Protecting Defined Benefit: managers, trustees, regulators, and the possibility of a pension scheme that works for us
John Murray, formerly of Zurich Insurance
Listen to this brief on SoundCloud
This is a substantially edited version of a blog post written by the author and published on 6 April 2018.
Introduction and executive summary
The precise methods which USS (the Universities Superannuation Scheme) uses to value its pension fund and deal with the supposed ‘risk’ of a deficit are not clear. Documents available in the public domain, such as two actuarial valuations which USS produced in 2017 (henceforth identified as ‘September’ and ‘December’, after their months of publication), do not spell out the full details. But a close study of them suggests that USS managers, actuaries and trustees are adopting an unduly cautious approach, and are committed to a highly conservative investment strategy which will destroy the fund’s value. They have been boxed into this corner by the current regulatory environment. However, their defensiveness may also reflect the trustees’ and managers’ lack of imagination or desire to protect the wealth of the Scheme’s members. This brief will review the Scheme’s plans and its relationship with the regulator and then close by suggesting how USS might extricate itself from its regulatory predicament and start acting in the financial best interests of its members. If USS were to do so effectively, a DB (Defined Benefit) pension would suddenly become affordable without increased contributions from employers or employees, and members could avoid having to take further industrial action in defence of their scheme.
Recent valuation documents give an impression that the Scheme’s much-discussed ‘deficit’ is a product of its managers’ deliberate intention to shift the fund’s investment portfolio from high-yielding equities to low-yielding government securities. When the managers speak of ‘de-risking’, they appear to mean that they plan to change the investment profile in this way, and then back-fill the resultant shortfall using the employers’ current deficit contributions of 2.1% of pensionable salaries. Without this change, there would be no deficit, and the closure of what remains of the DB scheme, which USS seem to regard as unavoidable collateral damage, could be avoided.
It is worth asking what the USS managers really mean when they speak of ‘risk’, and whether they are using the term in a helpful way. Discussions of this subject are often misleading. Commentators refer to long-term investment in equities, for example, as ‘betting’ or ‘gambling’. Such language is inflammatory, ill-advised, and should have no place in a discussion of this nature. Conversely, investing in lower-yield government securities is not as risk-free an approach as the USS managers seem to imply. And nowhere in the debate over risk and de-risking do actuaries address the risk inherent in the opportunity cost of missing out on potential equity returns. Ultimately, the primary ‘risk’ which the managers seem to be interested in avoiding is the risk of a problem with the Regulator: a risk, that is, to themselves rather than the Scheme’s members. This outlook produces further absurdities, such as the notion that the best way to make the pension fund ‘self-sufficient’ is to adopt a conservative investment strategy that will reduce the income available to it. But there is an alternative to the direction which the managers have proposed. USS is uniquely placed to seek a derogation from the much-criticised pension regulations. UUK, USS and UCU should put forward a joint case to the Pensions Minister to make an exception of the Scheme. If not, they may be able to find some other legal mechanism by which to protect the trustees from liability, in order to free them to adopt an investment strategy that preserves members’ current benefits.
The 2017 Valuation documents: some technical underpinnings
The September and the December valuations are presented as reviews of certain valuation assumptions, rather than as proposals for action. However, they clearly contain implicit proposals that involve a radical restructuring of the USS investment portfolio. Boiled down to their essentials, the proposals are aimed (perhaps not directly but certainly effectively) at reducing the investment income that the Scheme will receive in the future, thereby creating a funding deficit. One of the consequences of this action will be the need to remove what remains of the Defined Benefit scheme going forward. The reasons for this action are given as either ‘de-risking’ or creating ‘self-sufficiency’. These notes will argue that the proposed action will achieve neither of these objectives while unnecessarily depriving members of what is left of a valuable benefit.
The Scheme’s financial position is set out below:
The expression ‘Best Estimate’ in the first column should not be confused with the business term ‘Best Case’, the latter of which would be a more optimistic projected outcome than the former. In the Definitions at p. 57 of the September valuation, ‘Best Estimate’ is defined as ‘The trustees’ unbiased view of the future outcome for different variables without adjustment [or] with margins of any kind. It is consistent with the median (or 50th percentile) outcome’. This might fairly be translated as business ‘Mid Case’: a scenario that does not imply any optimism, and usually becomes the adopted business plan. Moreover, the definition of ‘Discount Rate’ in the same glossary says ‘The discount rate for technical provisions is determined by the expected investment return less a margin for prudence’, so it may be expected that this mid case outcome does in fact contain some element of caution.
The term ‘Approved Basis’ in the second column might be better described as ‘Managers’ Selected Basis’. The deficit that appears in this column is the result of the lower discount rate employed. But this column should not be confused with ‘Worst Case’. It is not an attempt to see what will happen if the margin above the gilt yield happens to drop by 48%. What it represents is the expected deficit that will result from changing the approach in line with the comments made in the body of the documents. In other words, the deficit does not come about as the result of some possible economic downturn but as the direct result of action that the trustees are, by implication, choosing to take. The actual discount rate is not made explicit but is expressed as a margin over the gilt yield. There are several sets of numbers given in the documents, but the mean rate is never made clear. The nearest clue is at p. 10 of the December valuation, where, under the general heading of ‘Investment Return (discount rate)’, we read that ‘This approach . . . includes a provision for a gradual investment de-risking [i.e. moving to gilts or equivalent] to take place over years 1 to 20’. As far as it is possible to tell, this statement seems to explain the movement in the mean discount rate between the ‘Best Estimate’ and ‘Approved Basis’. The deficit would not, therefore, be expected to emerge if things were to remain as they are, without any drastic change in the investment strategy.
The expected rates of return on different types of investment presently in use are set out below. The effect of the proposals — basically, to shift from equities to gilts — can be clearly seen.
Why is USS ‘de-risking’?
The reasons given for this shift to lower-yield investments are described as ‘de-risking’ or seeking to achieve ‘self sufficiency’. The two concepts, as used in the reference documents, are closely interrelated. Self-sufficiency is described at p. 58 of the September valuation as ‘The value of assets that are required to meet the Scheme’s accrued Defined Benefit liabilities while adopting a low risk strategy. By a low risk investment strategy we mean one for which there is a low probability of ever requiring additional employer contributions to fund benefits accrued to date.’ Note that this does not include benefits that might be accrued in the future.
This notion of shifting to gilts or similarly low-yield investments runs throughout both documents as though it was a given. Sometimes it is denied: for example, at p. 43 of the September valuation, we read that ‘The trustees apply a ‘discount rate’ to the liabilities which is based on assumed returns from current and planned future asset allocations. This does not adopt a formulaic approach to setting the discount rate linked to gilt yields’. But at pp. 49–50 of the same document, we read: ‘For the “self-sufficiency” and “economic” bases the discount rate assumes a term structure derived from the yield of fixed interest gilts appropriate to the date of each future cash flow (extrapolated for cash flows beyond the longest available gilts) as advised by the Scheme Actuary. For the ‘self-sufficiency’ basis a margin of 0.75% is added’. So practice would seem to differ from theory.
The real reason for the proposed change is never articulated, but it is not too difficult to divine. The objective of the Scheme managers is to switch the investments to what at p. 28 of the September valuation are referred to as ‘bond-like’ investments aimed at producing the gilt yield plus 0.75%, and then to fund the resulting deficit by use of the (employers’) ‘current deficit contribution of 2.1% of pensionable salaries’. This deficit, it is estimated, will be eliminated in eight years (pp. 24–25 of the September valuation). After that, they hope that the income from the ‘bond-like’ investments will be sufficient to ensure that the employers will not need to shoulder any increase in their contribution.
It is, of course, inevitable that Defined Benefit arrangements are discontinued at this stage. With such a poorly performing investment portfolio, the contributions required to keep these benefits going would be enormous.
There is a driver behind the driver here, however. That is current pensions legislation, which pushes schemes down the gilts route. The essentially flawed thinking behind this legislation is considered below. It is in this area that the real battle lies, and it would be valuable if the trustees could be brought on board. The managers’ intention is to reach some notional ‘safe haven’ whereby if things do go wrong they have assured the trustees that they will be free from any liability because they have followed the ‘safe’ route. This should secure a quiet, comfortable life for all involved — except for the Scheme members.
Investment, ‘gambling’, and risk
Before moving on to consider a way out of this morass, it would be opportune to take some arguments off the table before they are employed as a counter. It is sometimes contended, for example, that investing in equities constitutes gambling and that using this form of investment is equivalent to ‘betting’ the scheme’s funds on the vagaries of the stock market. As we will see below, there is risk inherent in all types of investment, but it is important to differentiate buying equities to hold for the long term from buying and selling in short order with a view to a quick profit: the former is investing, whereas only the latter may be considered a form of gambling (like the activities of a day trader). Moreover, it is worth bearing in mind that equities, as a class of investment, have outperformed all other forms of investment, including UK house prices, over the long term (and what is a pension fund if not a long-term investment?), or at least in the period from 1900 to end 2017. Talk of ‘gambling’ or ‘betting’ is inflammatory, ill-informed and inappropriate in this type of discussion.
Government bonds, on the other hand, are by no means risk-free investments, as is sometimes claimed. Some of the risks involved are implicitly acknowledged in the reference documents, but it would be useful to refer to some of them here. The problem of ‘Matching Term Risk’, for instance, is acknowledged by implication on p. 10 of the September valuation, which refers to ‘cash flow (extrapolated beyond the longest available gilts)’. Pension funds are very long-term undertakings, whereas government bonds have end dates, and it is often impossible to match the pension terms with government securities. This risk does not exist with equities.
Equally, there is an assumption by the managers that bond-like rates will revert to ‘normal’ in 10 years’ time. For evidence of this assumption, see the jump in the discount rate at year 11 in the table on p. 51 of the September valuation and also p. 5 of the December valuation. At p. 8 of the former, it is observed that ‘If interest rates do not in fact revert as forecast to the levels proposed [sic] by the trustees, then future contribution requirements could increase.’ A stronger word than ‘could’ would perhaps be appropriate here. This risk is a heavy one, and it could be avoided. USS got it wrong in 2014 when they thought that rates would have reverted by 2017. See p. 9 of the September valuation: ‘Between valuations, long-dated index-linked gilt yields have fallen from already historically low levels by a further 1.5%, making them more expensive than in 2014. As a result, the trustees could not de-risk the portfolio under the funding triggers agreed at the 2014 valuation’.
These are not the only risks of investing in bonds or similar. Because government securities are issued for a fixed period, they are automatically redeemed when the period ends. It is by no means certain that a similar rate of return will be available on issues then coming onto the market, as USS acknowledge when they speak of ‘Matching Term Risk’ (see above). Furthermore, if everyone is seeking to invest in bonds, this will drive the prices up and the returns down. This eventuality is implicitly acknowledged at p. 9 of the September valuation. However, no mention is made of the fact that, if the government is looking to reduce the level of its borrowing, there is the risk that there may not always be sufficient bonds issued to meet demand. This is already a factor in Germany. Finally, while the British Government has never so far defaulted on a bond, sovereign default is not unheard of and the future is difficult to predict. Going almost exclusively for government bonds introduces a risk concentration element which would not exist with a well-diversified portfolio of equities.
Risk to members’ benefits: the risk that dared not speak its name
There is a very significant risk that is not even alluded to in either of the valuation documents, namely that the members may be deprived of what remains of a most valuable benefit for no good reason. Future events could well show that by maintaining a portfolio of well-diversified equities, sufficient income would have been produced to keep current benefits in place. This is the clear implication of the ‘Best Estimate’ valuation, which is neither optimistic nor pessimistic. Trustees owe a duty to every participant and not just the Pensions Regulator.
It would be difficult to convince the man on the Clapham omnibus that one achieves self-sufficiency by reducing the income available to the fund. It is only the very strange definition of self-sufficiency contained in the document, whereby a deficit is deliberately created and then back-filled, that will be secured by the proposed approach.
A way forward
The approach proposed by USS fails in its objectives and, by way of collateral damage, ends Defined Benefit going forward. It would be wise, therefore, to abandon ‘de-risking’ and revert to a more balanced portfolio of investments, including a substantial equity element. This will not be easy to achieve, however. There are regulatory issues to consider. It is most probable that the managers will have intimidated the trustees and convinced them of the absolute necessity of reaching ‘safe haven’ if they are to be sure of avoiding any risk of personal liability. It would not be the first time that such a thing has occurred, and the natural caution of the trustees is understandable in the circumstances. Further, the Pensions Regulator may be expected to support the managers’ position. The Regulator’s role is to enforce the rules, not to consider the best interests of the scheme members.
The easiest way to arrest the planned journey to Armageddon would be to seek a derogation for the Scheme. There are grounds for this. At pp. 35 ff. of the September valuation, reference is made to the strength of the employers’ covenant. Headings on p. 35 state that ‘The covenant is uniquely robust’ and that it is ‘rated as “strong” [the highest rating] by PwC’. For all the reasons explained in the document, this looks to be a fair assessment. One could therefore argue that the scheme, in its unique circumstances, should be allowed to derogate from the real or implied rules of pensions legislation and to continue with a truly balanced portfolio approach. It follows that the trustees should at the same time be granted immunity for any adverse outturn that might follow the granting of this derogation.
There is already support for this view. In a letter published in the Financial Times on 13 March this year and referring directly to the USS, the joint signatories called for ‘A sustained large-scale academic assault on decades of regulators’ mistaken focus on point-in-time market values and inappropriate discount rates’. And it should be emphasised that the requested derogation would cost the taxpayer nothing. To have the maximum chance of success, the application should be made jointly to the Pensions Minister by UUK, the trustees and UCU. A useful analogy here is the insurance sector, where individual companies have been allowed to develop their own risk and capital models to fit their own circumstances, avoiding the prescriptive constraints of a one-size-fits-all regulatory system.
Another, though less certain, approach would be to see if the trustees could avail themselves of what is known as ‘the business judgement rule’ which is available to members of company boards. Effectively, this rule says that directors cannot be held liable to their shareholders for an adverse business outcome if they made the decision in good faith. The rule can apply even to contrarian decisions such as assuming a rise in oil prices when, in fact, they fall. Broadly speaking, the claimant, to have a chance of success, needs to be able to show that the directors acted in their own interests and against the interests of the shareholders or favoured one group of shareholders over another. It is a powerful defence, and it might be worth finding out if the trustees could take advantage of it or something similar. If the current advisers still insist on the proposed course of action, then it should be possible to find some advisers who will take a different view. After all, the shortcomings of the current pension rules and their consequences are well known.
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 hashtag #USSbriefs17; the author will try to respond as appropriate. This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.
