Markets & views

The Investment Journal (volume three)

Liability driven investment: An exercise in risk management

This is without doubt the most turbulent, volatile and extreme set of market events ever faced by pension funds. Pension fund trustees, asset managers and investment consultants face very serious difficulties in deciding what steps to take next. Every course of action will subsequently, in the fullness of time and with the benefit of perfect hindsight, be judged to have been prudent, prescient, inspired or, alternatively, reckless, naïve and ill thought out. There is unlikely to be a halfway house.

To add to the pressure, industry-wide decisions thus far have not inspired confidence. Most pension funds have witnessed asset deterioration across many traditional asset classes on an unprecedented scale, as Chart 1 illustrates.

The long established theory of “diversification benefit” has all but been suspended and it is unclear when normal service will be resumed. Each month is worse than the last with all signs pointing to even further attrition in 2009.

The speed with which this Category 5 hurricane has developed has taken pension funds almost universally by surprise. It began innocently enough with a relatively unremarkable problem in June 2007 in a Bear Stearns credit fund which was forced to take a large loss as it became clear that the value of its credit assets was nowhere near the value attributed to them. However, it soon became apparent that the problem was not confined to a single, particularly arcane, type of asset. Indeed, in the last two years, there has been a global re-pricing of risk across virtually every asset class and pension funds continue to face a hostile economic environment and outlook.

Understanding risk

In 2009, therefore, it is particularly challenging to establish a reliable and robust risk management framework. All pension funds sign up to the principle that it is crucial to understand and manage “risk”, but beyond that there is plenty of differing opinion. For example, how should a pension fund define risk? In order to achieve badly needed return on assets, one pension fund’s trustee board may regard as acceptable an investment that another set of Trustees would consider imprudent. There are so many influencing factors – funding level, strength of sponsor covenant, market conditions, level of governance and Trustee expertise, extent of existing risk management framework, etc – that it is simply impossible to prescribe a finite set of risk limits or acceptable asset classes. And yet, it has never been more vital to have a clear picture of risk and a carefully planned approach to risk management. The alternative – maintaining the status quo – is, for many pension funds, simply not viable.

Our view is that risk – however defined – cannot be discerned in these markets in reliance on any single set of risk parameters. There are several examples of pension funds (and major financial institutions) that have suffered serious losses because they relied solely on a single risk measure that was, on its own, inadequate.

In other words, a pension fund requires a set of multi focal “risk lenses”. This is true for both the assets and the liabilities and there are at least three such risk lenses which should be utilised.

Risk lenses

As a first step in a basic risk analysis we advocate using an asset/liability “microscope” to determine the fund’s sensitivity to micro-factors – which should be considered separately, then together. The results usually contain a wealth of important information. This type of 100x magnification of the fund’s sensitivities highlights those exposures that are simply too large and have the potential to cause significant losses.

This first lens shows the impact on the liabilities of small changes in certain “unrewarded” factors on the aggregate liabilities e.g. interest rates, inflation or longevity assumptions. However, the behavioural characteristics of each member class within the plan are very different. It is important to turn the lens on separate component member classes – the pensioners, actives and deferreds. A tiny rise (or fall) in inflation for instance, ripples through the future liability cash flows and changes their present value by a defined amount. A minute adjustment to the interest rate used to discount the cash flows payable to tomorrow’s pensioners can materially alter the present value of the liabilities.

However, this market factor sensitivity analysis does not provide any information about the likelihood or expectation of loss due to market movements. An important complementary second risk measure is Value-at-Risk or “VaR”.

VaR

VaR provides a single minimum amount (the value) which the fund might expect to lose (at risk) one year, say, from now. So, if the fund’s liabilities are £1 billion, its assets are £800m and its 1-year 95% VaR is £300m, then, simply, VaR ascribes a 5% risk to the deficit increasing from £200 million today to £500 million or greater in a year’s time.

Crucially, the VaR calculation relies on estimates of future volatility and correlation together with an assumed probability distribution. Often these are derived from historical information. Sometimes that historical data inaccurately forecasts the future. VaR also does not provide reliable information on how significant the “or greater" loss may ultimately turn out to be and a common error in relation to the use of VaR is to equate the VaR amount with a maximum (rather than a minimum) likely loss.

There are, consequently, many detractors of VaR, the most vocal of whom is perhaps Nassim Taleb. In his oft quoted book “The Black Swan: The Impact of the Highly Improbable" 1 Taleb eloquently argues that users of VaR are typically lulled into a false and sometimes fatal sense of security about the future, since, he says, events perceived by a VaR model as impossible or extremely unlikely (Black Swans) are actually the main sources of risk.

Taleb has a point in his criticism of VaR. And yet, there is another perspective. In an article for the New York Times, (January 2, 2009) business columnist Joe Nocera interviewed several risk managers and quotes one:

"One risk-model critic, Richard Bookstaber, a hedge-fund risk manager and author of “A Demon of Our Own Design," ranted about VaR for a half-hour over dinner one night. Then he finally said, "If you put a gun to my head and asked me what my firm’s risk was, I would use VaR." VaR may have been a flawed number, but it was the best number anyone had come up with."

And that’s the point. For all its flaws, VaR provides useful additional information with regard to the risk profile. In addition, VaR itself has several sub-lenses and risk “filters" which help to build a clearer picture of where the fund may find itself in the future. For example, whilst VaR only tells us the minimum loss we can expect, “Conditional VaR" goes one step beyond basic VaR and provides an average expected loss should the “tail event” actually occur.

But even with these additional add-in VaR lenses, it is a serious mistake to rely solely on this risk metric, because of the inherent inability of risk models based ultimately on historical experience to predict the sheer magnitude and effect of future market events.

Stress testing

So the Trustees need a third perspective from which to view the fund’s risks – a kaleidoscope of large scale moves in markets. This comprehensive collection of simulated bleak and dismal market events is designed to measure just how well equipped (or not) the pension fund is to withstand sudden and severe jolts or sustained market dislocations. Financial regulators have always advocated the importance of stress testing but one of the lessons from the current global crisis is that the tests were insufficiently severe. Trustees can and should simulate their own bespoke kaleidoscope of rising inflation, falling interest rates, plunging equities, falling property and weakening credit – in other words, a brutally tough obstacle course to test the real-time robustness of the fund and its current investment strategy.

Conclusion

In conclusion, in an environment in which most asset classes have steeply declined in value, it is notable that the best performing assets have been gilts and long dated interest rate and inflation swaps (please see Chart 1) – the very instruments used to hedge volatile pension liabilities. With a simple set of guidelines, it is possible for any defined benefit pension fund to analyse its risk and to start the process of re-aligning and improving its risk profile.

Contact us

Please share any comments or views you have regarding The Investors Journal by contacting:

Chris Wagstaff
Editor, The Investors Journal theinvestorsjournal@avivainvestors.com

Unless otherwise stated the views expressed are those of Aviva Investors. They should not be relied upon as indicating any guarantee of return from an investment in funds managed by us.

The information shown should not be viewed as investment, regulatory, tax or legal advice nor as a recommendation of any nature.

09/0200/280210