Why USS’s Test 1 is not too risky and how Test 2 should be changed
This is a USSbrief, published on 9 August 2018, that belongs to the OpenUPP (Open USS Pension Panel) series, and was submitted to the UCU-UUK JEP (Joint Expert Panel) by the author on 3 August 2018. The section on Test 1 was first published on the author’s webpage on 1 August 2018.
1. Why Test 1 is not too risky
As a follow-up to my first JEP submission (also published as an OpenUPP submission: #USSbriefs30), I would like to submit the following: ‘Why Test 1 is not too risky’.
This engages with David Miles’s submission to the JEP, which has been publicly posted (see #USSbriefs36: ‘The state of USS: some observations’).
Most of the critics of USS’s Test 1 maintain that it is overly prudent in setting too demanding a level of aversion to risk. In his submission to the Joint Expert Panel, however, David Miles of Imperial College raises the opposite worry:
that passing Test 1 should give members of the USS and the sponsoring universities little comfort. To put it another way: the position of the scheme is more precarious than the apparently ultra-prudent USS calculations suggest.
I explain, in contrast, why Test 1 remains sufficiently prudent.
The underlying objective of Test 1 is that it be possible to close the defined benefit (DB) scheme 40 years from now and go into run-off. At that point, pensions benefits would be paid out of a low-risk ‘self-sufficiency’ portfolio weighted towards bonds, which has a 95–97.5% chance of being able to cover all DB promises that have accrued.
David Miles raises the following objection [*]:
self-sufficiency’ means only that pensions can be paid, without extra support, with ‘a high degree of confidence’. How high is that? This is not entirely clear from the documents USS make easily available, but I believe the answer is with a probability of 95%. That would be considered scandalously risky for a bank and completely unacceptable to bank regulators. I suspect that USS scheme members would not consider the scheme to be on a solid foundation if they got their promised pensions 19 times out of 20.
This fear that ‘self-sufficiency’ leaves members with a 5% chance of failing to receive their promised pensions is unfounded. It rests on the implausible assumption that there will be no UK higher education sector around more than 40 years from now, whose employers will be able to make at least modest deficit recovery contributions in the 5% likely event that the self-sufficiency portfolio underperforms. Given the low variance in the value of assets in a low-risk self-sufficiency portfolio, any deficit from which it might be necessary to recover is unlikely to be very large.
Although we can safely set the above worry to one side, there remains the following further challenge to which Test 1 gives rise, to which I shall devote the remainder of this submission.
In order to be able to close the DB scheme by year 40, Test 1 requires that it be possible to purchase a self-sufficiency portfolio by then, via a supplementation of the scheme’s assets through an increase of 7% — i.e., from the current 18% to the maximum affordable 25% of salaries — in employer contributions from years 20 to 40. The assets must reach a market value by year 20 that is sufficiently high that it is possible to get from there to self-sufficiency via such an increase.
Insofar as this possibility of getting from year 20 to year 40 is concerned, it does not matter whether the mixture of assets in the scheme at year 20 remains along current lines — namely ‘broadly half in equities, one-third in bonds and the balance in infrastructure, property and other assets’ — or whether the assets have been ‘de-risked’ in the manner USS now proposes, via a 20-year shift towards bonds. Rather, what matters is the market value, not the composition, of the assets at year 20.
The challenge is that it must now be possible, in year 1, to ensure that the value of the assets reaches the required level by year 20.
On the assumptions USS uses for Test 1, there is, however, only a 67% chance that the assets will reach the required level by year 20. It follows, of course, that there is a 33% chance that the assets will fall short of this level, in which case it will not be possible to purchase a self-sufficiency portfolio by year 40.
Therefore, assurance must be provided that it will be possible to recover from the 33% of scenarios in which the value of the assets falls short of the required level by year 20. This recovery would need to come out of extra deficit recovery contributions beyond the current 18% regular employer contributions. All of these extra contributions would need to be paid before year 20. This is because Test 1 assumes that, from years 20 to 40, the maximum affordable 25% (18% + 7%) is going towards raising the required funds to purchase a self-sufficiency portfolio by year 40.
USS might appeal to the need to provide this assurance in order to try to justify their plan to ‘de-risk’ the portfolio through a shift to bonds between now and year 20. They might maintain that the current growth-weighted portfolio carries too much downside risk for it to be possible to guarantee that the scheme could recover from losses, via extra deficit recovery contributions between now and year 20, and reach the required asset level by year 20.
In any event, some such explanation is needed, for why we cannot remain continually invested in the current growth portfolio between now and year 20.
Such an explanation is needed, in light of the fact that we now have a long 20 year period in which to make adjustments out of extra deficit recovery contributions to get to the required level of assets by year 20. As one gets closer and closer to year 20, there will be less scope for making the needed adjustments to recover from unexpected falls in the value of the assets. If, however, the covenant remains as least as strong and visibly long from triennial valuation to triennial valuation, the relevant target date will always remain 17–20 years in the future, as the 40 year horizon and its 20 year midpoint will always move forward in time by 3 years at each valuation.
If the strength and visible length of the covenant deteriorate at future triennial valuations, then one might need to revise Test 1 in a more conservative direction at that point. But there is not now any call for such revision.
[*] Miles also raises other challenges for USS in his submission, which I do not address here. In particular, he questions the credibility of USS’s claim that there is a 67% chance of achieving full funding. He does not, however, raise the particular challenge to Test 1 to which I have devoted most of this submission.
2. Test 2
I would also like to urge the JEP to encourage USS to adopt the following, which USS proposes on p. 26 of ‘Methodology and Inputs for the 2017 Valuation February 2017’:
USS proposes Test 2 is replaced by ongoing monitoring of the required contribution rate for the current benefit using a model that calibrates to the underlying internal rate of return assumptions used by the trustee rather than a fixed margin over gilts …. USS feels that estimating the required contribution using a model calibrated to the latest view of the expected return on assets will be a more reliable indicator of the employers’ short term risk exposure.
In my blog post of 21 December 2016 — ‘Alarming deterioration in USS funding is based on an incoherent valuation methodology’ — I spell out the case for adopting such a monitoring approach which coheres with the approach to the valuation that USS takes at its triennial valuations.
This is a USSbrief, published on 9 August 2018, that belongs to the OpenUPP (Open USS Pension Panel) series, and was submitted to the UCU-UUK JEP (Joint Expert Panel) by the author on 3 August 2018. The section on Test 1 was first published on the author’s webpage on 1 August 2018. This paper represents the views of the author only. The author believes all information to be reliable and accurate; if any errors are found please contact us so that we can correct them. We welcome discussion of the points raised and suggest that discussants use Twitter with the hashtags #USSbriefs39 and #OpenUPP2018; the author will try to respond as appropriate. This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.