Beyond 40 years, here be dragons: how to meet USS’s rationale for Test 1 without ‘de-risking’ the portfolio for at least 30 years
Number 30: #USSbriefs30
This is a USSbrief, published on 7 July 2018, that belongs to the OpenUPP (Open USS Pension Panel) series. It was first published on the author’s webpage on 6 June 2018, and was submitted to the UCU-UUK JEP (Joint Expert Panel) by the author on 10 June 2018.
USS’s portfolio is currently ‘broadly half in equities, one-third in bonds and the balance in infrastructure, property and other assets’. According to their investment forecasts, this growth-weighted portfolio has a 67% chance, over the next three decades, of achieving returns of 1.67% above CPI per annum. In order, however, to meet the requirements of USS’s much criticised ‘Test 1’, the current valuation calls for a shift during the next twenty years towards bonds, which have a much lower expected return than equities and property. The shift would be sufficiently great that the average return per annum that USS expects its ‘de-risked’ portfolio to have a two-thirds chance of achieving over the next thirty years is just 1% above CPI.
What explains USS’s commitment to Test 1 in spite of its great cost? A few months ago, I submitted an ‘internal’ critique of Test 1, in response to an invitation from USS Chief Executive Bill Galvin for our comments. In March 2018, I was provided a detailed written response and a meeting with USS executives to discuss my challenge and their response.
At this meeting, they informed me that the rationale for Test 1 was to ensure that it is possible to close the DB (defined benefit) scheme, within 40 years, with no further contributions needed to be made into the scheme beyond that point. They said that this requirement was based on Price Waterhouse Coopers’ assessment of the strength of the covenant, which is the employer’s ability to support the scheme. According to Price Waterhouse Coopers, the covenant appears strong over a 30 year horizon, with greater visibility of its strength during the first twenty years. Year 30 isn’t a cut-off. It isn’t the point where the sun falls below the horizon and things become dark. Rather the strength of the covenant becomes hazy after year 30. Once, however, we go as far as 40 years, we cannot see further and therefore can no longer be confident of the ability of employers to continue to support a DB pension. Beyond 40 years, here be dragons. So it must, according to USS, be possible for the DB scheme to be closed and to go into run-off by year 40.
Many of us had therefore been misreading USS’s Test 1 language of ‘self-sufficiency’ as involving an investment portfolio which secures a high probability of employers having to make no additional contributions beyond regular employer contributions of 18%. But USS has now made clear that, by a self-sufficiency portfolio, they mean one that is intended to fund a closed scheme that goes into run-off, with no further employer contributions in any form whatsoever.
I do not believe that any of USS’s literature on Test 1 and its underlying justification draws a link to the possibility of scheme closure at the point at which the covenant goes over the visible horizon. This is probably because they regarded it as too explosive to mention DB scheme closure. They should, however, have done so, because Test 1 is otherwise inexplicable and unmotivated in the context of an ongoing scheme such as USS.
When the rationale for Test 1 is made clear, we can see how USS’s approach flows from an objective which is not obviously unreasonable: namely, to ensure that it is possible to close the DB scheme at the point at which the covenant goes over the visible horizon. USS has emphasised that closure in 40 years’ time is not a target or a goal. It’s just something that will have to be possible to achieve by then, given lack of visibility of the covenant beyond that point. That is not an unreasonable basis on which to manage the scheme.
Rather than trying to persuade USS to reject this objective, we should challenge their chosen means of realising it.
We should emphasise that, in order to make sure that it is possible to close the scheme in 40 years’ time, there is no need to start de-risking the portfolio now. We could instead adopt the following revision of Test 1 which preserves the possibility of closure of the scheme in 40 years’ time:
Revision of Test 1: Set the technical provisions discount rate on the current growth portfolio so that it is prudently adjusted to a 67% chance of success in meeting or exceeding this rate of return. Discount the liabilities on the assumption that the scheme will remain invested in this portfolio for the full duration of these liabilities. Measure the gap between the liabilities, so discounted, and the liabilities discounted on a ‘self-sufficiency’ basis of gilts +0.75%. Calculate whether a 7% increase in employer contributions (from 18% to 25%) during years 20 to 40 would be sufficient to close this gap. If this is sufficient, then the revised test is satisfied. If this is insufficient, then lower the technical provisions discount rate from year 40 so that the gap between the technical provisions liabilities under the lowered discount rate and the technical provisions liabilities under the original discount rate is the same size as the residual unclosed gap. Now the test is satisfied.
When this revised test is satisfied, it will be possible to engage, at year 40, in a ‘bullet de-risking’ of the assets of the portfolio from the current growth portfolio to a self-sufficiency portfolio, in order to close the scheme. My inspiration for ‘bullet de-risking’ comes from Universities UK’s (UUK’s) own submission to the 2014 consultation, where they maintained that ‘some compelling arguments have been put forward’ for ‘“bullet de-risking” in 20 years’ time’, as opposed to ‘assuming arbitrarily that de-risking is carried out on a linear basis over a 20 year period’.
Bullet de-risking might be problematic for two reasons:
- For a portfolio as large as USS’s, it might be impossible, or else adversely affect the prices of the assets one needs to sell and purchase, to engage in such a rapid shift out of equities and property and into bonds.
- Even setting aside the above problem, year 40 might be an inopportune time to sell so many growth assets, especially given the volatility of the price of such assets.
For these reasons, one might need to modify the revised test so that it involves a more gradual shift out of growth assets and into bonds throughout, for example, the decade leading up to year 40.
An upshot of this revision of Test 1, even so modified, is that we will have at least 30 years ahead of us of returns on investments as weighted towards growth assets as the current portfolio. At each valuation, this 30 year period of investment in growth assets could be renewed, so long as the covenant remains as strong and the distance to horizon as long as they are now assessed to be.
In their submission to the consultation, First Actuarial, actuarial advisers for UCU, takes a different approach from the one I’ve just sketched. They write: ‘Being an open scheme brings significant investment advantages, which can be exploited to the benefit of the employers and members. The investment time horizon is infinitely long’.
On the approach I have sketched, by contrast, we do not assume an infinite investment horizon, since it is now clear that, and why, USS rejects such an assumption — namely, that even Price Waterhouse Coopers’ review of the covenant as uniquely robust and strong does not support the assumption of an infinite investment horizon.
On my approach, we would remain able to run an ongoing scheme, continually invested in, and discounted by rate of return on, growth assets. We would do so by means of continual renewal of a finite, but four decade long, investment horizon, at least the first three decades of which would involve continual investment in the current growth portfolio.
Update 11 July 2018
At the author’s request, an update was made on 11 July 2018 to this brief. The figure of 1.69% in the second sentence has been revised to 1.67% in order to reflect a more precise and accurate calculation that Sam Marsh has provided.
Update 8 February 2019
In ‘USS must be planning for the possibility of DB scheme closure in 20 rather than 40 years’, the author updates his understanding of the length of the horizon that is assumed by Test 1.
This is a USSbrief, published on 7 July 2018, that belongs to the OpenUPP (Open USS Pension Panel) series. It was first published on the author’s webpage on 6 June 2018, and was submitted to the UCU-UUK JEP (Joint Expert Panel) by the author on 10 June 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 #USSbriefs30 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.