It is suggested from time to time, often more in hope than expectation, that investment benchmarks could and should be dispensed with entirely because they have distorted the investment decision-making process. They are said to have constrained investment managers’ freedom and forced them to invest in things they otherwise would not, or prevented them from investing in those opportunities they regarded most favourably. This criticism is usually applied to particular types of benchmarks (such as the established equity market indices) but is framed in terms of a more general criticism of benchmarks, which is unjustified.
Every investor with a liability to meet has an implicit benchmark, and that is the liability itself. Equally, every mandate given to an investment manager must have a benchmark of some sort if there is to be any means of knowing whether the portfolio has been managed well or badly. An absolute return target or a return in excess of cash or index-linked government bonds is a benchmark like any other. So is a requirement to beat a peer group. From this perspective, asking whether it is good to have a benchmark is much the same as asking whether it is good that grass is (mostly) green; we have no choice in the matter.
Even so, the simple fact that a benchmark of some form must exist does not mean the benchmark that is actually set for an investment manager’s portfolio best serves the interests of the investor. This article discusses two different approaches to benchmarking, primarily focusing on the investment of pension fund assets.
The first, the approach generally followed by institutional investors today, constitutes what is colloquially the ‘strategic benchmark’ approach. It involves the investor taking responsibility for the level of strategic risk in the benchmark by setting defined weightings to equities, bonds, and other asset classes and specifying benchmarks for each asset class (eg, the FTSE All Share for UK equity).
The alternative approach shifts much or all of the responsibility for decisions on risk (strategic or otherwise) to the investment manager and is usually referred to as an ‘absolute return’ approach, (although the name is not necessarily always suitable). This provides scope for much greater investment manager freedom and an escape from the alleged constraints of more traditional benchmarks.
Which of these two best serves the investor’s interests?
The strategic benchmark approach: It is worth emphasising the two basic skills required to set a ‘traditional’ strategic benchmark effectively. The first is an understanding of the liabilities and the second is an understanding of the risks inherent in different asset classes relative to the liabilities. The fact that expertise on the liability side does not often lie with the investment manager is the principal justification for this approach to setting benchmarks. Responsibility for the acceptance of strategic risk within a benchmark has therefore lain with the investor and their consultant advisers owing to their liability-side expertise.
In deciding on what is tolerable in risk terms, the investor should take into account both the level of strategic risk inherent in the benchmark and the level of risk taken by the investment managers (active risk) selected to manage it. The investor can reduce strategic risk in order to increase active risk, or vice versa, but the overall ‘risk budget’ should stay the same.
In theory, there is a clear separation under this approach between active and strategic risk. Strategic risk is the undiversifiable risk in an asset that would lead to a risk premium – an excess expected return – compared to a lower or risk-free asset assuming that both were priced efficiently. Equities exhibit significantly greater strategic risk than government bonds and so it is reasonable to expect an equity risk premium, even in a wholly efficient market. Active risk, on the other hand, is risk taken with the aim of benefiting from a mispricing by the market. It has no inherent risk premium attached to it; for active risk to add value, an investor must have superior information to its peers or the ability to better apply the same information. To be worth accepting, active risk must improve the risk/reward balance and this would happen only in an inefficient market.
Under the traditional approach, active managers are not expected to meet their targets by taking on greater levels of strategic risk on their clients’ behalf: they are expected to exploit mispricings. In practice, the two risks are impossible to separate completely, but it is reasonable to say that the level of strategic risk taken by investors with traditional benchmarks is largely defined by their benchmarks and strategic risk is often a significant component of total risk. It is important to keep this in mind in discussing absolute return benchmarks.
Absolute return approach: As mentioned at the start, it is not possible to just dispense with benchmarks entirely. The investor with liabilities should always choose an investment policy that considers those liabilities. However, an alternative to the traditional approach is to transfer all decisions on risk relative to the liabilities to an active investment manager. To do this, the investor would define a benchmark that represents as close a match as possible to the liabilities. This means that ‘benchmark’ risk taken by the investment manager equates to ‘liability’ risk. For a defined benefit pension plan, a benchmark of bonds would be the result and the excess return target would be set by reference to that. If the excess return target were substantial (say +3% per annum or more), significant discretion would need to be given to the investment manager. The manager could, for example, be permitted to invest in equities, even though they were unrepresented in the benchmark. This is not in fact an absolute return mandate at all, but an actively managed high-risk mandate with a bond benchmark as there is no requirement to produce a positive return when bond returns are significantly negative.
This approach blurs the distinction between strategic risk and active risk but this does not mean the investor should be agnostic as to the balance between the two. For example, an investment manager that outperforms a bond benchmark through investing in equities is not necessarily adding value because outperformance was expected anyway. If the manager was charging a passive manager’s fee, the investor might be unconcerned, but this is improbable to say the least and there is surely no case for paying high active fees simply to take strategic risk. The outperformance target could be adjusted to reflect the level of strategic risk actually taken but this would require heroic (and controversial) assumptions for quantities such as the equity risk premium and would weaken the investor’s control over the level of risk taken relative to liabilities. It would also require a high degree of transparency.
In short, absolute return mandates provide the most appealing prospect in value for money terms when it is clearly active management skill that is generating the bulk of the excess return. Investors wishing to adopt such mandates as an alternative to a more traditional asset class benchmark would therefore be swapping strategic risk largely for active risk. Some may find this attractive but, as active risk has no inherent risk premium, it does depend on identifying above-average skill. Depending on it entirely requires a great deal of faith and, unless active risk can be expected to offer proportionately greater rewards than strategic risk, it is likely that a combination of active and strategic risk will be preferable.
As a result, ‘absolute return’ is unlikely to replace established approaches to benchmarking on a widespread basis. However, allocations to such vehicles within a broader strategic benchmark are more probable. The hedge fund industry, whilst lacking transparency, is arguably already working to performance mandates similar to those just described, and is therefore the more likely entry point for institutional investors to absolute return mandates.
Steve Woodcock is a European principal with Mercer Investment Consulting in London