Accounting and risk-based regulation for pensions remain contentious. The accounting standard is mixed attribute in nature using market prices to value assets and net present values for liabilities, based on an exogenous market-observed discount rate. Clearly, this was heavily influenced by the funded model of defined benefit pension schemes. Regulation and regulatory intervention operate on current balance sheets reported in this manner.
Again we use a stylised model; consider a perpetual pension payable of £10 p.a. due from a scheme which has a current endowment of £100 held in cash and no future income. It is immediately obvious that this scheme will become insolvent, unable to pay any pension beyond the tenth year. This is the point of equitable insolvency, the inability of the scheme to meet current liabilities.
The net present value, under different discount rates, of scheme liabilities and solvency ratio for this scheme are shown in Table 2 together with the regulatory interventions to restore perceived full funding based upon balance sheet solvency.
Though the fundamental position is the same in all cases with the scheme surviving until the tenth year equitable insolvency, the reported solvency and regulatory intervention vary wildly. Depending upon the choice of discount rate, the scheme is in surplus, deficit or perfect balance, though, in fact, it will fail after ten years. In the 10% discount rate case, regulatory intervention reduces the scheme to a pay-as-you-go basis after one year though it has assets of £90 at that point in time. In the 5% discount rate case, £100 is required immediately and the scheme then continues beyond that point in time on a pay-as-you-go basis. By contrast, under a 15% discount rate no intervention is made until three years have elapsed and then it is that fourth year’s shortfall of £6.67 which is required. Thereafter, it continues from year to year on a pay-as-you-go basis.
It is clear that, under this accounting standard solvency is, at best, a very noisy signal of equitable insolvency, and may even be positively misleading. The discount rate used contains an implicit required survival time. There are, respectively, 20, 10, and 6.67 years for the 5%, 10% and 15% discount rates; the lower the discount rate, the greater the implicit survival time. The solvency ratio is merely the ratio of the actual to the implicit survival time. It may be considered no more than a risk buffer in time; lower discount rates require larger buffers extending the time between balance sheet and equitable insolvency.
Closure of a scheme to new members is also interesting in this regard. If we introduce a new member contributing, say £2 annually over a 45 year working lifetime, then clearly we extend the time to equitable insolvency to year 13. Closing a scheme to new members reduces the time gap between balance sheet and equitable insolvency. The balance sheet is also rather interesting – see Table 3.
This balance sheet uses a projected benefit obligation view of the scheme and uses a 10% discount rate. The present value of the full perpetuity, just £1.51 as it commences 44 years in the future, is shown. The current cash assets, which have been augmented by the contribution received and the net present value of future contributions are also reported. The scheme appears comfortably solvent even though it will fail equitably in year 13. Even in that thirteenth year it will have an asset, the present value of future contributions payable, £19.14, available to it. In financial theory, this asset could be borrowed against and further extend the time to equitable insolvency; in the jargon of academic financial theory it is ‘pledgeable’. This ability to borrow against future revenues is a central definition of liquidity, to which we will return later.
However, in this case it should be recognised that future contributions will not be pledgeable in the absence of credible sponsor support. As soon as the employee member recognises that their perpetuity may not ultimately be paid, they will cease making contributions; a default event would make this common knowledge. In turn, as the liquidity supplied by capitalisation of the pledged contributions is exhausted after two years, a creditor will also recognise that it will face the equitable insolvency problem after that time (Having received, by then, repayment of only a small part of its advance) and that credible sponsor support is then necessary.
Of course, the current accounting standard is based upon the accumulated benefit obligation. This means that, with the payment of just the first year’s contribution, only 1/45th of the present value of the current employee’s perpetuity is recognised (£0.034) and that liabilities will be stated as £100.034 and assets at £102.00 – the scheme appears solvent.
The current accounting standards are often supported by the claim that they make schemes more comparable among themselves, yet here we have schemes which fail after ten and thirteen years but have very similar balance sheets. To reinforce this point, we might alternately consider the £100 of assets to be held in an income bearing demand deposit bank account paying 2% per annum and only the original perpetuity to be outstanding. Here the balance sheet shows a liability of £100 and assets of £102, yet the scheme will experience equitable insolvency in year twelve. This trivial illustration offers the insight that, if and only if the assets offer the same future return profile as the discount rate function will this ‘mixed attribute’ accounting standard offer a true and fair view of the solvency of the ongoing scheme. Note this is the same return profile, not the same risk-adjusted return profile.
The standard is reliant upon a particular interpretation of financial theory: that equal current market prices of all assets must reflect similar risk-adjusted future returns. This is tenuous in the extreme and there is a very substantial volume of evidence against this hypothesis. Moreover, regulators have shown a reluctance to accept this hypothesis; if they did, why should they be as concerned as they evidently are by the asset allocations and investment strategies followed by pension funds?
This balance sheet view is often justified by the assertion that this is sound risk management. In an inter-temporal context, it is not. The time line for pensions is all-important. The need for a pension arises in the first place from the asynchronicity of labour income and consumption. A pension is simply an income in retirement; it is not savings for retirement, which will require realisation to generate income. Sound risk management would be based upon survival times, that is to say the time to equitable insolvency, predicated on the current state of affairs.
It is notable that, when deficits are stated as proportions of the present value of liabilities, these survival times are an inverse function of discount rates. Table 4 below shows the relation between discount rates, current proportional deficits and the survival time of a scheme in the absence of any further contributions or investment income.
The effect of lower interest rates is to increase the survival time buffer; the time until explicit sponsor support is necessary. Regulation should reflect this characteristic but does not; it uses some simple heuristics, such as deficit repair schedules with terms beyond ten years requiring regulatory approval. All of this is reminiscent of Charles Goodhart’s condemnation of banking liquidity buffers by the analogy of the railway station taxi. Here a weary traveller arrives late at night at a remote and desolate railway station and finds to his relief there is a solitary taxi waiting at the cab-rank. But when he attempts to hire it, he is told that he cannot, as local regulation requires there to be a cab at the station rank at all times. It also raises questions as to why scheme deficit statistics receive so much publicity – as is evident from table 4, a 10% deficit at 10% discount rates represents a shorter survival time than a 50% deficit with a 2% discount rate.
It is frankly doubtful that any pensioner would be able to obtain a Court insolvency judgement based upon the argument that the scheme would likely prove to be unable to meet a pension payment fifty or more years in the future. Apart from anything else, the accuracy of the balance sheet is vital and based upon numerous assumptions. It is critical that both asset and liability values are consistent and correct.
However, under regulatory pressure, the practice has developed of using conservatively biased valuations in the pursuit of ‘prudence’. This has several unfortunate effects. It renders the scheme accounts inaccurate, more volatile and increases the difficulty of achieving accurate hedges should these be desired. ‘Prudent’ behaviour is better described as rational and well-informed. This is the unbiased expectation, the best estimate of technical provisions.
Though risk buffers are discussed in more detail later, the discount rate can be seen to function as a form of risk buffer; the lower the discount rate, the longer the time to equitable insolvency. The old adage “time is money” contains more than a grain of truth.
It should be realised that the affordability of a liability, a pension or whatever, is a determinant of the security of that liability. The cost of a pension is, as shown earlier, fixed at award. The security of that is determined by the sponsor’s wealth and earnings on that wealth. If we have a company’s rate of earnings above the pension cost rate, then the affordability of that liability is determined by the ratio of the rate of (pre-tax) earnings to that cost rate.