Were the deficits real?

The reported deficits during the years before 2023 were the result of the actuaries overstating the liabilities. That happened due to their practice of thinking about interest rates (and also investment returns and discount rates) as if they were driven by gilt rates determined in the bond market. This, combined with the use of the mathematical formula for present value, adds up to a flawed methodology.

So when gilt rates were driven down to near zero by government monetary policy after the financial crash of 2008  (so-called quantitive easing by the Bank of England), pension liabilities ballooned.

The problem for the USS is illustrated in this diagram showing the assets and liabilities reported by the actuary at each valuation, both statutory valuations and interim valuations. Figure 1.

This was not a coincidence but due to their use of a flawed methodology. In reality the affordability of the pension promises was unchanged –  USS was still in good financial shape. That is because the returns from the kind of growth assets in the scheme’s investment portfolio were not linked to gilt yields and continued to be good enough despite QE.

How important this has been is illustrated below in Figure 2. As gilt yields went down the liabilities grew enormously. Now that gilts have returned to a kind of ‘normal’ range since 2023, liabilities have likewise returned to normal.

The obvious lesson from this experience is that liabilities should be valued with a discount rate more in line with the expected returns from the actual portfolio, that is, one based on what the actuaries call ‘best estimate’. There is no legal reason why the trustees of USS should not choose their discount rates in this way: it is provided for in the regulations.

The devil is in the mathematical detail of how the liabilities are calculated. Some members have forgotten that when the crisis began around 2011 and UCU appointed pension officers to look into it, we said the valuation was not real because of this – that they were using the wrong discount rate – essentially the wrong economic model. That argument has not gone away just because the deficits have evaporated.

The USS has to move away from gilts-plus discount rates. Until that happens we cannot take the valuation deficit as a true indication of the basic financial health of the scheme. A valuation based on a gilts  plus discount rate is more a reflection of the state of government monetary policy than of the health of the pension fund.

Deficits and Conditional Indexation

One awkward legacy of this is that many members have got themselves into a way of thinking that deficits happen because of occasional poor investment performance. They talk about the  scheme as if it were an investment fund that sometimes has “good years” and sometimes “bad years”. They think that the valuation is a kind of objective reality check rather than the result of using a flawed model.  Hence the case for Conditional Indexation which is being seriously considered as the answer to instability.

But a DB pension scheme is not an investment fund. The health of the scheme does not follow investment returns in a simplistic manner. So it is not a question of maximising investment returns but rather of investing so as to be able pay the pensions promises that members have built up, when they retire.

If we accept CI without sorting this out there is a risk of long term damage the next time gilt rates go down for reasons of government policy – such as following another market crash like 2008. Limiting benefits under CI because of the resulting actuarially calculated ‘deficit’ – when in reality the scheme’s investment portfolio is performing well enough would be a disaster. In other words CI is not an answer to instability.

Incidentally this kind of event – such as following a crash resulting from a market bubble – as widely expected to be occurring now with the mania to invest in AI – is not covered by the simulation studies the USS has commissioned from consultancies like Ortec to help it argue for CI. Simulating thousands of economic scenarios drawn from a stable universe provides no guidance for such fundamental uncertainty incapable of mathematical modelling.

See the second CI report. https://www.uss.co.uk/news-and-views/briefings-and-analysis