162.This Chapter describes a number of technical matters that were raised in the written and/or oral evidence:
b)Transitional Measure on Technical Provisions (TMTP)
d)One Year Value at Risk
163.The first two are relativity crude adjustments to help counter some of the limitations of Solvency II and received a great deal of comment, particularly from Life firms. The last two challenge the conceptual framework of Solvency II.
164.The Volatility Adjustment is similar to the Matching Adjustment, and can be applied to insurance products which are not eligible for the Matching Adjustment. It is a mechanism “albeit only of limited effect and artificial in application, that allows liabilities to be reduced on a prudent basis when asset values are particularly low”.
165.The Volatility Adjustment is only available in certain currencies (US Dollars and European currencies), limiting its usability for international insurers. A further limiting factor is that HM Treasury decided that firms must seek regulatory approval in order to use the Volatility Adjustment, unlike some other EU member states on the ground that it may encourage pro-cyclical behaviour. However, the IFoA point out that requiring approval “could also lead to pro-cyclical behaviour, in which insurers would become forced sellers in an economic downturn”. For this reason the regulator’s approach has been labelled “unnecessarily rigid”.
166.Most of those who gave evidence on the subject to the previous Treasury Committee noted that the rules could be changed without needing EU permission as “different countries in Europe have done different things”, setting a precedent for change. They made several suggestions on how the Volatility Adjustment could be improved. For example, the IFoA argued that “it would be appropriate for the ‘emergency’ VA approval to be granted were necessary”, and the Volatility Adjustment “should be permitted to vary in assumed stressed conditions i.e. it should be treated as dynamic rather than fixed”. This would align the UK approach with some other European jurisdictions where the Volatility Adjustment is recalculated in a stress. In oral evidence to the previous Committee, Andrew Chamberlain noted that:
“It is well documented in the public arena that the Dutch regulator has already taken a more liberal view than the PRA. As far as we are aware, no action has been taken against the Dutch regulator in regard to that, so there must be some scope”.
167.The Volatility Adjustment was briefly touched upon in oral evidence given to the previous Committee by the regulator, who noted that £1 billion of Volatility Adjustment had been “handed out” to the insurance industry. It was not explored in the PRA’s written submission.
168.The Committee agrees with the oral evidence that its predecessor heard from Andrew Chamberlain that the Volatility Adjustment is “fairly crude”. From the oral and written evidence it would appear that there are two main problems with the Volatility Adjustment. First, whether it should require approval from the regulator, and second, whether firms should be allowed to use a ‘dynamic’ Volatility Adjustment. Given the problems of procyclicality examined in Chapter 4, it would seem prudent for the Volatility Adjustment to vary in certain prescribed circumstances, without regulatory approval. Given the actions of other regulators, notably the Dutch, this action seems uncontroversial, and the Committee sees no reason why it should not be undertaken without delay. An update should be included in the PRA’s progress report requested at paragraph 90.
169.In the absence of any transitional relief, the introduction of Solvency II resulted in a significant increase in the level of capital required to be held for regulatory reporting purposes compared to the previous UK regime. This was a bigger issue for the UK industry than most others because of the nature of UK products, and a particular problem for the life industry due to the impact of the Risk Margin. This could have resulted in reported solvency problems for many companies, despite their unchanged assets, liability and risks from levels previously accepted by the PRA. Insurers may therefore, with supervisory approval, apply for transitional relief (the Transitional Measure on Technical Provisions or “TMTP”), which allows the initial increase to be phased in over a period of 16 years by means of a deduction from technical provisions as calculated under Solvency II.
170.In 2016, the first year of Solvency II, the availability of TMTP was considered to be “beneficial and generally effective”. However, in order to gain PRA approval for any re-measurement that may be required in future, there is a risk that insurers will in effect need to maintain their old models running alongside Solvency II models and calibrations. The ABI asserted this will be “costly and not practical”, while the calculation itself has been labelled “highly complex”. For example, concerns have been expressed on the drafting of the rules, particularly “the so-called ‘double run-off’ issue in relation to in force technical provisions, which superimposes the 16 year run-off of the initial impact of Solvency II onto the natural run-off of the business existing at 1 January 2016…and the degree of onerousness in the recalculation for the transitional provisions, particularly in the light of falling interest rates”.
171.The transitional benefit only applies to provisions for business written prior to 1 January 2016, not to new business. As the TMTP is reduced (on a linear basis) the potentially higher capital requirements can be expected to re-emerge as an issue in absence of further action. These concerns have led the ABI to conclude that that “we do not believe the PRA’s approach to the implementation of TMTP is practical or flexible enough”.
172.The Transitional Measure on Technical Provisions was discussed briefly by the regulator in oral evidence to the previous Committee; Sam Woods noted that the PRA have allowed it to be dynamic so it can be recalculated as risk-free rates move. Sam Woods also acknowledged the ABI’s request for simplification, stating “I am absolutely in the market for simplification ideas; if they can bring something forward, that would be great”.
173.The Bank of England published an exposure draft in October 2016 (CP 47/16) on the calculation of TMTP and replies were due in by 25 March 2017. This focused mainly on how firms should take account of changing circumstances and assumptions in recalculating the TMTP as time progresses. The conclusions were published in a supervisory statement in April which was then updated in May, reflecting the complexity of the subject.
174.The PRA’s consultation on the calculation of the Transitional Measure on Technical Provisions was welcome, but was limited in scope. Broader issues remain to be addressed including: (i) whether and how the transitional benefit should be allowed to run-off over time; (ii) whether the principal causes of the increase in reserves that accompanied the introduction of Solvency II can be “corrected” when determining the future UK solvency regime issue; (iii) reducing the cost, particularly over the potential need to maintain dual models for sixteen years in case a re-measurement is required; and (iv) developing an approach which is practical to implement.
175.In Solvency II, the rules for contract boundaries (to identify where one contract ends and another begins in cases where an insurance contract includes options for extending it, for a separate premium) follow a legalistic rather than economic approach. Solvency II therefore ignores the conventional economic and accounting convention of matching income to costs. This is a key issue because it differs both from insurers’ current practice and the recently published international accounting standard, IFRS 17.
176.KPMG told the previous Treasury Committee that:
“In our view, the Solvency II rules on contract boundaries are artificial and are not economically based. The latest drafting for the forthcoming IFRS 4 Phase 2 (now known as IFRS 17) financial reporting standard is much more economic in this area”—ie it allows for future premiums to be projected, subject to a best estimate of future cessation rates, and then subjecting this best estimate to stress testing in the capital requirements. We believe that the solvency approach should be similarly economically based.”
177.They cited reinsurance as an example of the distortion it creates.
“Currently, a consistency issue arises where reinsurance protection is purchased to cover a class of business over a period of time—typically one year from the date of reinsurance purchase or renewal. The full cost of purchasing the reinsurance is taken into account, but the benefit of the reinsurance can only be recognised in respect of the business in force at the valuation date in question (as opposed to the full expected amount of business). Given that most reinsurance programmes are renewed on an annual basis, this is causing an artificial strain.”
178.One consequence of this distortion is that firms may need to maintain an alternative model in order to show the true economic picture. As the Investment and Life Assurance Group said in its evidence to the previous Committee:
‘’A number of areas of the Solvency II calculation are not considered to be truly economic, for example the implementation of contract boundaries. This means that even insurers that are using an internal model may need to maintain an additional model to allow for what they would consider the true economic impact to be’’
179.In Solvency II, the basic capital requirement is that firms hold enough capital to be 99.5% confident of surviving extreme events over a 12 month period. The risk margin is an additional margin which increases the amount of the provision to a level that is supposed to represent its market value, although clearly there is little or no market for insurance liabilities
180.Most firms did not challenge this concept but preferred to advocate remedies to a rigid interpretation of the one year concept. However, Prudential’s evidence to the previous Treasury Committee outlined why they disagreed with the core one year at risk concept.
“There are major problems in applying this construct to insurance. Given the length of a life insurer’s liabilities, it is highly likely that the construct of 1 Year value at risk will fail to capture the path dependency of risk factors beyond the 1 Year period. This is a highly significant constraint given the number of dynamic management actions which can be taken over the lifetime of the liabilities, the effect of which would then need to be re-modelled.’’
181.Lloyd’s of London cited the IMF’s suggestion that for non-life insurance, the framework’s one-year horizon is potentially insufficient for some of the long-tail risks insured in the London market, and went on to say how it will compensate for the limitations of the one year approach:
“It should be noted that Lloyd’s bases member capital setting on the ultimate Solvency Capital Requirement (SCR), i.e. adverse development until all liabilities have been paid, rather than the one-year SCR, because it believes that this is more appropriate.”
209 Sam Woods giving oral evidence to the Committee on his appointment as Deputy Governor for Prudential Regulation and Chief Executive of the PRA 19 July 2016, HC567, Q71
214 For clarity, this terminology does not relate to the concept of ‘transitional measures’ for the implementation of Brexit.
25 October 2017