A devil in disguise?

As the current labour force looks set to fall short with their retirement funds, the moves to extend Solvency II to pension funds are defended. This paper looks at the facts

Providing retirement income for the ageing populations is a major challenge for Europe in the coming decades. As an economist I see no way to escape the fact that the current labour force must save more for their future pensions. Expanding occupational pension schemes is a well-known way to do that. However, I doubt that workers are willing to defer wages unless they are confident that they get value for their money.

Lack of confidence may be considered as an agency cost. Beneficiaries may be concerned that pension providers - due to their higher information level or better bargaining positions - may give priority to their own interests. In defined benefit schemes one should expect that the management is better equipped than the beneficiaries to judge if the schemes are fully funded and adequately capitalised. Better bargaining positions may be the result of discretionary features in contracts - for example, with regard to conditional indexation or bonuses.

Financial regulation and supervision should mitigate agency costs and thus add value to the relevant contractual relationships. The prudential part of financial regulation and supervision is concerned with protecting depositors, policyholders and other relevant owners of credit-sensitive contracts against risk-shifting.

Risk-shifting may occur because depositors, policyholders, and so on assume part of the downside in a risky investment strategy, while owners of a limited liability firm will get the full upside. In economist jargon, risk-shifting may occur because the owners of credit-sensitive contracts have written an option to default to the owners of the firm. Risk-shifting occurs when the value of the option to default is higher than anticipated - if not promised - when contracts were entered into. In that case, owners of credit-sensitive contracts get less value for their money.

Households and firms want access to assets that are almost default-free, that is, assets where the value of the option to default is close - but not equal - to zero. Writing contracts with many financial firms in order to diversify default risk would be prohibitive in terms of transaction costs. Hence the core of prudential regulation is to set a cap on the value of the option to default - properly defined. Prudential supervision is to observe if firms comply with this requirement and take actions if not. Setting a cap on the value of the option to default implies that firms should have higher financial strength when they increase risk assumption - for example, by investing in more volatile assets.

Pension contracts are a subset of financial contracts and hence share some common characteristics including agency costs. However, for obvious reasons they are also a subset of wage contracts. Pension contracts are thus less regulated on the European level than other financial contracts. They are mostly covered by national social or labour law or considered to be in the domain of the social partners. The exceptions are when pension contracts involve commitments from financial firms - for example, banks or insurers - or from occupational pension funds with actives in several EU-countries.

The occupational pension funds that are subject to European regulation - by the IORP directive - are with regard to prudential aspects in many ways regulated as life assurers. Consequently it is natural to ask whether these funds should continue to be subject to equivalent prudential regulation as insurers when the coming Solvency II is implemented. If the answer is yes, one may ask: do we need to do more and subject all occupational pension funds to prudential regulation and supervision equivalent to Solvency II? I will argue for a yes to both questions.

Solvency II is the coming European prudential regulation and supervision of insurers - including reinsurers. Earlier this summer, the Commission issued a draft framework directive incorporating Solvency II into European legislation. In my view, Solvency II is a major step forward. It will bring prudential regulation and supervision of European insurers in line with prudential regulation and supervision of European banks and securities firms, which has recently been updated by the implementation of the Basel II accord. The current European prudential regulation of insurers is completely outdated - first of all because it does not take investment risk into account.

The core of Solvency II is risk-sensitive financial strength requirements for each insurer based on a market-consistent valuation of its assets and liabilities. For practical reasons, risk sensitivity is measured by probability of default and not by the value on the option to default. However, qualitatively the results are equivalent. All insurers are required to have enough financial strength to keep the estimated probability of default within one year below 0.5%. Default is here defined as assets less than liabilities - that is, negative own funds.

Financial strength is mainly measured in terms of regulatory capital - that is, own funds after some corrections plus subordinated capital. If the own funds of an insurer is less than the Solvency Capital Requirement (SCR), action will need to be taken by the firm and the supervisor in order to restore the situation. The level and speed of action required depends on how much own funds are lower than the SCR. If the own funds become lower than the Minimum Capital Requirement (MCR) the ultimate supervisory action is required - that is, authorisation is withdrawn.

The SCR is calculated in two steps. The first is either made by using a standard model or an internal model approved by the supervisor. The second step is an 'Own Risk and Solvency Assessment' by each insurer. If the firm - or the supervisor - believe that more capital is needed than required by the SCR from the first step, they increase the SCR. In that way even insurers with unusual risks should not be in a situation where the value of the option to default is too high.

Financial strength requirements are complemented by risk management requirements in Solvency II. Even a very high financial strength may be inadequate if the risk management is flawed.

Market-consistent valuation of assets and liabilities implies that assets and liabilities are reported at the prices they are currently traded in financial markets. This is also known as fair value and is in line with International Financial Reporting Standards (IFRS). Some (parts of) assets and liabilities are not regularly traded. In such cases - notably technical provisions - prices need to be estimated as if these assets and liabilities are regularly traded. I will spare the practical details - but note that as a consequence insurance obligations may be subject to similar interest rate risks as fixed income bonds. This will be most relevant for insurance obligations with financial guarantees and a long duration, for example, life or pension insurance obligations. As a general rule, the longer the duration, the higher the interest rate risk.

Use of market-consistent valuation of assets and liabilities implies that even if regulatory capital is not too much below MCR, the firm may enter into solvent run-off if it is not feasible to transfer the insurance obligations to another firm.

Putting more emphasis on risk-sensitive financial strength requirement makes it natural to abandon the current system of admissible investment for insurers and introduce 'the prudent person principle' in Solvency II. That will enable insurers to invest in additional asset classes. The basic requirement is that the asset mix is considered to be prudent, taking liabilities and financial strength into account. In other words, when insurers hedge less of the interest rate risk on their insurance obligations they need more financial strength. The same applies when insurers increase their risk with regard to stock prices, liquidity or in terms of concentration.

The prudent person principle already applies to the occupational pension funds covered by the IORP directive. If such pension funds will be subject to prudential regulation equivalent to Solvency II, they will able to invest in the same asset classes as today - for example, private equity. Hence, provided they have the required financial strength they need not change their asset mix. If they do not have the required financial strength, they need to change their asset mix - but not more than insurers with identical assets, liabilities and financial strength.

On this backdrop, one should expect that subjecting occupational pension funds to prudential regulation equivalent to Solvency II should be uncontroversial. During my four years as CEIOPS chair, I have, however, learned that it is not the case.

Firstly, it is claimed that occupational pension funds are different from insurers. That is undeniable in general terms. However, I will be surprised if there are pension obligations that may be assumed by occupational pension funds but not by insurers. According to European regulation, there may be no regulation of products provided by insurers. When designed properly, Solvency II should be able to deal with all kinds of insurance obligations - including pension insurance obligations.

Secondly, it is asserted that Solvency II is indeed poorly designed because it exaggerates the risk of holding stocks. Even though stock prices may be volatile in the short term, they generate an expected return above the risk-free rate of interest. This means that in the long run - where only average returns matter - it is (almost) risk-free to 'hedge' nominal obligations with stocks. The former part of this argument is in line with modern financial economics. The latter is not. Yet Solvency II is supposed to "fully reflect the latest development in ... actuarial science and risk management".

The basic argument against the assertion is one of arbitrage. If true, there is no scope for equity premiums. Stock prices would regularly be bid up until equity premiums are negligible. Modern financial economics are, of course, not perfect. However, if it has any bias it probably underestimates risks - including the volatility of stock prices.

So what if insurance obligations are not nominal - but indexed to inflation or wages? Such obligations are rarely fully hedgeable. It is likely that there is some correlation between stock market returns and wage or price inflation. However, I would expect the correlation is higher for returns on commodities, some types of real estate and, naturally, price-indexed bonds.

Thirdly, it is said that promises from occupational pension funds may be subject to renegotiation. Promises may only be intentional and may be written down if necessary to avoid negative own funds. For this reason renegotiable pension promises need to be supported by less financial strength than otherwise identical non-renegotiable promises from insurers. I share this view to the extent promises from occupational pension funds in fact are subject to renegotiation. However, this is not the real issue. A fully renegotiable promise is the economic equivalent of a unit-linked insurance contract where the liability equals the value of the assets in the unit. The financial strength requirement on such a contract should, according to Solvency II, be independent of the asset mix. In case the promise is partially renegotiable, it is economically equivalent to a combination (weighted average) of a fully renegotiable and a non-renegotiable promise - and the financial strength requirement on such a contract according to Solvency II should also be a combination. Pension promises with conditional indexation and life insurance contracts with discretionary bonuses are economically equivalent to partially renegotiable contracts. In other words, the important matters are the assets, liabilities/promises and financial strength of pension providers - not what they call themselves.

Finally, it is mentioned that in order to keep promises, occupational pension funds may increase contributions from sponsors. This means their financial strength requirements should be lower. I agree to the extent that the assumptions are correct. However, the value of this put option issued by the sponsor may be limited if the value of his own option to default is already high. In other words unless the sponsor constitutes a low credit risk - for example, due to a high rating - the impact on financial strength requirements of taking additional contributions from the sponsor into account should only be marginal.

Actives tend to have three major assets: their homes, human capital and accrued pension benefits. Due to skill specificity, the earning power on (value of) the human capital may be substantially reduced if one is laid off - for instance, due to the default of the employer. Hence it may be risk concentration if the accrued pension benefits are also heavily exposed to him.

I have argued for why occupational pension funds covered by the IORP directive should be subject to prudential regulation equivalent to Solvency II. I will now turn to my second question - why should this be extended to all occupational pension funds in Europe? Let me start by simply asking: why not?

Prudential regulation and supervision equivalent to Solvency II would imply less asymmetric information. Foreign workers would be less concerned to contract with firms that offer membership of occupational pension funds as part of their pay benefit packages. Pension portability would be facilitated too in a system where all pension providers use market-consistent valuation of assets and liabilities and are subject to identical risk-sensitive financial strength requirements.

If one, for a moment, ignores history and considers how to design occupational pension funds from scratch, it would be difficult to argue that they should not be subject to prudential regulation equivalent to Solvency II. What about the price impact of subjecting existing occupational pension funds in Europe to prudential regulation equivalent to Solvency II? Admittedly, this is a potential problem if the required financial strength -taking current assets and liabilities into account - is far above current financial strength. If contribution rates or benefits are non-renegotiable the asset mix will have to change. This may have a price impact. This may also be the case even when contribution rates are renegotiable. In this case, the impact will be on the value of the sponsors. If these price impacts are considered not to be politically acceptable, they may be mitigated by creating transitional mandatory guarantee schemes in relevant countries.

Let me finish with a remark on supervision. Previously I have focused on regulation. However, if we want convergence of supervisory practices within Europe, convergence of regulation is not enough. We need convergence of powers, resources and the objectives of supervisors too. Solvency II also contributes to this - with regard to powers and objectives. Supervisors are also required to be more transparent and accountable - including how supervisory heads are appointed or dismissed. However, one element is left out. This is that supervisory decisions should be free from political interference, just as monetary policy decisions are by central bankers.

I did not touch upon conduct of business regulation for occupational pension funds. This is the regulation of how occupational pension funds should treat actives and beneficiaries. The same applies to governance of occupational pension funds. The reason for this is not lack of interest. It is simply because I believe that this part of pension contracts should continue to be regulated as wage contracts - that is, by national law or by the social partners.

Henrik Bjerre-Nielsen has been Danish FSA director general since 1996. In that capacity he has chaired several international committees - including CEIOPS 2003-07, CESRfin 2001-03 and IAIS Reinsurance Subcommittee 1999-2003. He was on the board of directors of Danish pension fund ATP 1986-94.

References

BCBS (1995)

An internal model-based approach to market risk capital requirements

BCBS (1996)

Amendment to the capital accord to incorporate market risks

EU Commission (2002)

Risk models of insurance companies or groups

CRO Forum (2005)

Principles for regulatory admissibility of internal models, June 10, 2005

CEIOPS (2005)

Consultation paper 7 - Answers to the second wave of Calls for Advice - Solvency II, October 2005

CEIOPS (2006)

Consultation paper 20 - Draft advice to the European Commission in the framework of the Solvency II project on Pillar I issues - Further advice.