Yesterday saw the implementation of the funding code, tomorrow may see it rewritten

We are fast approaching the Mansion House Speech (Thursday) when we can confidently expect Government to address the issue of valuing and funding defined benefit pension schemes. In advance of this speech I will be publishing two important blogs by Pension Oldie, the first – today – address the question

Should Pension Schemes Ditch Gilts?  – (at least for valuation of liabilities…)

Part 1

These thoughts are being put forward to promote a discussion. I do believe the time is ripe for a strategic reappraisal of the legislative and regulatory framework that surrounds pension schemes, particularly defined benefit pension schemes which are recognised by many as the most efficient way to transform a given level of current contributions into a pension income in retirement. The legislative and regulatory framework was established reflecting the conditions applying towards the end of the last century and to me seems to be unthinkingly applied both in risk and scheme management in the significantly changed environment of the 21st Century.

I welcome comments and criticism. (please use the comments boxes below or send feedback to henry@agewage.com for transmission)

Introduction

These thoughts are being put forward to promote a discussion. I do believe the time is ripe for a strategic reappraisal of the legislative and regulatory framework that surrounds occupational pension schemes, particularly defined benefit pension schemes which are recognised by many as the most efficient way to transform a given level of current contributions into a pension income in retirement.

The current legislative and regulatory framework was established reflecting the conditions applying towards the end of the last century and to me seems to be unthinkingly applied both in risk and scheme management in the significantly changed environment of the 21st Century.

I welcome comments and criticism (please use the comments space on the blog-ed)

Background

The yield on UK Government Gilts is used as the base for virtually all the methods of estimating the assets and contributions required to pay pensions and other benefits, often very many years into the future, of a defined or targeted benefit pension scheme. Yet there is no intrinsic relationship between gilts and the pension scheme except as one of the range of investments available to the pension scheme. At present [almost] all the actuarial valuations of a pension scheme, including those required under legislation and by regulations, are based on a historic gilt or bond yield measure.

It has often been said that it does not matter that the value of a pension scheme’s assets has gone down because the liabilities have also gone down simply because of an increase in the gilt yield. This is blatantly incorrect – there are only two things that alter the liabilities of a pension scheme: The period over which the pension is paid, primarily represented in the mortality assumptions; and secondly, the rate at which that pension increases before and after retirement, represented in appropriate inflation assumptions.

The pension scheme should invest in such a way that it has the cash to pay the future liabilities as they fall due. The cash required can be provided by interest or dividends or rental or other income from investments, or by contributions from the Members or the Employer.

Later in a pension scheme’s life cycle when the member and employer contributions have dried up, the excess cash outflow required over that received from investments has to come from the sale of the “capital assets” of the scheme.

Over the life of the pension scheme the goal should be to maximise the cash flow from income streams and dividend income stream as that reduces the dependency on further contributions and the significance of short and long term fluctuations in market values.

The relative significance of the market values of its investments rises and the significance of the anticipated income from those investments decreases as the scheme matures.

The Issues

The issues that arise from using Gilts as the valuation base include:

1. Gilts have been the most volatile of the asset classes available to pension schemes, both on a long term and a short term basis. Over the 30 year period to 31st March 2024, the yield on 20 year gilts has varied from 0.60% to 8.33% with the market price of the bond varying inversely with the yield, while the price of the 2068 Index Linked Gilt fell by 87% during 2022.

2. Valuations are based on a spot historic measure of the gilt yield which is not indicative of future yields and even less representative of the future cash flow the pension scheme can expect to obtain from its capital assets.

3. As a pension scheme matures the relative significance of the future return discount rate diminishes relative to maintenance of the total asset value available to the scheme (to raise the cash required to pay benefits as they fall due).

4. Solvency valuations based on gilt yields are used as a proxy for the cost of buying out the liabilities with an insurance company. However insurance companies seek to invest in different assets and hence the buy-out cost should not be based on the gilt yield. However the insurer is fully aware of the “provision” being made by the pension scheme. They are therefore able to price profitably to that provision and also take steps, such as exclusivity agreements or restricting in specie asset transfers, to limit the pension scheme’s capacity to obtain truly competitive quotations.

5. When first proposed that pension schemes adopt a liability driven investment (“LDI”) policy, the policy was defined as matching the projected cash outflows with targeted investment income and proceeds. Unfortunately over the years, the “liability” measure has been distorted to be in common parlance the actuarially calculated present value of those liabilities using gilts or bond yields to discount (or enhance in times of negative real gilt yields) the projected future liabilities. It has therefore become purely a strategy to match the volatile valuation assumptions rather than an investment strategy.

6. Following volatile gilt yields as an investment policy (particularly through a hedged LDI product) without regard to cash flow requirements of the pension scheme has resulted in a substantial loss of assets, measured in Billions if not Trillions of Pounds that would otherwise have been available to pension schemes, their sponsoring employers, and ultimately to unnecessarily restricted pension payments..

7. The Pensions Act 1995 in s75 sets out a basis to establish a debt on the employer when it severs and cannot fulfil its obligation to guarantee the deferred consideration obligations of its past DB pensions promises. The legislation states that the debt shall be an actuarial estimate of the cost of buying out those promises with an insurance company at the triggering event date (apart from scheme merger this was the only alternative available in the 1990s). Significantly the debt is not adjusted to the actual cost if the estimate proves incorrect. The actuarial profession determined at that time that the appropriate estimate should be based on Gilt yields and has enshrined that relationship in actuarial standards ever since, despite the fundamental market volatility.

8. When the Pensions Act 2004 established the Pensions Protection Fund (“PPF”) and The Pensions Regulator (“TPR”) to protect the PPF, TPR decided to adopt a slightly modified s75 debt calculation as its risk measure in the Technical Provisions calculation used to estimate the amount of additional contributions it would require employers to make over the deficit recovery period. This introduces the volatile discount rate as a risk factor in scheme management. The alternative “asset based” valuation model using estimates based on the weighted prior year average achieved returns for each category of asset held by the scheme was disregarded as not appropriate for measuring the risk of scheme failure deficits. The s75 debt model based on gilt yields at a single historic date has therefore been applied to all the subsequent valuations without any consideration of changes in market conditions or an assessment of the historic value as a predictor of the future.

9. In the period 2020 to 2023, real gilt yields were negative and as a result the inflation linked cash liabilities of the pension scheme were no longer discounted as they ran into the future but successively further enhanced in each future year in the present value calculation. This resulted in mortality assumptions assuming a greater and undeserved significance.

10. In recent years the risk premium between AA corporate bonds (used as the valuation base in Company Accounts) and 20 year gilts has diminished from 2.4% vs 0.86% at 31st March 2020 to 4.9% vs 4.43% at 31st March 2024. As Corporate Bonds are more representative than Gilts of the assets used by insurance companies to generate the cash flows required to service their annuities, the assumption that the buy-out endgame can be measured using gilts yields is increasingly challenged. As the premium over the pricing model decreases, insurance companies have to look to potentially higher risk investments (such as funding sub-underwriting outside of UK solvency regulations). This raises questions over the capacity of the unfunded Financial Service Compensation Scheme to cope with the failure of one of the big players in the bulk annuity transfer market. Incidentally these are effectively the same risks as the employers through their pension schemes have already paid £10BN in levy payments to the PPF to protect.

11. While it is customary to apply a lower discount rate to pensions in payment and to predicted future pensions coming into payment than to discount the accrued benefits before they come into payment, in both cases the same discount rate is often applied throughout the forecast duration. This ignores the critical importance of the early forecast years whose net cash flow determines the capacity of the scheme to meet its obligations as they fall due in the future years.

12. U.K. Index linked Gilts (“ILGs”) pay negligible interest on a regular basis; it is purely the terminal redemption payment that reflects an inflation adjustment equivalent to the rise in the relevant Index since issue. Unless the pension scheme holds the ILG to maturity, the redemption proceeds will be subject to market determination and likely to reflect a discount of the estimated terminal value using a market determined discount rate (e.g. the yield at the date of sale on fixed interest gilts of similar maturity). ILG are therefore at best an imperfect hedge against the inflation of the pension payments as they fall due.

In the second part of these thoughts I will provide my comments on possible alternatives to using Gilts as a valuation basis and my conclusions.