The use of Latin tags in titles is frowned on by this blog. But as Con is well-loved we will permit him one this once. If you want to know what “ex abundanti cautela” means, follow the link. Not all members of USS have a classical education and my O-level Latin let me down too!
Con has been off-line for a couple of weeks and we welcome him back with applause and thanks!
This piece was written in response to my daughters’ questions concerning the press coverage of the condition of the USS pension scheme as disclosed in the latest report and accounts for the year ended March 2017. Both are academics and active members – they are respectively a mathematician and a materials scientist. Its publication is a pure personal indulgence on my part.
The title reflects my view of the manner in which the USS scheme is managed – not quite reckless prudence, but overly cautious. The motivation for caution is good inasmuch as it enhances member security, but when caution reaches the point that it reduces the benefits offered in future it ceases to be so, for either party to the pensions contract.
The motivations of various commentators and observers are harder to identify – but a good scare story will usually get publicity for them that money simply cannot buy. These narratives need have no connection with reality.
Bill Galvin, the CEO, of USS has circulated a note to scheme members clarifying the position. It offers much reassurance to members.
Some press commentary described the asset investment performance as failing. It also attributed underperformance to insufficient index-linked gilt holdings. The scheme assets produced a return of 20.1% in the most recent year – that is not failure by any sensible standard. It may be around 2% below the internal benchmark set but that is to a degree an arbitrary measure.
Index-linked gilts now yield around 1.9% below retail price inflation. They have been significantly below RPI all year. The objective of any pension scheme investment must be to produce positive real returns over the long term. Should the scheme have speculated on index-linked rising in price and the real returns they offer being even lower? Clearly not – speculation is not the proper business of pension fund investment managers. Indeed, purchasing an index linked gilt on negative real returns (and indeed nominal bonds with negative yields) is the acquisition of a prepaid liability not an asset.
One of the more important questions that should be associated with extremely low discount rates is the extent to which they undermine the rationale for saving. When the investment (discount) rates are as low as we have seen in recent years in the gilt market, does it make sense to save at all. The rational criterion for saving should be that the consumption foregone today should be equal to or less than that received in retirement, or some close variant to that. There is clearly a problem with saving massive proportions of today’s income in order to assure a smaller future income.
This achieved rate of return is far above the discount rate applied in the 2014 actuarial valuation. The annual report notes that in the period 2014-2017: “The actual investment returns achieved of £18.2bn have been higher than those originally expected at £7.2bn, and this has decreased the deficit by around £11.0bn over this period.” There is an open question as to the discount rate estimate used in valuations and the actual performance of the investment assets – with the five-year average performance reported as 12% p.a. there is a wide disparity with a discount rate in the 4% – 5%. It is also notable that over this five-year period achieved portfolio returns exceeded their benchmark.
The report is clear that the overwhelming proportion of the increase in liabilities is due to changes in financial market conditions. This is a question of lowered gilt yields and lower expected returns on assets. The report states that changes in financial market conditions has added £17.2 billion to liabilities. It also states that £5.9 billion has been added due to the accrual of new benefits (and £5.2 billion of contributions received).
The use of the term liabilities, though standard actuarial practice, is somewhat confusing. The liabilities here, and throughout the report, are the discounted present value of the projected values of the pensions awarded. In other words, changes in the discount rate have added £17.3 billion to the present value of liabilities over the three-year valuation period while the ultimate pensions projected, the true liabilities have not changed materially.
In various other places, I have discussed the use of discount rates and their selection for specific purposes, as well as the suitability of the use of solvency estimates in scheme assessments – see for example: http://www.longfinance.net/publications.html. Here I shall confine myself to what we may learn from a situation which is independent of the discount rate. One such measure calculates the number of years of the ultimate pension liabilities which could be discharged by the assets currently held (with no further investment income). In 2014, this was 18 years; In 2017, it was 25 years. This 33% increase implies a very considerable strengthening of the financial position of USS.
With a scheme which has liabilities extending out 80 or 90 years, this may appear relatively short. It isn’t, it is among the longest coverage periods that I have seen, and the improvement the greatest. With such lengthy periods, it is worth thinking back to the situation 25 years ago to 1992.
The UK was in a profound recession, with residential house prices falling sharply – negative housing equity was a major political concern. Neil Kinnock was yet to snatch defeat from the jaws of victory in the General Election. Corporate insolvencies had just reached 1.6% of the active population – the highest ever recorded, before or since. Gilt yields (ten year) began the year at 9.76% and ended it at 8.49% – the year was marked by the September departure from the exchange rate mechanism and the end of the defence of sterling, which saw short rates as high as 15%. It was also the year in which Professor Sir Roy Goode began on the Pensions Law Review – whose 1,000 pages of report(s) marks the beginning of the new age of DB pensions.
Let me ask a question: how many or how much of the subsequent developments might credibly have been foreseen? At this type of timescale there are simply too many imponderables for any “weather” forecast, this is the land of climatology.
There is another metric which might be used, but is not usually reported in valuations or other assessments. This is the rate of return required on the scheme assets held that will discharge the liabilities fully. The likelihood of achieving this rate may be estimated and the ongoing security of the scheme and degree of reliance upon the sponsor derived. It is not possible to calculate this figure based upon public disclosures, but my best guestimate would be that it lies in the range -0.7% to -1% real. This is highly unlikely not to be achieved and dependence upon sponsor institutions is rather a moot point.
It is most unfortunate that the current accounting and actuarial standards so misrepresent reality, and that the press and some commentators seize the opportunity for irresponsible commentary. There is a further concern – it appears that many advisors feel that it is their duty to seek as much funding as the sponsor can bear, with no regard to the principle of equity and good faith.
