How often have you looked back and thought "it was obvious, why didn't I act?" Hindsight is a wonderful thing; the challenge is recognising when you are at a seminal moment in time. As we enter 2017, are we at such a moment?
Well, let's consider a few of the dynamics currently at play:
1. During what seemed like a never ending decline in gilt yields, the commentary was often along the lines that they couldn't get lower – but they did. Now, the talk has changed with the constant theme being "lower for longer". Are we in for a surprise around the corner with yields rising more than expected? If so, what are the implications for pension schemes, in particular those with leveraged liability hedging strategies?
2. Actuarial valuations are typically driven off the pricing of gilts. However, the market's ability to sensibly project the level of forward rates has been poor of late – just look at the yield curve now compared to that implied say five years ago. How do you sensibly address this issue in agreeing an appropriate level of contributions where the aim is be fully funded on a gilts basis by a specified point in time?
3. There has been a running commentary on how much pension fund deficits have moved from one quarter to the next. Whilst we are all for transparency, is this analysis really helping or simply stoking unwarranted fears over the ability of pension schemes generally to pay members benefits over time?
4. Risk is typically measured by the volatility of the assets relative to a gilt based liability benchmark (or a close proxy thereto). This is enshrined in much of the practical workings of the pensions industry which has driven, in part, the decline in gilt yields and is just another example of how benchmarks can drive behaviour. However, is such "Value at Risk" (VaR) analysis and the investment arrangements built around it really appropriate for many pension schemes and if not what are the alternatives?
5. The use of actuarial valuation discount rates with a relatively constant margin over gilts (albeit often with separate rates pre and post retirement) is a practical convenience. However, it implies the expected return on all assets moves in line with that on gilts. Is this really appropriate where assets are invested globally across a wide range of asset classes and what are the implications of continuing to use this approach?
We could quite easily add to this list. The common theme, however, is the constant reference to gilt yields and the industry's desire for an easy "benchmark" against which to judge the financial health of a pension scheme. The problem with benchmarks is that they drive behaviours rather than simply being the outcome of a broader based decision making process.
At times like this it is useful to reflect on how we arrived at the current juncture. The evolution of defined benefit pension schemes has been documented many times. However, there is one aspect we believe warrants particular attention, namely that when pension schemes were originally set up the benefits were "defined" not guaranteed.
In the early years, members of defined benefit pension schemes faced a number of risks besides the sponsor covenant risk. In particular, pension increases were often discretionary and not guaranteed – if investment returns disappointed or the sponsor would not cover the cost, increases would not be paid. In effect, members bore a material part of the investment risk. Changes in legislation requiring schemes to increase preserved pensions and those in payment transferred this investment risk back to the sponsor (until and unless there is an insolvency event), thereby increasing the degree of security over the benefits paid.
Alongside the changing benefit structure dynamics, the approach to pension scheme funding has changed. The current mark-to-market approach, which can result in material changes in the funding level over the inter-valuation period, replaced a methodology which typically smoothed the results from one valuation to the next. The widespread use today of gilt yields in the valuation of pension scheme liabilities stems from the ability to use gilts to build a portfolio that matches (i.e. effectively guarantees) the cash flows. Therefore, a valuation on a gilts basis tells you whether you have sufficient monies today to guarantee the benefits tomorrow. As a result, funding arrangements aiming for full coverage on a gilts basis have, in effect, assumed the nature of the agreement between sponsor and members has changed from when the scheme was originally set up.
The above dynamics, particularly those around the approach to funding, provide the clue to answering the question often posed by finance directors – "what is going on, the company has made significant contributions yet the funding just gets worse?" The answer lies in considering funding from the perspective of guaranteeing the benefits vs. an ability to sensibly invest the assets with a high degree of confidence in generating the returns needed to deliver the benefits.
Schemes targeting an objective of fully guaranteeing benefits have been chasing a moving target. For some time now the costs of securing a matching portfolio have remained one step ahead of most trustees as a result of falling gilt yields. However, from the perspective of the level of returns required to pay the benefits as and when they arise we suspect many schemes may be in a similar or better position than they were, say, 10 years ago before the financial crash. If you have not done the calculation already, ask your scheme actuary to:
a) calculate the absolute return (or real return) currently required to fund the scheme's benefits over the lifetime of the scheme (allowing for any outstanding contributions from the original recovery plan 10 years ago); and
b) repeat the same calculation 10 years ago.
To provide a sense of the dynamics at play, consider an actuarial valuation undertaken ten years ago using a "gilts +" valuation basis, noting that 20 year zero coupon gilts were yielding circa 4.2% p.a. at 30 September 2006. With returns on mainstream asset classes over the same period ranging from around 4% p.a. on property, circa 6% p.a. on UK equities and close to 10% p.a. on FTSE over 5 year index-linked gilts, some schemes may have achieved returns in excess of the absolute return assumed within the liability valuation basis back in 2006. This aligned with any deficit contributions in excess of those originally envisaged 10 years ago will have acted to push down the required return over the lifetime of the scheme. There are clearly other experience items to take into account along with changes in assumptions (notably inflation and longevity) that affect the projected benefit payments, some of which will act to increase the required return (all else being equal).
Whether or not the required return has in fact reduced or increased is, however, only part of the story. The key issue is whether in the current environment the required return is a realistic objective; the outcome of these discussions will provide a clearer sense of whether the scheme's financial position has improved over the last 10 years.
Viewing the finances from the perspective of the ability to deliver the cash flows provides clarity over the progress of the scheme against the original agreement between sponsor and members. This approach also starts shedding some light on the five issues we highlighted above:
Amount of leveraged liability hedging: if you have a high degree of confidence in delivering the required return (over the lifetime of the scheme) what is the rationale for incorporating leveraged hedging strategies? In particular, what are the implications for delivering the required return in an environment that results in losses (or unwinding of profits) on leveraged structures? There may well be good reasons for retaining leveraged structures, however, we believe this should be carefully reviewed given the current market background.
Full funding on a gilts basis: this is a moving target – why make life difficult for yourself? Whilst the cash flows are not fixed, they are more stable than a gilts based valuation measure. In focusing on the cash flows, we believe it is possible to develop strategies in which trustees can have a high degree of confidence that returns will be delivered by when they are needed. The emphasis here reflects our view that there are various risk premia that can be exploited but that the timing of when the returns can be expected to come through varies. As such, our approach is to notionally allocate assets to specific cash flows, choosing assets that reflect the time period until the cash flow is due.
Running commentary on deficits: the intention behind the provision of this information is to act as a barometer of the financial health of pension schemes. However, it is simply a snapshot in time and typically provides an indication of the amount of money that would be needed now to guarantee the benefits at a certain level. Whilst these numbers are a statement of fact, we are concerned that the headlines may well give rise to concerns that are overplayed in many cases. In particular, unless sponsor solvency is a real issue, the figures provide no indication of the likelihood of a scheme being able to pay member benefits given the assets, contributions and investment strategy.
Definition of risk: with a wide range of regulations and industry practices focusing on gilts based liability values it is hardly surprising that risk metrics such as VaR are widely used in building investment portfolios. However, this is just one risk perspective and, all else being equal, encourages investment strategies that focus on managing risk relative to movement in a benchmark linked to buy-out costs rather than the risks associated with delivering the cash flows. The resulting contribution and return requirements, along with the associated risks and how they are best managed, can look very different depending on which approach you adopt. This is not to suggest that one perspective is better than another. However, we do believe it is timely to consider these issues given the current market environment and the challenges thrown up by gilt based liability measures.
Constant margin over gilts: the issues around this are similar to the comments above on the "definition of risk", namely that it encourages strategies designed to manage the risk of movements in gilt yields. However, even with such strategies in place, the use of a constant margin is unrealistic and impacts on the implied level of prudence within funding discount rates. To the extent the effect of this dynamic is understood by all parties when discussing scheme funding, then there is no real issue. However, in our opinion, funding discussions and the associated communication with members would benefit from a more realistic approach that reflects the nature of the assets held and how the portfolio is expected to evolve.
To conclude, the evolution of our industry reflects many inflection points driven off a desire to improve matters which, as a matter of principle, cannot be faulted. However, our concern is that as with many events in the past, the ever increasing use of gilt based benchmarks may have unforeseen consequences. Benchmarks by their very nature change behaviours which may result in actions that are at best sub-optimal and, at worst, detrimental to the delivery of member benefits. Whether we really are at a seminal moment only time will tell. However, with the significant change in investment behaviours over the last few years driven off one perspective of risk, it would seem timely to reflect on whether there is a better way to deliver and enhance the security of member benefits.
Carl Hitchman
Head of Fiduciary Management Advisory
Stamford Associates Limited
The information and opinions contained in this article are intended for general discussion and do not constitute a personal recommendation. Past performance is no guide to future performance and you should seek independent advice before entering into any financial transaction. For professional investors only.
Stamford Head Office
19 – 21 Old Bond Street
London, W1S 4PX
+44 (0)207 629 5225
info@stamfordassociates.com
stamfordassociates.com
Stamford Associates Limited is authorised and regulated by the Financial Conduct Authority.
© Stamford Associates Limited 2017. All rights reserved.
