With the intention of keeping the Pension Regulator focused, Keating and Clacher have recruited Dr Anna Tilba of Durham University to their ranks. A powerful triumvirate of pension intellectuals whose latest blog is an antidote to the toxicity of pseudo de-risking so prevalent in received wisdom
This blog has been written in response to several requests for us to explain, succinctly, how we would measure and manage occupational DB pension schemes. We believe this is how schemes should be measured and managed and reflects the ongoing reality of a pension scheme.
The appropriate rate at which an employer should recognise DB pension liabilities is determined by the terms of the contract(s) creating those liabilities.
A DB pension contract specifies a future pension in terms of the scheme member’s future salary and is payable for their lifetime in retirement. The pension may have other attributes such as spouse and minor dependents benefits. These are all estimated using standard actuarial techniques; they involve assumptions with respect to wage and price inflation and longevity. They result in projected future values for the liabilities. This is well-travelled, well-understood territory. The liability comes into existence with the payment of the initial contribution. This contribution is the amount of liability which should be initially recognised by the employer.
The contribution and the projected values of liabilities uniquely and completely determine the rate at which the contribution should accrue in order to fully amortise the projected liabilities. This rate is intrinsic to the contract and fundamentally invariant over the life of the pension contract. We call this rate the Contractual Accrual Rate (CAR). Crucially, this rate will only change if the experience of wage, price inflation, or longevity varies from that originally assumed at the time of award or if the employer wishes to revise the assumptions with respect to these parameters.
The passage of time and the accumulation of experience ensures convergence to the pensions ultimately paid. For example, with a final salary scheme, one year prior to retirement we know, as a matter of fact, all but the final year’s salary change.
The Desirable Properties of a Discount Rate
The CAR captures the necessary conditions to discount correctly. First, it is time consistent. The present value of a liability using this rate is the same amount when it is calculated by accrual from prior values as it is when calculated by discounting of the projected future value.
Second, this rate may vary from employer to employer; its value will have been determined by the generosity (or otherwise) of the awards made. As the pension is an obligation of the employer, this rate is the cost of the award to that employer. Moreover, it is also the rate of return required from scheme assets in cases where there is no recourse to the sponsor employer (e.g., Collective Defined Contribution schemes).
Third, as occupational pensions are a significant proportion of the wealth of most members and often critical to their wellbeing in retirement, it is important that they should be secure. As sponsor insolvency is the only source of risk to members, given the employer is essentially guaranteeing the return on assets, collateral security is provided for them in the form of bankruptcy remote, funded trusts.
To this end, the correct level of funding needed to protect the accrued benefits of scheme members at a point in time is the value derived using the CAR. Note that this is not a replacement value. The value of this collateral may or may not be sufficient to purchase a similar pension in the market at the time of sponsor insolvency. The amount here represents full and faithful discharge of the obligations of the sponsor up to the point of insolvency. We should ask for no more. Funding the scheme to higher levels than this is inequitable to other creditors and stakeholders, whether insolvency occurs or not.
The Section 75 claim value can be seen to be entirely inappropriate in this context.
The pension fund should be valued by both employer and scheme at market prices as it is the immediate value of this as collateral which is of primary interest. Crucially, the pension fund returns defray some or all of the employer cost of providing the pension benefits to members.
When considering the asset allocation of the pension fund, it is the level and correlation of the fund returns with the earnings of the employer which is important. This suggests that it should be the employer, and not the trustees of the scheme who determine asset allocation and investment policy.
It is worth noting that the volatility of the contractual accrual rate will be a small fraction of that arising under current discount rate selection methods, which are essentially counterfactuals e.g., this is what the scheme would look like if it were targeting only gilts as its investment strategy. Unlike other approaches, which are influenced by the vagaries of markets (and government intervention in markets) changes to the CAR will are rooted in economic reality, such as unexpected wage rises and inflation experience.
With perfect foresight the volatility of the pension scheme in the employer account, if the CAR were to be the discount rate, would simply be the volatility of the asset portfolio. The volatility of scheme results under current arrangements has been the prime motivation for the closure and de-risking of all-too-many schemes by employers. Indeed, most of what passes in practice for de-risking is the elimination of the spurious volatility introduced by the incorrect discount rate being used.
As much of the woes of pensions started with financial accounting and FRS17 and subsequently IAS19 (with FRS 17 also influencing the original regulations of TPR) it is interesting that outside of pensions, accounting has moved on. The fact that the correct discount rate is fixed by the terms of the contract at the time of inception and initial recognition has now appeared elsewhere. It is the required method for the recognition of profits from long-term insurance contracts over their lifetime under IFRS 17. Perhaps, there are reasons to be hopeful that economic fundamentals may yet re-emerge in pensions.