Con Keating on the folly of applying Solvency II rules to defined benefit pension funds.

The European Commission argument appears to be that annuities and defined benefit (DB) pensions are essentially the same thing and that, in some way, this justifies the regulation of their providers in the same way. This is very strange. It is akin to arguing that, because I wish to compete in the Olympics 100 metres final, all other entrants should be handicapped to my level of incompetence. In the context of regulation, it is not seeking to eliminate or ameliorate any harm to the consumer, but rather introducing barriers to entry and provision.

It is most important to define the markets in which we are concerned about competition. The occupational pension scheme offers pension benefits to its employees - it does not offer them to the public at large. The insurance company provider offers them publicly. An insurance company may offer DB pensions to employees of private sector companies as part of their employment contract, but this is a matter of administrative convenience for the sponsor employer.

This is not the argument an employer voluntarily offers pensions and that these schemes are not organised for profit. A contract is, once consummated, a contract, whether offered voluntarily or not. An occupational scheme may be managed without profit in mind, but this is not relevant - schemes are usually managed and organised to minimise the cost of pension provision to the employer. In fact, then, both insurance providers and companies offer pensions with profit in mind. For the insurance company, the pension is its productive output and its profit based solely on this. For the private sector company, the pension is part of a contract with its labour force - this is part of its cost of production of other goods and services. The effects of regulation on competition would be felt in different markets. There is no reason to believe the costs of financial regulation of schemes would be similar in effect across the various markets for goods and services in which sponsor employers compete.

The Commission appears to regard both pension schemes and insurance companies as financial institutions and seems to be believe this merits equal regulatory treatment. It does not. A corporate treasury may provide financial services to other members of its group, but we do not require it to operate and be regulated as if it were a bank, precisely because it is not offering these services to the public. Indeed, the identification of an occupational pension scheme as a financial services enterprise is erroneous. It is no different from the corporate treasury case; it provides services to the employer though for the ultimate benefit of employees. Regulation of UK occupational schemes as if they were insurance companies offering pensions is inappropriate precisely because these regulations would constrain employers and distort investment, innovation and competition in other markets. It would also distort competition across the various markets in which occupational sponsor employers operate.

In the UK, schemes have recourse to their sponsor employer, through the balance of cost guarantee. This means they have access to their employer's equity to the extent necessary to pay pensions. A pension is a claim on future production. The employer sponsor is a producer and may credibly contract on this future production - this is exactly what it does when issuing debt or equity securities. This, in turn, means the investment opportunity set available to the occupational scheme is wider than for an insurance company; the insurance company must buy its claims in financial markets. This distinction is immediately obvious in the case of a book-reserve scheme. Here the claim is solely upon the sponsor employer.

Organisation of pension provision through financial markets is less efficient than occupationally. This arises in part because the pension scheme has recourse to the sponsor's production, but, in larger part, it is because financial markets are far more volatile than this. The difference is important: it is of the order of 1-2% a year in return.

It is worth considering the incentives associated with funding as a security mechanism. If we require full funding - that is to say that the contribution made in respect of a new pension award is actuarially fair - we reduce and perhaps eliminate any incentive to the sponsor employer to offer this pension. The discount rate used in this calculation is important, but not discussed here. If we require overfunding as a security mechanism, this can be seen as an attempt to eliminate any dependence upon the sponsor to some level of confidence that is socially acceptable for a form of consumer protection, former employee protection. This is the buyout value. This DB pension has now acquired one attribute of a defined contribution (DC) arrangement - recourse to the sponsor has no value.

For a corporate sponsor, this overfunding provides positive disincentives to provision. The drain on working capital is larger than the cost of providing immediate wages. The length of time for which overfunding is required is actually longer than most companies' capital expenditure and operating investment cycles and likely to remain so - it should rationally be considered a pure sunk cost by them. It is not available to them within the horizons of business operations and would not be capable of capitalisation through borrowing or equity issuance. The result is that companies can be expected to cease occupational DB provision.

The idea of own funds is deeply misleading - no institution possesses own funds. They are the property of some other person. In the case of insurance companies, these are usually shareholders. They have put capital at risk in the expectation that the management of financial asset and liability contracts can be effected profitably.

The question of ownership of such own funds for a UK scheme is material. Any surplus remaining after discharge of all liabilities is the property of the sponsor employer. Its position is that of a shareholder with unlimited but finite liability. It should also be understood that it would not be feasible for such own funds to be distributed equitably to pensioner members since we do not know, ex ante, how long any individual member will live.

The principal problem with overfunding, however, is collective. Overfunding is a solution to idiosyncratic risk, the possibility that a particular sponsor employer will fail. Even if funded at 100% of the technical best estimate of liabilities, there is a 50% likelihood the scheme will fail before all liabilities are discharged, which is socially unacceptable. The expected loss across all schemes is, however, zero. To require overfunding on the scale necessary is economically very inefficient, and the effect of this inefficiency would be experienced across the entire UK private sector, not merely in the financial sub-sector.

The fundamental nature of the problem is that funding as a security mechanism is operating on the consequence rather than the likelihood of sponsor insolvency. Indeed, funding can often be entirely counterproductive in this regard, actually increasing the likelihood of sponsor insolvency. This is not true in the case of insurance companies. Here, the purpose of funding is precisely to lower the likelihood of insurance company insolvency.

The likelihood of sponsor insolvency is determined only in part by the financial resources it holds. Of greater importance for most is their competitiveness in the markets for their goods and services, and these are not, for most, the markets for financial services. The proposal in any event is that these producer companies should be deprived ex ante of an amount sufficient to cater for an ex post problem.

In fact, it may trivially be demonstrated that the level of financial resource necessary to lower employer sponsor insolvency likelihoods differs markedly from that necessary for an insurance company or financial institution, which is subject to the vagaries of financial market prices, its principal non-labour factors of production. Ex ante, far lower risk buffers are necessary for private non-financial companies than for financial institutions. Post insolvency, a pension scheme, in its new standalone role, faces the same level of risk and requires capitalisation to the level of an insurance company.
The solution to this problem lies not in requiring ex ante funding of a scheme, nor in applying credit constraints upon companies wishing to offer occupational pensions. It lies in providing comprehensive pension indemnity assurance - cover against the consequence of scheme insolvency - to the scheme.

Such insolvency cover does not sit well in the public sector. It may be offered by private sector financial institutions. It should be noted that this is insurance, not participation by schemes in some state-sponsored mutual compensation fund. As insurance, it would properly be regulated under Solvency II, where the funding regulations lower its likelihood of failure and are unobjectionable in principle.

Con Keating is head of research at Brighton Rock