Moving smartly up the agenda
It used to be a simple enough thing. Allocate to equities, and ignore your risk exposure. Strong markets, and pension surpluses, meant you didn’t have to worry. With the heady days of the 1990s gone, pension funds are facing tougher times, and risk management has become an important issue. Not only are pension funds underfunded, with the combined European pension deficit standing at €116bn according to consultants Lane Clark & Peacock, but asset allocation approaches have had to change.
In a low-return environment where every basis point counts, pension funds are venturing further afield from mainstream investments, and turning to liability-driven strategies and alternatives. It means that risk budgeting has also become a more complicated task.
“Risk budgeting has moved up the agenda in terms of trustee awareness. The market to market of liabilities and the realisation that liabilities themselves are volatile is partly responsible for this, since this has been the primary reason for the dramatic deterioration of funding levels,” explains Gareth Derbyshire, managing director at Merrill Lynch.
The biggest risks, he points out, are often interest rate and inflation. “Eliminating those without impacting the rest of the strategy usually means using interest rate and inflation swaps. You can use long duration bonds, but for every pound of bond you buy, you have to sell something else.”
Consultants point out that the days of equity-only investing are long gone. “One of the major risks in the past has been the asset class mismatch, which has primarily manifested itself in the form of equity exposures of 70-80%. Now, people are recognising that liabilities are more bond-like, and running that level of equity exposure is a material source of risk relative to liabilities,” says Ralph Frank, worldwide partner at Mercer Investment Consulting. He points out that interest rate risk and inflation risk are embedded in the asset class mismatch risk, “although the extent of the contribution of these two factors to the overall mismatch risk is not always clear”.
In the UK, the use of inflation linked swaps tripled last year (2005) because of demand from pension funds, says consultants Watson Wyatt. In 2004, the market for end users, excluding intra-bank trades and executions with insurers, was an estimated £3bn (€4.8bn). But last year, the market could have exceeded £9bn, although exact levels of activity are hard to measure since most deals are private. Still, Watsons has advised on 20 swap and related executions over the last three years, with total nominal exposure standing at some £14bn.
Pension funds should ask themselves a number of key questions before making decisions about their risk budget, suggests Watson Wyatt. They must first look at how much risk their fund is taking, whether it is equity risk, duration risk, or currency risk, for example. But, says the firm, there is one overarching risk, “the risk that trustees have insufficient assets to pay members’ benefits”. Funds can look at future funding levels and contribution rates before measuring risk and, suggests Watson Wyatt, trustees should benchmark their investments against the growth in a measure of their liabilities.
After identifying how much risk is being taken, trustees need to decide whether it is an appropriate amount. A lower risk budget improves the security of members’ benefits, but higher risk budgets offer the potential for higher returns. The firm has criticised the traditional approach to determining investment strategy, where funds first determine their strategic asset allocation using asset liability modelling, and then create the policy benchmark. “With this approach, pension fund risk has historically been dominated by market risk. Manager risk has generally made a small contribution to the overall level of risk relative to the liabilities.”
Watson Wyatt points out that there are several weaknesses to the approach. The disciplines used to develop market exposures have been relatively weak, while strategic asset allocation benchmarks have typically included a number of duration, inflation, and currency risk exposures that were not expected to be rewarded. The concentration on benchmarks has limited the flexibility of managers, and the separation of asset allocation and manager skill has meant that the overall strategy has not been well joined up, according to the firm.
Others say pension funds have several further issues to consider. “A lot of risk management techniques are restricted to the relative world, but at the end what matters is absolute return,” argues Epco van der Lende, portfolio manager for global asset allocation at Morgan Stanley Investment Management in the Netherlands. He believes that the approach of completing an asset allocation study, and then setting a strategic benchmark using a combination of indices, and finally handing out mandates based on the tracking error of those indices, is problematic.
“Managers are not concerned with overall pension fund profiles; they are just trying to beat the index. The only link to the fund is tracking errors, and when the tracking error budget is set, it is usually not done very quantitatively,” he suggests. Van der Lende also argues that once strategic weights are set, it creates a “very rudimentary way of trying to stick within the strategic risk profile. The idea is that whatever happens in the market, you stick to the strategic risk profile and you do not deviate too much from that. The problem is these bounds are very rudimentary and sub-optimal; you never hit all the bounds at once for all asset classes.”
A far better approach, he suggests, is tracking error which also takes into account the correlation between asset classes. “What is still missing is the link between the assumptions on the strategic side, and the tracking error. If you look at it closely, the tracking error ranges that are allowed should depend on the strategy.”
For its part, Watson Wyatt highlights three challenges to overcome for successful risk budgeting. The first is the fact that there is limited historical information available. The use of an asset model is an intrinsic part of the risk budgeting process, but it is reliant on various assumptions. Problems include the fact that there is survivorship bias, and the existence of data mining, where past patterns are allowed to influence current assumptions. “As a tool, risk budgeting has been effective for getting to the nub of the debate about what pension funds should be doing regarding risk. It’s been incredibly powerful in that regard. But these are models. Nobody has perfect foresight here for how assets and liabilities are going to behave, so they have their limitations too,” says Kevin Carter, head of Watson Wyatt in Europe.
The firm started doing an annual appraisal of the risk budget of its schemes. “Then people said, well hang on a minute, since capital markets can move dramatically, why are we not monitoring this quarterly? I think it’s fair to say that only a minority of pension funds have gone down the route of close monitoring at present.”
Other challenges identified by the firm include the fact that there is limited capacity, where the returns from some strategies will vary by the amount of money allocated to them. “At the micro level it is relatively straightforward to see that an investment manager is likely to be less successful in outperforming with £10bn (e14.4bn) than he is with £10m. The same effect can occur at an asset class level when the entire industry changes its allocation to it,” suggests the firm.
According to Carter, pension funds “model their liabilities with a suite of bonds, some of which may not even be investible, thereby creating a proxy of the liabilities. You can then work out, from the yield curve, what the hurdle rates of the liabilities will be year to year. Now it is possible to see how this relates to the assets in the plan which enables one to quantify the risk, usually measured in VaR terms. Trustees can then use this to decide whether the level of risk is acceptable.”
Creativity, he suggests, has emerged from swaps, and liability driven investing, where the use of derivatives is key to better matching liabilities and assets. “Risk awareness has become the dominant conversation in trustee boards, as opposed to return seeking”.
But critics say the approach may have gone too far. In the past, liabilities were ignored. Now, pension funds are in danger of ignoring market risk. “Trustees are focusing a lot of liability risk, and it’s not their fault. There are changes in accountancy, in regulation, in tax laws. So I wouldn’t want to lay this all at the trustees or the plan sponsors or the consultants, but clearly, there is some misplaced advice. When liabilities have been taken into account, then, with whatever is left over, the portfolio managers need to take more active risk,” suggests Peter Jeffries, of Independent Risk Monitoring.
Others agree. “In the broad European discussions about liability matching, there is a lot of oversimplification and neglect of issues. Asset liability matching is somehow defining away risk. There is also another important point. Not every country in Europe has legislation where asset liability matching is the answer.
In Austria, for example, the liabilities in the pension plan are not discounted at market rates; they are discounted at fixed discount rates. “So, for us, asset liability matching is not the appropriate approach anyway. We have to go for an absolute return approach,” explains Günter Schiendl, director of investments at APK Pensionskasse, a fund that has been using derivatives for several years.
APK was founded in 1989 as Austria’s first pension fund. It is an independent scheme that runs 13 different plans, most of which are defined contribution. The bulk of its clients are industrial corporations. Schiendl himself has a background in the derivatives industry.
Placing your faith in fixed income is not always the answer, suggests Jeffries. “Liability management has to be an element of fixed income, but if the fund is reasonably well funded, the portfolio managers should be encouraged to take a view on the market and participate. The obvious example is the Boots pension fund. They felt they couldn’t take any more risk so they went from equities into fixed income, and for six months they looked fantastic. But two and a half years later the market has risen 40% and fixed income has gone nowhere,” says Jeffries.
He also points out that over the past decade, many pension funds have allocated their risk budget between passive and active strategies, the argument being that pursuing low costs passive strategies enables funds to spend their budget on higher-risk, higher-cost strategies. “However, the 2001/2002 meltdown and its impact on index funds, followed by relative under performance of hedge funds and fund of hedge funds, is calling these strategies into question.”
Still, pension funds that are investing in hedge funds have other risk implications as well, says Marc Gross, regional director for Riskdata in London. “It has been very difficult for institutions to get their heads around a risk budgeting perspective for hedge funds, because they don’t fit into the normal framework for traditional asset management risk,” he says. He argues that the pay-off structures of alternative investments can be non-linear, and not normally distributed, with fat tails and extreme events. “A pension fund might have an equal amount of allocation to 10 hedge funds, but three of those could account for 75% of the risk. What is important to understand is who are the risk concentrators, and who are the diversifiers.”
Governance a major consideration
There is the issue of governance to consider. In a recent study on developments in pension governance, the Organisation for Economic Co-operation and Development pointed out that good pension fund governance has been linked to improved investment returns. PGGM, the Dutch pension fund, is one fund that has included a risk management framework into its governance policies, for example.
In 2004, the fund introduced a risk management framework for the whole of its organisations. “This is a standardised way of identifying, systematically highlighting, and monitoring risks at strategic, tactical, and operational levels,” says the fund in its governance code.
PGGM identifies its risk into three stages: risks at a gross level, without taking account of existing management measures in place; determining the effectiveness of existing control measures; and devising action to improve situations where risks are deemed to be unacceptable. The executive directors and heads of various units are responsible for managing the risks within PGGM. This includes identifying, analysing and evaluating risks, with a risk report drawn up each quarter.