‘JEP2’s dual discount rate proposal for USS’: replies to comments on #USSbriefs89
Document created 18 January 2020
Updated 19 January 2020
The original piece is #USSbriefs89. Many thanks to those who have commented on it. I have paraphrased the comments below before responding. Further comments welcome: firstname.lastname@example.org or @aechitty1.
Q1 In section 1 you say that USS sets the discount rate at the rate of return which it calculates there is a 67% probability of its assets equalling or exceeding into the future. But surely USS calculates a different discount rate for each year from the valuation date onwards?
A2 True. When I say ‘the discount rate’ I mean the ‘average’ of those in-year discount rates, or more accurately that discount rate which, if applied to every year for the next (say) 50 years, would give the same value for the liabilities as the succession of in-year discount rates does. All this applies to the discount rate under the present valuation method and also to JEP2’s pre-retirement and post-retirement discount rates. (It also applies to SEDRp and SEDRf which strictly speaking should be seen as ‘averages’ of in-year SEDRp and SEDRf figures.) (Updated 19 January)
Q2 You say that “gilts + 1.33%” means “1.33% above the yield on long-dated index-linked gilts”. Can you clarify?
A2 Yes. First, “index-linked” is strictly incorrect. I believe USS’s benchmark when quoting ‘gilts plus’ figures is in fact not the yield on index-linked gilts but the yield on conventional (or ‘nominal’) gilts. This is because the yield on index-linked gilts (i.e. gilts where both the annual coupons and the final redemption payment are indexed to RPI) is not knowable. For we don’t know how RPI will change between now and the redemption date. (So the table on p. 13 of the 2018 consultation document is misleading in that it equates “gilts plus” with “spread over IL gilts”, i.e. margin over index-linked gilts). All the same, if the market-based prediction of RPI-inflation between now and the redemption date is correct (see A3 below) then the yield on index-linked gilts will be turn out to be the same as the yield on conventional gilts.
Second, “long-dated” is vague. USS’s formal statement of its ‘gilts plus’ benchmark in the 2014 consultation document (p.11) is “The gilt yield is calculated as the average equivalent rate of the gilt yield curve as at 31 March 2014 weighted by the profile of the scheme’s projected cash flows.” So the ‘gilts plus’ benchmark, at a given valuation date, is a function both of the yields of a range of gilts of different maturities, and of the distribution of USS’s liabilities across different periods in the future..
NB I am assuming that by default all gilt yields are measured in nominal terms. By ‘yield on a gilt’ I always mean its yield to maturity. See Fixed Income Investor’s explainer on bonds.
Q3 You say breakeven inflation is calculated as “the difference between the yields on nominal and index-linked long-dated gilts, called ‘breakeven inflation’”. But this can’t be right since, because we don’t know how RPI will change between now and the gilts’ redemption date, we can’t know the yield on index-linked gilts
A3 Yes. To be more accurate, breakeven inflation for the next, say, 30 years is calculated as that rate of inflation under which, given the current market prices of nominal and index-linked gilts, £100 would give the same yield to maturity whether invested in 30-year conventional or 30-year index-linked gilts. This breakeven inflation, minus an ‘inflation risk premium’ that is supposed to represent the value of having security against inflation provided by index-linked bonds, and that USS estimates at 0.3%, serves as a market-predicted measure of RPI inflation over the next 30 years. (Updated 19 January)
Q4 You say “Given that the yield for 30-year index-linked gilts is currently about -1.75% (see Figure 2)”, it follows that a return of gilts + 0.75% “currently means a return of only about -1.0% in RPI terms”. But surely the figure of -1.75% taken from the right hand side of Figure 2 assumes a particular RPI inflation rate, namely 3% over the next 30 years.
A4 Yes, the chart in Figure 2 shows the yield on a 30-year index-linked gilt, calculated on the assumption that average RPI inflation will be 3% over the next 30 years, with 3% then subtracted. So it shows the RPI-inflation-adjusted yield on the the bond given the assumption that average RPI inflation will be 3% over the period to its redemption date. However the market-predicted RPI inflation for the next 30 years, as indicated by the Bank of England’s breakeven inflation rate for that period (which is currently 3.13%) minus the inflation risk premium (which USS estimates at 0.3%), is 2.83%. This is very close to the 3% assumed by Figure 2.
For details see A3 above, and the Bank of Englands’s latest yield curve data, spreadsheet ‘GLC Inflation daily data current month’, tab 4 ‘spot curve’, column 30.00 years, row 16 Jan 20. (Updated 19 January)
Q5 What is the point of your first modification in section 8? Surely any rate of return can be expressed in terms of gilts+ or CPI+.
A5 The JEP report is unclear on the status of its figures of ‘gilts + 2.5%’, ‘gilts + 3.0%’ and ‘gilts + 3.5%’ for the pre-retirement discount rate. If it is suggesting that USS, UUK and UCU should agree on one of these figures for the next valuation (or even the next several valuations), then the modification is a rejection of that. If instead it is just using those figures as a measure of the aggressiveness of different possible return-seeking portfolios, then the ‘modification’ should be seen as developing the JEP2 proposal in a certain way rather than rejecting an aspect of it.
Q6 Following your second modification, if the USS was 100% pensioners, would you conclude it should be invested fully in a self-sufficiency portfolio?
A6 No. If the scheme had 100% pensioners the scheme would effectively be closed. Under the second modification, at any one time the fund would be fully invested in a return-seeking portfolio, except for the part of it associated with pension payments falling due in the following five years, which would be invested in a self-sufficiency portfolio.
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 #USSbriefs89 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.