be required to allow for a smooth transfer as well as sound on-going management of the assumed liabilities.

 For example, firms will need to consider how to allow for features specific to bulk transactions such as:

 
1.   ensuring that scheme rules, if applicable, continue to be applied to the correct policies.
 
2.   making sure any changes to pensions legislation are appropriately allowed for; and,
 
3.   obtaining appropriate assets to match the inflation indices to which the pension scheme liabilities are linked.

 For buy-in contracts, options and benefits provided for the scheme itself are also potential ways in which insurers can differentiate themselves to win business. We have seen a wide range of such features including surrender options, collateral arrangements and the options to defer premiums or include additional business at a later date. Whilst we have no particular concerns in principle about this, we have seen instances where contracts have had to be revisited in the light of such options, with the end result turning out to be quite different from what the insurer originally envisaged. Such options can also often present particular challenges for matching adjustment eligibility and so need careful consideration before being included in a matching adjustment portfolio.

 Few firms will have either operational or financial capacity for more than a handful of medium-sized bulk deals in any one year. I have touched on a number of the operational issues already so now let me turn to the financial capacity of firms to take on liabilities in bulk as my final point on this market.

 Specifically, where balance sheet constraints are in play, firms will either need to limit business amounts or find a way to increase capacity. One such way to increase capacity is via transfer of some or all of the risks assumed under the bulk purchase transaction. This leads us back squarely to the topic of longevity risk transfer.

 Longevity Risk Transfer

 Let me start by considering what longevity risk transfer means in practice.

 Risk transfer is an essential tool for managing risk exposures and profiles. For a number of years there has been an active market in the transfer of longevity risk although we recognise that this is not as well developed as market participants would like. A lot of the issues turn on two essential characteristics of longevity risk:

 Longevity risk can be quite bespoke to individual firms and exposures. Off-the-shelf index-based solutions can therefore be deeply unattractive to an insurer seeking to manage its risk profile.

 Moreover, longevity risk is a long-tailed risk. However, there is a limit to the number of capital market participants that are willing to provide long-term risk transfer solutions. Exposure to rollover risk on the risk transfer instruments is not something that would normally be acceptable to a transferring entity. As a prudential regulator we also have material reservations as to the adequacy of risk transfer that would be achieved through such solutions.

 As to the risk transfer market that currently exists, insurers may choose to transfer longevity risk for a variety of reasons bespoke to the business and its business requirements. However, one or more of the following reasons could also be at play:

 Firstly, longevity transfers enable firms to balance their overall risk exposure and reduce concentrations of risk. The growth of the bulk purchase annuity market (which is essentially another mechanism enabling pension schemes to transfer longevity risk to insurers) increases, as a second order effect, the demand for longevity reinsurance by insurers as they lay-off some of the additional risk concentrations assumed.

 Secondly, as I noted earlier, the costs of the risk margin for new business under Solvency II may be creating incentives, at least in principle, for firms to amend their business models towards greater use of risk transfer.

 Looking across this market, it is clear that longevity risk transfers are on an increasing trend. Longevity risk transfer transactions have covered at least £131bn of insurance and pension scheme liabilities since June 2009. After a slow start, with just over 20 longevity swaps transacted between 2009 and 2012, the longevity swap market gathered pace with the highest number of transactions taking place during 2014 with a record of £40 billion liabilities hedged[2].

 I hasten to add that reinsuring risk can form part of a genuine risk management strategy. But as part of the decision to transfer risk to a reinsurer we expect boards of insurance companies to understand rigorously the mechanics of the risk transfer taking place and to take into account any wider associated risks which may arise. This is because such transfers are not risk free; to transfer one set of risks, other risks have to be assumed. In particular, when an insurer transfers risk to a single or small number of reinsurers, it inherently exposes itself to counterparty credit risk and, potentially, concentration risk. In a nutshell, it falls to boards to ask themselves in what circumstances might the transfer breakdown such that the liabilities end up on their balance sheet, and the probability and consequence of that.

 We therefore look to Boards to monitor, manage and mitigate these risks with a critical, searching perspective and with a particular focus on collateral arrangements.

 In particular, boards should ensure that the firms’ overall risk transfer strategy is consistent with the Solvency II requirements on risk management and the prudent person principle. Specifically, firms need to have in place a risk management system which monitors, manages and mitigates ‘concentration risks’ arising from its risk transfer strategy and ensure that it is appropriately reflected in the SCR. Boards should satisfy themselves that that the associated capital requirements, whether calculated using standard formula or internal model, appropriately reflect the risk profile of the firm.

 Our expectations surrounding reinsurance have not changed as a consequence of Solvency II. It is, however, important that we are sighted when insurers are entering into these transactions and understand fully the rationale and terms on which they are doing so. The PRA will use this information to inform our supervisory strategy and keep up to date with market developments. I emphasise that it would be unacceptable to us if firms were to use this market primarily as a tool to achieve regulatory arbitrage and to avoid key in-built requirements of our new solvency regime rather than to manage their risks in the interests of their policyholder and the firm.

 Conclusion

 There are a number of external factors at play which could well act as a driver for insurers to consider making changes to their business models and consequently their risk profiles, by way of risk transfers. Whilst the insurance market may be operating at a time of increased transformation and change, to be clear, there has been no change in our expectations of firms when it comes to ensuring the necessary prudence in business model decisions and reinsurance. We remind firms that their usual approach to making strategy decisions should continue. Boards should make certain that they fully appreciate the risk transfer taking place and any knock-on implications there may be. Appropriate risk management systems should be put in place to reflect the changed risk profile of the firm.

 With regulators and regulated alike now adapting to the brave new world of Solvency II, we ask that insurers maintain an open dialogue with us as their prudential regulator on all of the market developments and risk areas that this conference is usefully covering today. For our part, we shall be watching developments closely and will continue to welcome an informed and detailed dialogue with market participants on all aspects of the BPA and longevity transfer markets. Doubtless that dialogue will be keenly pursued by both sides of the regulatory relationship.