99.Insurance is a long term business, particularly for many life products—“insurers could have customer relationships for 50+ years”. And insurance works on an aggregate basis as “a large number of uncorrelated idiosyncratic risks…diversify to reduce the overall level of risk”.
100.UK insurers have sold large volumes of annuity products, which provide a guaranteed stream of income to the customer until death. They also provide pension products with guaranteed returns and equity release products, where the insurer takes whole or part-ownership of a homeowner’s property in exchange for providing a regular income, and allowing the homeowner to remain in residence.
101.Life companies have been able to offer these products as the cashflow profile of a large portfolio of pooled liabilities is highly predictable. They have been able closely to match the liability cashflows by investing in long term, relatively illiquid assets where the return includes an illiquidity premium to the investor (e.g. corporate bonds, equity release mortgages or infrastructure projects). After allowing for default risk, these assets provide higher returns than other more liquid assets, enabling more competitive rates to be offered to consumers.
102.While some European countries have similar product lines, they are not on the same scale as the UK, making it unique in providing long term savings of this type. This was problematic because the original Solvency II proposals did not allow credit for the illiquidity premium. Both “HMT and the PRA played a key role in winning an important concession for the UK insurance industry”, namely the Matching Adjustment (MA), which allowed such credit to be taken, by adjusting the discount rate used to calculate the present value of future liabilities so that when an insurer invests in assets using a “buy and hold” strategy, it is less exposed to short term market volatility, with the risk that the market value of its bond assets may fall and can only be sold for less than expected. However the design of this Adjustment “incorporated stringent restrictions and constraints required by a sceptical non-UK audience designed to prevent it being ‘misused’”. Its particular configuration has led respondents to argue that it is too restrictive and “over-engineered”—reflected in the fact that “with the exception of Spain, the UK is the only country to use the Matching Adjustment”.
103.The IFoA quoted several problems with the design of the Matching Adjustment, and these were repeated in other submissions. Most notably, some asset classes are ruled to be ineligible, “such as Equity Release Mortgages (which are considered to be appropriate assets to back annuity liabilities), or commercial property”. These are either “excluded from investment mandates which leads to lower levels of funding in the wider economy (e.g. for some infrastructure assets),” or firms have to put in place artificial constructs to convert them to eligible assets.
104.Even where the Matching Adjustment is allowed, the rules (and the PRA’s interpretation of them) has added cost and complexity. For example, there is an “unnecessarily onerous approval process for using the MA…inappropriately harsh consequences of breaches in the Matching Adjustment (no matter how minor)…and a requirement to use allowances for default and downgrade risk set by EIOPA (the ‘fundamental spreads’). These can differ from a firm’s underlying view of the risk for the assets held.” The IFoA concluded that this has “increased the complexity and cost of maintaining annuity portfolios. Some of this cost is likely to be reflected in annuity pricing, to the detriment of consumers.”
105.It is unsatisfactory that significant monetary and time costs are being incurred as insurers create artificial structures to “get round” the rules—for example in restructuring (reasonable) assets so that their cashflows meet the exhaustive qualifying criteria set out in the rules. This causes consumer detriment to the extent that the costs might be passed (indirectly) to consumers.
106.In developing the future regulatory model, specific efforts should be made to avoid creating situations where artificial structures are encouraged to achieve an appropriate regulatory treatment for any class of assets or liabilities.
107.Although the PRA has correctly attempted to reflect the long term nature of assets matching long term liabilities via the Matching Adjustment (see below) the difficulty of achieving eligibility, and the additional problems caused by the penal Risk Margin (see Chapter 6) have caused several insurers to curtail writing long-term savings products. In addition to reducing choice and value for consumers, this transfers risk to the State if and when individuals run out of money in old age.
109.In oral evidence to the previous Committee, Julian Adams of Prudential took up this general theme, noting that Solvency II is not working in the way envisaged:
“…because of the excessively rigorous implementation of rules around cash flow matching. For instance, equity release mortgages, residential or commercial mortgage-backed securities and commercial mortgage lending all have uncertain payback periods. Yet we would argue that, as an asset class, each of those would be appropriate in aggregate to match the liabilities that we would be looking to hold. As a result, it has restricted our investment in those assets, and you could make similar arguments for infrastructure lending as well.”
110.The oral evidence sessions held by the previous Committee also noted that if an insurer’s base currency is sterling, then they cannot hold and get the full benefit from matching their liabilities with and dollar assets because of the standard formula. This means that an insurer is “penalised for being a sensible and appropriate risk manager”, as they still lose out, even though they may have a perfect match of assets and liabilities. Innovation is also discouraged as the regulator does not keep up with the creation of new asset classes as fast as the industry creates them.
111.The “cumbersome nature and rigidity of the framework” has caused a “real global phenomenon” as insurers are forced into the selection of similar assets for the benefit of “limited additional policyholder protection”, at the cost of reducing long term investments by insurers. Legal & General argued that this has resulted in insurers losing out to non-insurers with lower capital costs:
“The table below compares the capital requirements for a 1 year and 10 year duration A-rated corporate bond:
Solvency II standard formula credit risk capital requirement
Basel credit risk capital requirement
Furthermore, for pension schemes there is a clear advantage for non-insurers to hold those liabilities relative to insurers. Insurers have to provide capital under the assumption that scheme members rank alongside policyholders, whereas non-insurance companies do not have the same constraint”.
112.Absent Solvency II, there are a “number of areas of socially useful investment” where insurers would be the “natural investors”, such as infrastructure projects, student loans and retirement funding. However, Solvency II appears to be hampering both investment from outside the EU, alongside investment within the EU. For example, Nigel Wilson of Legal & General stated that:
“There is billions of capital available outside of the UK that would come into the UK that could be invested with a better regulatory set up.”
Meanwhile Aviva noted that “EU insurers may have more capital tied up as regulatory capital than an equivalent insurer operating in or from a different country outside the EU”.
113.In oral evidence to the previous Committee, Sam Woods noted that “insurers are natural holders of long-term assets” and the PRA “want to encourage that”, asserting that the Matching Adjustment provides this mechanism. Accordingly, the regulator has “handed out…£59 billion of Matching Adjustment and £1 billion of Volatility Adjustment” (to be discussed in Chapter 8). Furthermore, the regulator argued that the industry had “put their case much too strongly” because there are other factors to be considered besides the rules when assessing the lack of investment in infrastructure by insurers, including the availability of such assets. Elaborating on this point, Sam Woods noted that the “rough average of the intentions of those firms across the next four years is to raise” the share of illiquid assets backing annuity business “from 25% to 40%”. When asked whether insurers would have invested more in infrastructure under the old rules than they will under the new rules, Mr Belsham responded:
“the availability of the assets is the problem. It is very difficult to find a pipeline of attractive longterm illiquid investmentgrade fixedinterest assets”.
114.Legal & General responded on this specific point, noting the possibility of securitising some assets, but pointed out that this was a costly process, and that there was no certainty that the PRA would be comfortable with the structures that firms put in place.
115.In terms of taking steps to address the problem, David Belsham noted that “the Solvency II rules are very clear that the Matching Adjustment is only allowed on assets with fixed cash flows” which “cannot be changed”. However, he added that “the PRA is allowing the firms to carry out a restructuring to get a senior tranche of that asset, which could be 8090% of the cash flows, to be eligible”, so while they are “stuck with the rules as they are written”, they are “trying to operate them in as flexible a way as possible”. Sam Woods also commented that:
“There are some bugs that need to be ironed out and there are some design features that I would like. On the Matching Adjustment specifically, if I had complete control of it, I would probably do something slightly different. I would want to take account of unexpected defaults and I would probably be a tiny bit more flexible around this notion of “fixed”. That is how I see it.”
116.Despite insurers facing a “more complicated process” compared to the old ICAS regime, David Belsham felt that the level of investment in infrastructure assets is “very similar”. Furthermore, Sam Woods argued that the Solvency II regime had not had an impact on annuity rates, with these instead being driven by risk-free rates and corporate bond spreads.
117.UK firms believe that Solvency II makes it harder for them to invest in longer-term illiquid assets, such as infrastructure and equity release mortgages. This is a concern as the disincentive could have negative economic consequences and act as a restraint on UK plc.
119.The Matching Adjustment has given some relief to the industry in that it attempts to reflect the long term nature of assets matching long term liabilities. Nevertheless, the Adjustment is a “workaround” solution, bolted on to the core Solvency II rules, which is cumbersome, and unnecessarily constraining. For these reasons, the PRA needs to conduct a fundamental review of the Matching Adjustment and its eligibility criteria, in order to achieve a more principles-based approach to the Matching Adjustment.
120.The regulator and industry should work together to ensure that opportunities are taken to develop a more effective long-term approach to the treatment of insurers holding long-term assets to match long-term liabilities.
121.Infrastructure is one such asset. Others may include equity release products. It is concerning that, when asked by the previous Committee, the PRA appeared to be of the view that the problem is down to a simple lack of availability of such assets, whereas the industry suggest that it is the PRA’s interpretation of the term “fixed cashflows” which limit the number of appropriate long-term illiquid investment grade assets.
122.A robust process will still be needed for approving and maintaining any approach that gives credit to an illiquidity premium. Evidence suggests that there is scope for more pragmatic rules to be agreed, with more flexibility alongside strict criteria to avoid the need for artificial restructuring of firms’ cashflows for purely regulatory purposes. The Committee would expect the PRA’s progress report (requested at paragraph 90) to address the extent to which this is the case, and set out what action it proposes to take.
25 October 2017