Given at the Investment and Life Assurance Group Limited (ILAG) Conference, London
A chart relating to this speech is available in the speech PDF.
Introduction – reflections on the last year
It is a great pleasure to speak to this important gathering of the Life sector today.
So, the brave new world has dawned: we in the insurance sector are now operating under a new prudential regulatory regime. Solvency II has become ‘business as usual’ – a regime that has appeared, at times, a distant, indeed receding prospect, and which has undoubtedly consumed enormous efforts from all of us in this room - and, let us be frank, as with all far-reaching regulatory change, significant cost - has now been in place for no less than 48 days.
Towards the end of last year the Bank of England approved 19 internal models for the insurance sector; and of these 9 were life groups. As a result, reflecting the relatively concentrated nature of the life sector, we are now in a position where the prudential capital ratios for firms that in aggregate represent over 80% of the UK life sector balance sheet are determined by approved internal models. That is a formidable achievement, one that sets the UK life sector apart from much of Europe and one of which I believe the industry can be proud; but as I shall also explore today, it is also a development that carries profound strategic, supervisory, risk management and operational implications for firms and regulators alike.
All that said, transitioning to Solvency II has required colossal efforts on both sides of the regulatory fence and it is due to this effort that we have been able to make Solvency II a reality.
I do not think it an overstatement to describe Solvency II as a “once in a generation” reform of insurance supervision. Although it has not been without some criticism and controversy, it is important not to lose sight of the fact that Solvency II replaces a system of prudential regulation dating back to the early 1990s and before. A lot of water has flowed under the bridge since that time, making Solvency II an important and much needed modernisation, which harmonises many requirements across the EU.
That harmonisation is not yet perfected; it was perhaps unrealistic to expect a single big bang reform that imposed, overnight, a complete uniformity of capital adequacy calculation on a patch-work of national regimes. Significant transitional measures, stretching over many years, have been necessary to make the Directive work in practice for individual jurisdictions. These differences, especially in the approach to the long term valuation rate, will continue to complicate comparisons between firms across markets. Nevertheless, it is essential not to lose sight of the substantial step forward and increased harmonisation that Solvency II has brought in areas such as capital requirements, governance and risk management standards, group supervision, data gathering and reporting, and public disclosure. In sum, this represents major progress, with sizeable benefits for the industry and consumers alike.
From a competitive angle also, Solvency II represents a major step forward in levelling the playing field across Europe. A regulatory modernisation on the scale of Solvency II will take time to embed fully. The PRA has been clear on what elements within Solvency II may need further reflection and debate. That said, the remaining differences in Europe appear to be relatively modest when compared to what I have characterised as the patchwork of regimes under Solvency I.
The PRA continues to focus on how to make the transition from ICAS to Solvency II as smooth as possible. We have been acutely mindful of the perennial risk, whenever a new policy framework is introduced, that implementation crowds out everyday supervision. I can assure you that, mindful of some of the problems seen in the banking sector pre the financial crisis, we are determined to maintain our focus on business-as-usual supervision. Here we all recognise that the business environment faced by the life sector over the last few years has been exceptionally challenging by any standard as a result of the continued low interest environment and far reaching changes in the annuity and savings market.
Additionally, the dynamics of the Solvency II capital regime, in particular the introduction of an additive and highly interest rate sensitive element to the financial resource requirement in the form of the risk margin, whilst prudent, have been uniquely challenging for the life sector. Life firms are evolving distinct strategic responses to these dynamics, particularly in the area of longevity risk transfer, and I will have more to say later on our perspectives on that. But overall, given an economic and business environment that is generating major strategic challenges, our prudential supervision will continue to pay sharp attention to emerging risks we see in the market; to ensuring that governance and risk management standards are maintained, not least in terms of the adequacy of boards’ understanding of new business and diversification initiatives; and to conducting rigorous business model analysis.
What will our supervision look like under Solvency II?
As the title of my talk suggests, I want to focus on what you can expect of the UK prudential regulator in the new Solvency II landscape.
Many of you may ask what Solvency II will mean on a day by day basis – can we expect to see changes to the supervisory relationship? The simple answer to that is no: the core of our forward-looking, judgement-based, proportionate approach to supervision will remain the same, forged as it was by the lessons learned from the financial crisis. So too will the nature of our dialogue with firms that will continue to be polite, responsive and co-operative, but always frank, challenging and ruthlessly focussed on those few key areas of a firm’s activities with the potential to determine ultimately whether a firm succeeds or fails. I am very confident indeed that Solvency II complements, indeed enhances, this approach, and gives us a suitably modernised policy framework within which to take it forward.
It so happens that the PRA, will in due course, publish its updated “Approach to Insurance Supervision” for 2016. This document will re-state, in clear and simple terms, the way we will go about our business, and I encourage practitioners to take a look at our website to find out more when this goes live.
The thrust of the document is that we will continue to focus on strengthening the UK financial system through being a forward-looking and judgement based prudential regulator. I want to highlight the phrase “forward looking” here because to me it encapsulates a key regulatory lesson of the financial crisis. This was that it was simply not enough to focus on what I might call “here and now” compliance, to take undue comfort from the fact that currently a firm might – just about – be meeting its prudential ratio requirements. Rather, both boards and regulators alike need to understand deeply the prospects for a business relative to the emerging risk and plausible threats it faces. The protection of policyholders lies at the heart of the Solvency II regime and is, of course, a statutory objective of the PRA under its founding legislation. Where, therefore, the PRA judges that the risks and prospective threats are sufficient to endanger the protection of policyholders, it will not hesitate to intervene at an early stage. Nothing in Solvency II opposes or changes that approach; quite the reverse.
Since its inception, the PRA has focused on those firms and those matters that pose the greatest risk to the stability of the UK financial system and, on the insurance side, to the protection of policyholders. We believe this approach to be the right one and the one our wider stakeholders, indeed society as a whole, would want us to follow. So we do not intend to change in this regard. Being risk-focused entails our adapting our supervision to take account of an insurer’s individual circumstances and ensuring that the frequency and intensity of supervision we apply is commensurate with the risks posed to our financial stability and policyholder protection objectives and, subject to that, facilitate effective competition within the insurance market.
Solvency II will change none of that. Rather, the founding aims and principles of the new Solvency II regime, in particular the emphasis on proportionately and policyholder protection, underpinned by the Directive’s enshrining of the prudent person principle in assessing risk strategies, complements, indeed enhances, the PRA’s supervisory approach very well.
A key area of improvement under Solvency II is the richer and more-timely data we will receive under the third pillar of the Directive. Insurance is now one of the most intensive data gathering operations of the Bank of England and, whilst that has presented significant technological, systems and analytical challenges, there has been a massive step change in the quality and scope of the data we receive. This will greatly assist our analysis of trends and emerging risks. In the last year, the PRA has been intensively developing tools to utilise this new data in order to gain as much value from it as possible. Supervisors will be using these new tools and data set both to inform our supervisory dialogue and as the foundation of our judgement-led, forward-looking supervision.
With Solvency II in place, insurers and supervisors alike may be tempted to breathe a joint sigh of relief. But I hasten to add that there is still a large body of work for firms and supervisors to do this year.
As I mentioned, the PRA approved 19 internal model applications at the end of 2015. While this is a sizable number, and represented by far the largest number of any European jurisdiction, we expect to see a “second wave” of model applications this year. Additionally, firms with approved models may be contemplating model changes that will fall above the change parameters that we have approved and so will need our fresh approval – although I hope that firms will be relatively sparing in this regard. And I would flag that, whilst we are of course open to model improvements and refinements, we will bring to bear an appropriate scepticism where the only rationale for such changes is manifestly and consistently that of reducing capital requirements.
In practice, the IMAP process has stretched over many years firm by firm. I believe that both sides learned a tremendous amount in this vital “pre app” phase and that it made a key contribution to the smooth and successful implementation of the final phase of model approval in the second half of last year. For those who do not have an internal model but are looking to make an application over the course of this year, the PRA will endeavour to make the model review process as smooth and transparent as possible through providing timely and comprehensive feedback, just as we did in 2015 and before that.
A clear and undoubted lesson from the pre app and model approval process so far has been that it benefits from early, transparent and comprehensive engagement with the PRA at the earliest opportunity. I would encourage all those contemplating a model application this year to take this message to heart!
At this juncture, I would like to touch briefly on some common perceptions about the internal model approval process, perceptions that I know are sometimes strongly held in the life sector but which I believe are mis-placed. There are some broad overview points I would like to make here:
- Under the previous ICAS system, we used a series of internally derived benchmarks to inform our judgements on individual firm capital guidance. Under Solvency II, these benchmarks have been succeeded by a new quantitative indicator (“QI”) framework that forms an integral part of our assessment of internal model applications. Of course, we have not simply abandoned the learning and intellectual perspectives that we built up over many years and were embodied in the ICAs benchmarks. But, contrary to the perception I sometimes hear, this framework is far from being a simple re-badging of the ICAS benchmarks. For a start, whereas the ICAS benchmarks were single numbers or ratios, many of the QIs are structured in the form of ranges of acceptable outcome, recognising that in certain areas there is a range of reasonable judgement that can be brought to bear.
Furthermore, our thinking has moved on in many crucial respects: recognising that Solvency II imposes tighter constraints on the liquidity premium that a firm can recognise, we have adopted a range in respect of credit risk decomposition within corporate bond spreads to reflect the substantial variations we observe in firms’ asset compositions and hence inherent levels of credit risk.
And, recognising developments in industry thinking, and the latest data trends, we have adjusted our longevity calibrations and, once again, introduced a range within which reasonable judgements may fall.
- As was the case with its ICG benchmark predecessor, the QI framework is not a hard and fast, pass-fail metric, but rather one of many tools designed to help us in making our decision on whether a model application meets the Tests and Standards laid down by the Directive. It is, moreover, a starting point, not an absolute: as with the ICAS regime, our objective remains to arrive at a capital cover that is expressly tailored to the individual risk profile and characteristics of the business.
We exercise a great deal of judgmental latitude in this respect.
- To be clear, whereas under ICAS we set a capital number, the cardinal difference under Solvency II is that we approve the model that produces that number, not the number itself. And we are not just interested in the capital output the model produces. As required under the Directive, and recognising the step change in modelling complexity that has occurred in the last decade, we assess the components of the model itself, rather than the capital output in isolation. And we do not have a single capital number in mind to which we intend to hold all insurers.
- Finally, I would emphasise strongly that the PRA has not been seeking to use the internal model approval process as means to increase capital across the insurance sector. We consider that the UK life sector has been adequately capitalised. We are confident that the ICAS regime, with its 99.5% calibration, has served well policyholders and the industry alike and has stood the test of time, including the sternest of challenges during the 2007-9 financial crisis. Solvency II works to that exact same confidence standard and so we have had no incentive to seek to depart from what is tested and proven.
- Moreover, because it is underpinned with a very solid prudential floor of individual capital guidance, calculated according to the parameters of the proven ICAS regime, we have full confidence in the transitional regime negotiated as part of the Long Term Guarantee package. We have been, and will remain, very clear in our message to the market as to our view of the transitional regime: provided they meet the technical calculation parameters we have set, and are subject to the prudential floor of the previous regime, Solvency II transitionals will be fully eligible as capital. Nor should a firm’s inclusion of transitionals in its capital ratios be an obstacle to dividend payments, provided that the payment of a dividend can be sustained by the business and does not threaten its overall capital adequacy. That has always been the case of course; the introduction of Solvency II transitionals makes no difference here.
- In terms of absolute financial resource, however, it has to be acknowledged that Solvency II is additive in its requirements for the life sector, through incorporating a risk margin into the base liability calculation. Unpopular though it may be, that is an unequivocal Directive provision aimed at enhancing policyholder protection and so we have no option but to implement it. Nor is there any basis in the Directive for this additional requirement to be offset through a weakened calibration standard in the SCR. So the effect of the risk margin should not be taken as indicating that the PRA has hardened its view on the SCR or the capital requirements of the life sector. That said, the interest rate sensitivity of the risk margin is very striking and I expect that there will be plenty of room for debate as to whether its design is optimal in relation to its intended purpose and developments in the market.
Of these points, I would like to spend a bit more time explaining the quantitative indicator framework because, understandably, this has been a key focus of debate with firms seeking model approval.
You would have seen that on the 15 January the Bank issued a letter and accompanying technical paper on this very subject. I would like to highlight a few key points from this.
Firstly, we use the framework as one of a variety of tools when considering whether a firm’s model meets the Tests and Standards under Solvency II. The QI framework essentially captures the PRA’s own view of what a 1-in-200 year shock could look like.
It is in no one’s interests, of course, for the model approval process to have the character of a rubber stamp. Whilst the system was designed to make substantial room for firm’s assessments of their risks, on the basis that they are best positioned to do so, plainly the prudential regulator, without in any way claiming a monopoly of wisdom, must also have a view in relation to the Tests and Standards laid down by the Directive. Moreover, that view needs to be well informed and independently and credibly arrived at. As the prudential regulator, we cannot be expected credibly to form that view without ourselves having reached an assessment of what is a reasonable calibration level.
We do not adopt a black-and-white approach to applying these QIs; instead we apply judgement. This means making sensible allowances where a firm’s model is stronger in some areas and weaker in others, within reasonable tolerances, as long as we are satisfied that this would be robust in the face of balance sheet change and plausible stresses.
Finally, I would like to address the claim I sometimes hear that the PRA ‘really’ expects insurers to hold much higher levels of capital and is looking somehow to introduce “buffers on buffers”.
In particular, I want to dispel any notion that the PRA is intending insurers to hold much higher capital levels than is required under the SCR. Solvency II requires firms to capture all quantifiable risks within the SCR and that the SCR itself should represents a 1-in-200 year standard. As part of the internal model review process, we have sought to be robust but fair in our assessment. This invariably means that it should be unnecessary for the PRA to look to firms to hold substantial further capital on top of the SCR amount.
In practice, firms have revealed a clear preference to operate some way above their SCR, not least to meet rating requirements and to provide a margin of safety against an inadvertent or unforeseen breach. But given that Solvency II is a highly sophisticated and risk sensitive system of capital measurement, we would not necessarily expect to see the sort of large multiples of capital requirements that typically characterised the published ratios of the previous Pillar I regime, most notably the IGD and FGD ratio measures, neither of which was economically arrived at or risk sensitive. For that key reason, comparisons between firms on these ratios as compared to those they report under Solvency II are unlikely to shed much light or otherwise be worthwhile.
We have been very clear, and indeed are fully on the record, in saying that we are not deploying a single ratio some way above the SCR, as a formal intervention point across the insurance sector. Nevertheless, should a firm’s capital level trend sharply downwards towards its SCR, you can expect the current and prospective capital position to be given progressively more attention from your supervisor, as we need to be in a position to activate the ladder of intervention should that become necessary. We will be taking into account the volatility of each firm’s capital position, the nature of its business and idiosyncratic risk exposures, and what scope it has to improve the capital position through credible and timely management actions. We will also be mindful of the Own Risk and Solvency Assessment (ORSA) that firms themselves produce. Any departure from the level of capitalisation indicated by the ORSA should of course be a matter of intense enquiry for the board as much as for the regulator.
One of the risks of a prudential regime that permits firms to calculate their own capital requirements is that the system, over time, is gamed. The experience of the banking sector in the decades before the financial crisis is instructive here. In the years following the introduction of Basel II, the risk weights bore, at best, an increasingly tenuous relationship with real levels of risk being taken on balance sheets.
Chart 1 plots the average risk weight applied to the assets of 17 major international banks over the period 1993 to 2011, together with a trend line. The trend is steeply and strikingly downward-sloped, falling relentlessly on average by 2 percentage points each year. Banks’ average risk weight (risk-weighted assets per unit of assets) almost halved, falling from over 70% in 1993 to below 40% at end 2011. At the same time, you will see leverage in the banking sector increased dramatically.
In other words, coincidentally with, or perhaps as a consequence of the introduction of a new capital regime, a secular downward trend in capital strength steadily occurred.
As a prudential regulator, we have a responsibility to ensure that the lessons of the banking sector are learnt and we are determined to do so, whether in our regulation of insurers or banks.
There are, of course, fundamental differences between Basel II and Solvency II and so we need to guard against overly simplistic mantras or comparisons. But we also need to recognise that, left to its own devices, competitive pressure can exert a steady and determinedly downward pressure on capitalisation. We will accordingly monitor trends and developments in insurers’ model capitalisations very closely indeed, looking to guard against any pronounced downward drift. This will involve monitoring firms’ capital requirements and, in particular, ensuring that firms’ capital levels do not fall when real-world risks either stay the same or even increase.
But that is not a role for the regulator alone, or even primarily of course. Boards are the front line in this respect. That is why the Solvency II model approval framework, through the use test requirements, explicitly requires boards and senior management to understand the drivers and key sensitivities of the model, to incorporate the model fully into their governance, risk monitoring and strategic and business planning, and to ensure that the model continues to capture the key risks of the business. Our routine supervision will regularly review how well boards are fulfilling these Directive expectations and requirements.
A forward look
I mentioned earlier that the capital dynamics of Solvency II were prompting firms to consider strategic and business model changes. These changes are also being spurred by intensifying competition in the sector and the prevailing low interest rate environment, pressure on profit margins, especially in the annuity market, and the emergence of alternative retirement products. Of these changes, I would like to focus particularly on longevity risk transfer.
A number of UK life firms make widespread use of risk transfer mechanisms through reinsurance and intragroup reinsurance. Indeed reinsurance has for many years been a staple activity of both the Life and GI sectors. We recognise that there has been an active market in the transfer of longevity risk for a number of years from which risk management has undoubtedly benefited. Traditionally, insurers have transferred longevity risk for one or more of the following reasons:
- Firstly, to balance their overall risk exposure. The growth of the bulk purchase annuity market
(which enables pension schemes to transfer longevity risk to insurers) increases the demand for longevity reinsurance by insurers as they lay-off some of the additional risk assumed.
- Secondly, balance sheet capacity constraints, and/or a deliberate strategy of switching the deployment of capital to other business lines.
- Thirdly, pension fund de-risking: a number of recent transactions have involved firms insuring the longevity risk in their staff pension schemes and thereby reducing the overall risk of the schemes.
Solvency II potentially provides firms with an additional incentive to undertake transactions to transfer longevity risk by way of reinsurance (especially for new business which does not benefit from any transitional measures) as doing so may ease the requirement for holding a risk margin in respect of the reinsured liabilities. We recognise that the current design of the risk margin makes it acutely sensitive to interest rate conditions. However, we will be monitoring closely if firms become active in this market consistently and solely for reasons other than seeking genuine risk transfer.
It is also worth noting that such transfers are not risk free: should an insurer transfer risk to a single or small number of reinsurers, it inherently exposes itself to counterparty credit risk and, potentially, substantial concentration risk. The PRA expects approved models to calibrate these risks robustly. Additionally, we expect firms’ boards to monitor, manage and mitigate these risks intensively. Indeed, Solvency II requires firms to have in place a risk management system covering ‘concentration risk management’ and this risk management system should include risks that are covered by the SCR as well as those that are not.
In practice this means that holding capital under the SCR may not be sufficient in itself to mitigate this risk. Additional measures besides capital, for instance collateralisation, may be necessary and we will focus intensively on the adequacy of these case by case in relation to our end goal of adequate policyholder protection.
The PRA’s expectation of firms’ appetite for reinsurance, and their identification, reporting and mitigation of major reinsurance counterparty default risk, has not changed under 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. Accordingly, Chris Moulder, my counterpart on the GI side, and I issued a joint Directors’ Letter on 9 February 2016 explaining that, in order to supervise firms’ risk management practices, the PRA expects to be notified of proposed longevity risk transfer and hedge transactions, and the firm’s proposed approach to risk management, well in advance of a firm completing such a transaction.
Having covered some of the challenges of Solvency II, I would like to end with a moment of perspective. Sam Woods recently spoke at the Lloyd’s library on ‘the history of insurance regulation in the UK’. The scope of the speech was remarkably broad and sweeping: it goes right back to insurance in Babylonian times and dwells on such curiosities as the development of insurance products in the Florence of the Medici; it then turns to the many innovations that characterised the twentieth century and the way that these which went hand in hand with profound social changes. I would commend it as a good read for anyone with even the most passing interest in the sector. But what most clearly comes out of this grand narrative is the sheer durability and adaptability of the UK insurance sector. Seen against that backcloth, the changes wrought by Solvency II appear as no more than a continuation of the steady progress of the sector and its regulation. Furthermore, they highlight the sector’s intrinsic resilience and innovation. Over several centuries, the UK insurance sector, and the life sector in particular, has tackled and adapted to new and emerging challenges whilst continuing to play an essential role in serving the economy and society in which it functions. I have every confidence that will continue.