Given at the Association of British Insurers, London.
Before we start today’s panel session, I wanted to take this opportunity to say a few words. I promise to be brief.
As a general rule, I am not a fan of sporting metaphors in speeches. But as a New Zealander I find it impossible not to introduce a minor element of rugby given the events of last weekend.
Like the Rugby World Cup Final, the run-up to Solvency II has been a long, arduous undertaking. The players, including many of you and many of my staff, have been grinding down the pitch for a decade. But by my reckoning, at T-minus 59 days to the arrival of Solvency II we are now at the equivalent of 1.62 metres away from the try line. I feel like the last back in the line, ball in hand and (hopefully) moments away from scoring.
So the first thing I wanted to do is thank you – whether you are on my team or someone else’s – for the huge effort that has gone into getting us to this point. Next I would like to recap what’s in store over the next few months and offer a few of my own early reflections on the new regime.
Solvency II: using stoppage time
I am confident that we have made the most of any stoppage time afforded by policy delays during the negotiations over Solvency II. We introduced the internal model approval process, developed ICAS+ and have been running “pre–applications” for some time now. These processes and pre-applications were brought in to provide opportunities for feedback and allow firms to kick the tyres before switching their models on.
But where are we now? The PRA is busily reviewing around 300 Solvency II applications, of which around 20 applications are from firms seeking approval to use an internal model on day one of the new regime.
It is essential that our process for reviewing these applications is fair and consistent across our firms, not least in order to preserve orderly markets. In order to deliver this, each firm’s internal model application is being taken to a recommendation panel made up of PRA senior management. This will be followed by a further review by a separate senior decision-making panel which will look at all the decisions in the round and make the final calls.
We are intending to communicate our decisions on matching adjustment applications to each individual firm simultaneously in early November, and to notify those firms which have gained internal model approval in advance of 1/1/16 at the same time in early December.
So, November and December will see the culmination of some excellent preparatory work, numerous years in the making. However, I should stress use of the word ‘preparatory’. Come January, we will not be sitting on the side-lines. I know better than most that there is still work to be done, not least because of the significant number of firms whom I expect to submit applications that will need to be considered over the course of 2016.
To be clear, internal model approval is not an end in itself. For some firms, an internal model might not be the right answer – the standard formula is perfectly appropriate for the large majority of our firms, and for others a partial internal model might be the right answer. For those who do gain model approval this December, they should make sure their models remain fit for purpose on an ongoing basis and will need to assure us that this is the case.
A margin for risk
In the UK, firms have already been operating under the ICAS regime for a number of years and many of the principles of ICAS will continue under Solvency II. However, as with any new regime, Solvency II will introduce some changes.
I would like to touch briefly on one such change – the introduction of the risk margin. Today I want briefly to consider the role of the risk margin in the new regime, because in my opinion the legislation is not explicit about this (though it is clear about how it should be calculated).
The intellectual construct for the risk margin is that it is the compensation that a third party would require in order to take on the assets and liabilities of another firm that runs into trouble, as compared to the current regime where we assume the firm goes into closed run-off. This would reduce the risk for the purchaser of the liabilities turning out to be bigger than expected. Many of us have therefore used, as shorthand, a description of the risk margin as facilitating a shift from a gone to a going concern regime. Whilst this is crudely true, this interpretation does not adequately, to my mind, explain the role of the risk margin.
The reason for this is simple. Any firm that takes on another firm’s assets and liabilities will itself have to hold a risk margin against that balance sheet, just like the transferor had to the day before. So the compensation for the transferee only materialises when the liabilities run off and the risk margin is released.
This is a somewhat attenuated conception of ‘going concern’. It would therefore perhaps be better to talk about the risk margin as a shift to a fully market-consistent regime. Nonetheless, one thing is clear: the risk margin should provide an extra degree of protection for policyholders, which should advance the PRA’s objectives. For instance, the presence of the risk margin may enable ‘white knight’ deals to take place at a wider range of sale values, because of the additional security it affords the buyer – and this may prove beneficial if we are faced with a firm in distress.
So as a supervisor I think the idea of the risk margin is a good one. And as a practical matter, it will be the law from 1 January and we will implement it properly. However, in the medium-term I think legitimate questions could be asked about whether the way it is calculated under the Directive is really optimal.
I will touch on two points here:
i. Because of the flat 6% cost of capital in the Directive, the risk margin is very sensitive indeed to risk-free rates. For example, we estimate that a 50bps increase in risk-free rates would reduce the risk margin for the UK life sector by around 20%. There are reasons why investors might seek more compensation when rates are low, but it is also possible that the rate of return threshold could become lower in such conditions. Personally I am sceptical that such a high degree of volatility is desirable from a microprudential or macroprudential point of view.
ii. The risk margin is intended to capture ‘non-hedgeable’ risks such as longevity. But it is plain that in some circumstances life insurers can shift out longevity risk through swaps or other methods. We are monitoring this activity carefully – but it does beg the philosophical question: what is non-hedgeable? The reinsurance activity also raises practical questions about how we manage the interaction of different regimes across the world, many of which have no concept of a risk margin. I am worried that the unintended result could in fact be reduced, not enhanced, protection for policyholders.
With rates where they are, the risk margin is currently large for UK life insurers. In the near-to-medium term, transitional measures absorb the impact of this – and I would like to repeat what I said in July: the Bank will allow full use of transitional measures by those firms that qualify to use them. And when we consider whether or not firms are in a position to pay dividends, one of the main quantitative yardsticks we will use is capital levels after the benefit of transitionals.
Level playing field
I have no doubt that, when compared to the situation today, Solvency II will be a massive leap forward in terms of establishing a level playing field across Europe. Indeed, for the first time we will all be playing the same game!
But of course, as we get into the nitty-gritty of implementation then some differences in national styles will emerge. It is even possible that a few nasty things might go on in the scrum before the referee is able to intervene.
Looked at in the round, I think these differences are surprisingly modest so far. There is no equivalent of the northern hemisphere / southern hemisphere divide. But I do want to touch very briefly on one current area of difference – how to implement the volatility adjustment.
It remains our view, and that of some others in Europe, that firms should not assume any change to the level of volatility adjustment when calculating their capital requirements. The position we have taken is based on our reading of the Directive, and the purpose of the volatility adjustment. On the former, the Directive states straightforwardly that the SCR shall not cover the risk of loss of basic own funds resulting from changes in the volatility adjustment. On the latter, our view is that to take full credit for a dynamic volatility adjustment upfront would make the volatility adjustment a less effective lever for the authorities to pull when the stress actually arises, by increasing the discount rate insurers can use to value their liabilities.
But there is another school of thought, favoured by some others in Europe, which is that if the purpose of an internal model is to forecast changes to a firm’s balance sheet in a 1-in-200 stress, then it is illogical to isolate this one element and hold it fixed. This is not an incoherent view, but I think that the prudential arguments, both macro and micro, point firmly the other way.
I mention this issue not to do the arguments full justice, but just to illustrate the sorts of differences which will inevitably emerge as we bring all of our regimes together into a single new one. This issue, and any others like it, will need to be ironed out before too long.
The height of the bar
To conclude my talk to you today, I would like to touch on what ‘normal’ will look like under Solvency II.
Solvency II is clear that firms must hold sufficient capital to meet their Solvency Capital Requirement (SCR). I believe that we have done a thorough and robust job in making sure that for UK firms that requirement represents a 1-in-200 year standard, as set down in the legislation. Although this calibration is the same as the one we have been using under the ICAS regime, this has not been easy for anyone on the pitch, and while no one has been stretchered off I’m afraid that a certain amount of blood has been spilt.
Solvency II sets the bar at 1-in-200, broadly equivalent to a BBB rating. However, Solvency II does not require us to place an extra bar on top of that bar. So I am concerned when I hear market participants assuming that there is some other bar, very much higher than the SCR, which all firms must somehow meet in order to satisfy regulators.
Now of course analysts, investors and people running insurance firms may pursue a higher bar if they wish to – for instance, in order to achieve a higher credit rating. And as a practical matter, the frictional costs to a firm of running so close to its SCR that it is constantly breaching it might be quite high. So as in all regulatory regimes that I have experienced, I expect that many firms will choose to hold some kind of buffer above their SCR.
Supervisors also naturally tend to monitor capital more closely as a firm comes closer to its regulatory requirement – this is common sense. We will of course do this under Solvency II, taking into account the volatility of each firm’s capital position and the risks to which it is exposed as set out in its Own Risk and Solvency Assessment (ORSA). However, I would like to caution market participants against the idea that we have in mind a single ratio, somewhere very high above the SCR, which we have set as an intervention point across the insurance sector. That would be illegal, would frustrate the intention of legislators to deliver a
1-in-200 level of solvency across Europe, and would not take into account the different risks different firms run.
Approaching the try line
Now, I calculate that during the course of this 15 minute talk we have moved approximately 0.28 millimetres closer to the Solvency II try line. Actually the truth is that I am terrible at running, one reason among many (you should see my ball-handling skills) why I am pursuing a career as a public sector official rather than an All Black! But despite that, between us all I am confident that we will ground the ball.
Many thanks for your time today.