This essay is motivated both by the Pensions Regulator’s consultation on its proposed code of practice for scheme funding and by papers published in recent years by the Institute and Faculty of Actuaries, such as “Actuarial valuations to monitor defined benefit pension funding”.[i]
There have been calls for the consultation to be withdrawn and revised to reflect the post-Covid world. We do not agree with these calls, but do believe that both this consultation, and this crisis, provide an opportunity to revisit and correct many of the misconceptions and errors which persist and bedevil DB schemes.
Our approach is normative; we address how schemes should be valued and managed. We shall not invoke or rely upon either economic or financial economic theory as this is unnecessary, and indeed, often unhelpful. Starting with economic or financial economic theory and applying it to the real world (a non-ergodic system - irreversible time has been a subject of great interest to Long Finance) is not a sensible starting point for understanding the mechanics of what is really going on. The guiding principle in all that follows is that the relation between sponsor and scheme, and by extension members, is equitable, that is to say: fair and just. We note that the Regulator invokes this principle extensively in the consultation paper, as “equitability”.
We start with valuation i.e. the determination of the amount of the liabilities of a company. First and foremost in valuation is a fact that is rarely considered in any discussion of liabilities - the amount of a liability or liabilities is fully determined by the terms of the contract which created the liability, and is independent of the manner in which it is financed or funded. In the case of a DB pension award, the liability incurred is a function of the contribution(s) made and the projected benefits. These give rise to an endogenous rate of return, which we call the contractual accrual rate (CAR).[ii]
This rate is time continuous, meaning that the amount returned by discounting the projected benefits and the amount arrived at by accrual of the contribution are the same. Moreover, this amount is equitable among the scheme and its members, and the sponsor and its stakeholders. It is the amount arising from the agreed terms. It is notable that no financial market volatility is introduced into the liability estimate. It is clear that the discount rates specified in legislation and in other areas such as accounting standards are misconceived; they are counterfactuals and answer the question "What would the liability be, if the terms of the contract were based on [insert discount rate of choice]?"
The discount rate is technically a measure serving to reduce a sequence of projected future benefits to a present value. For consistency in interpretation of results, time invariance of the discount rate is a desirable property. None of the rates admitted by regulation possess this property; the discount rate used in any valuation of DB pension liabilities is the largest source of variability in valuations, and this variability is spurious. The contractual accrual rate is time invariant; it will vary only as the experience of the factors driving projections, such as wage and price inflation, and mortality, vary from their expected values, or where expectations and assumptions are revised. Unlike the situation with existing measures, changes to the contractual accrual rate and valuations reflect only changes in the projected benefits payable; and most importantly, they are grounded in reality.
The valuation generated using the CAR is a best estimate of the accrued amount of the sponsor’s obligation. Other approaches, such as ‘prudent’ estimates, introduce a bias into estimates of the sponsor’s obligation and are therefore inequitable to the sponsor and its stakeholders. This best estimate of the sponsor’s obligation is the amount which would be admitted in equity as the amount of the scheme’s claim at the time of an insolvency event. This amount (or its asset equivalent) may or may not be sufficient to purchase similar benefits in the open market at the time of insolvency. However, there would usually be a shortfall. This arises because the open market from which benefits are secured is through regulated insurance companies. The Section 75 value, which is based on replacement cost, is misconceived, and is inequitable to other stakeholders and the sponsor. Simply put, it is an overstatement of the amount of the DB scheme’s claim in insolvency.
The next question which arises concerns the purpose of funding a DB scheme. The idea has taken hold that the funding of DB schemes is to pay pensions as and when they fall due. This is also misconceived. The primary purpose of the pension fund is to secure the accrued liabilities of the scheme, valued based upon the CAR. The secondary purpose of the assets of the pension fund is to defray the cost to the sponsor of providing these pensions. The trust structure serves to ring fence these assets, protecting them from the claims of other stakeholders in sponsor insolvency.
Attractive though the ‘pay pensions when due’ idea may appear, it leads to a wide range of potential abuses of the equitable relations among all stakeholders. Specifically, it leads to demands for funding in excess of the amount of the sponsor’s accrued obligation. Such overfunding could, in principle, be subject to clawback by the receiver of an insolvent company. The consequence of this is investment strategies and objectives that are inappropriate e.g. liability driven investment. It also places an overly onerous set of duties and responsibilities on trustees, which have been systematically imposed by the Pensions Regulator through an ever-increasing burden of regulation.
By contrast, with the fund as security for the sponsor liability, trustees are concerned solely with the current market value of the fund in relation to accrued liabilities of the scheme. The management of the fund is a matter for the sponsor; this is appropriate as they bear the risk of its performance. In the event of a shortfall of assets relative to the liability value (as measured by the contractual accrual rate), trustees have a duty to call upon and require the sponsor to redress this shortfall within an agreed period.
The problem at the heart of DB funding regulation is the possibility of funding being inadequate at the time of sponsor insolvency, and this has led to notions of self-sufficiency i.e. that a scheme can pay pensions whether the sponsor supports the scheme or not. This is not an equitable solution for sponsors or other stakeholders. Any regulation which attempts the impossible, i.e. self-sufficiency, of course, propagates ever more complex permutations of itself as its failings become evident. If we wish to eliminate this possibility of shortfall, and given the importance of DB benefits to scheme members, we should, it is then necessary for the pension contract to be written with an independent supplier, such as an insurance company (though even those may fail). Pension indemnity insurance covering full benefits in the event of shortfall is both feasible and affordable and works in other jurisdictions e.g. Sweden.
History tells us that inequitable arrangements do not endure. Earlier arrangements in the sphere of pensions, such as the priority or preference given to pensioners in payment, were inequitable and predictably gave rise to public outcry; the decline in the provision of occupational DB is the result of extremely costly but ineffective regulation.
The situation we face demands a bonfire of regulation, not the introduction of masses more.