Dennis Leech

Valuations

A valuation – comparing its assets with its liabilities – is a legal requirement for every pension scheme. A valuation is only truly a guide to funding adequacy for a closed scheme. But it is a poor rule for deciding if there are likely to be enough funds to pay the pensions in other cases, especially schemes open to new members. Using it as the sole funding methodology regardless of other criteria leads to bad decisions and the closure of viable pension schemes. So it is with the USS whose DB existence is under threat from a blinkered focus on valuation which says nothing about funding.

The performance of a pension scheme (as any business) is income v expenditure. That is to say that as long as contributions plus investment income exceed the cost of pensions payments and expenses, and look likely to remain so, the valuation of the investment portfolio itself is a secondary matter. When the income starts to fall below payments and expenses things change. In practice it should be possible for trustees to anticipate this and start moving into liability-driven investments in time to reduce the exposure.

The USS is a million miles from this: it does not need to touch its investment income and makes a large and growing annual surplus – last year its net surplus was over £2 billion –  yet the asset based valuation is pointing to a large deficit. This situation is so paradoxical that it ought to raise suspicions and prompt questions:

Why does the deficit figure vary so much from month to month? Isn’t that suspicious? Surely if it is a real shortfall it ought to be a stable number

If the scheme is in deficit, when, in how many years is it going to go cash-flow negative? Surely there must be an actuarial estimate of this for planning purposes.

Why have other methodologies not been used? The valuation methodology discussion forum, that was set up on the recommendation of the joint expert panel following industrial action, should have reported on the funding methodology yet has failed to do so, despite repeated requests from the UCU, instead seemingly focussing on the mathematical detail of the valuation methodology.

Open and closed DB schemes

So there are two types of schemes that the regulatory code ought to distinguish and treat differently. This was described succinctly by Baroness Sherlock in the recent House of Lords debate on the Pension Schemes Bill:

“A closed DB scheme will, of course, see contributions decline and the remaining scheme members progressively age. As more and more of the assets will be needed to pay the pensions, they will need to be lower risk and provide liquidity to ensure that members receive their benefits when they become due. A sustainable, meaningfully open scheme has an ongoing flow of new contributions, including from future members. These can be invested for the long term, providing higher returns. Their investment profile does not need to be as risk-averse as that required for a declining DB scheme. If sustainable, open DB schemes are unnecessarily pushed into the same investment and derisking strategies required for declining closed schemes, there is the risk that the regulator will push up the ongoing contributions of members and employers to such a level that, perversely, they encourage open, sustainable DB schemes to close. This cannot be right. It does not benefit employees, employers or the economy.”

(Hansard, HoL Pension Schemes Bill, 19 January 2021)

Open and closed schemes should be managed differently in terms of investment strategy. It should not be necessary for an open scheme like the USS to go down the de-risking route yet that is precisely what is being done.

And the cost of prudence…

In this paper I present a further argument that has been little discussed: that because there is difference in the volatility of asset values in the market place and investment income received from company dividends and the like, the two methodologies lead to differences in the cost of prudence. Prudence is formally assumed to correspond to a given probability that the investment funding, together with the contributions, will be adequate for the pensions. Probability calculations treat volatility as risk. Since asset prices are known to be very much more volatile than the investment income they command (a market failure), the valuation methodology leads to greater cost of a given level of prudence than the income funding methodology. Thus the use of the valuation methodology artificially increases the cost.                          

Which methodology is used is crucial to the cost of pensions because they imply different risk measures, and hence the cost of prudence. The cost of prudence is approximately proportional to the volatility of asset values; so pensions are considerably more expensive to provide if the risk metric is asset price volatility rather than investment earnings.

This is a very important and urgent issue that suggests that the DB sector deficit is exaggerated by an ill designed regulatory system.

The full paper, which includes a partial critique of financial economics, and further evidence about the USS, can be downloaded from my blog which can be accessed here

Dennis Leech is Emeritus Professor of Economics at Warwick University