There has been much criticism of the way investment benchmarks are used by institutional investors, such as pension funds. Many commentators have argued that the ‘tech boom’ was exacerbated by institutional investors piling into technology stocks simply because they were in the index. Many investment managers have complained that risk controls around an index benchmark force them to take positions in stocks they do not like. The Myners Review in the UK questioned whether pension fund trustees were allowing their investment managers sufficient freedom to take risk and add value.
One reason why this issue has drawn so much attention recently is because falling equity markets mean that all equity portfolios with tight risk constraints around a benchmark have seen sharp falls in value over the past year. Even good fund managers, who have delivered strong out-performance relative to their benchmark, have still delivered large negative returns in an absolute sense.
Many commentators and some fund managers have been prompted to ask why institutional investors are so wedded to index benchmarks. Why not give the investment managers freedom to invest in those stocks they expect to deliver the best returns? Then perhaps, a good fund manager can deliver positive investment returns whatever the market conditions. It seems like a nice idea – just ask the fund manager to pick the 40 stocks they think will perform best. It is instinctively appealing, and simple to understand.
Where did the index benchmark come from in the first place? The idea of the index benchmark originated with the Capital Asset Pricing Model (CAPM). This has been accepted academic theory for many years, and it concludes that investors can achieve the best trade off between risk and return by holding all the stocks in the index according to their market capitalisation weight. CAPM does depend on assumptions, in particular relating to efficient markets, that may not be fully borne out in practice.
So what is the problem with an index benchmark? There seem to be two main complaints, one of which is valid and one of which is not. The valid complaint is that an index benchmark does not allow a fund manager to underweight smaller capitalisation stocks. A typical index will have large weights in a few stocks and small weights in many stocks. For those stocks with small weights, the fund manager cannot take a significant underweight position if they are not allowed to short the stock. At the same time, if the fund manager has limits on the size of any one overweight or underweight position, they will be forced to hold a certain minimum holding in the largest stocks – this could be described as a quasi-passive holding as the fund manager does not have discretion to sell this holding.
The other complaint, which is not valid, is that measuring a portfolio against an index benchmark constrains the manager because it stops them from taking enough risk. This is only true if the investor has an intolerance of under-performance, which is inconsistent with their out-performance objective. It may be true that many institutional fund managers set themselves artificially tight performance constraints – but this is only because they fear being sacked if they under-perform significantly. The fund manager may rationalise that they are unlikely to be sacked for close to index performance. If you are a successful manager with large assets under management, it is more important to protect those assets by avoiding underperformance than it is to attempt to win new assets by outperforming.
This problem could be solved by investors setting their managers realistic risk targets, and then showing a willingness to accept that even good managers will have periods, where they underperform in the short term.
So is the answer to take away the constraint of the benchmark index? Perhaps this sounds like a nice idea. However the concept of an unconstrained benchmark raises its own questions. How is the benchmark defined? How is performance relative to the benchmark measured? How do you monitor whether the manager is undertaking their job in a manner that is consistent with achieving the objective they have been set?
The simplest solutions revolve around equally weighted benchmarks. However, I believe this approach leads to a potential danger. If everyone started getting rid of the index as a benchmark, would create small cap bubble. Investment managers would start selling out of large cap stocks such as Vodafone and Nokia. Money might pile into smaller companies in whichever sectors might be in vogue, pushing up their market capitalisation to levels far higher than might be justified by the economic reality of their businesses. Ironically, by trying to remove the market distortion which contributed to the tech bubble, we could end up helping to create a small cap bubble.
A truly unconstrained long only global equity portfolio with an absolute return benchmark might work as an alternative allocation for a small part of the portfolio. I do not believe such a benchmark should be considered as part of a mainstream/ core asset allocation to equities for the reasons given above. However, I also think there are other interesting alternative approaches for both core/ mainstream portfolios and satellite/ alternative portfolios, once you start to give the investment manager more freedom.
What are these alternatives? For mainstream equity portfolios, institutional investors could retain index benchmarks, but allow the investment managers greater freedom. The manager would be set a higher out-performance target, with higher risk and a greater tolerance of underperformance. For more conservative institutions, rather than set active managers lower risk parameters, the overall risk could be reduced by balancing this type of mandate against a core index or enhanced index portfolio.
For institutions that are concerned about taking too much equity market risk in the benchmark, another answer is to put some of your “risky” assets in asset classes that diversify from equities, eg property, private equity and hedge funds. Property is back in vogue, although how much of this is simply down to good recent performance is hard to tell. Private equity is interesting but investors need to be happy with the idea that their money is tied up for the long term. Other investment opportunities might be developed in the future to capture long-term economic growth without direct equity market risk, for example investment by pension funds in long-term infrastructure projects. The difficulty with many of the alternative asset classes and investments is the lack of a traded market and consequent difficulties in valuation and performance measurement. However where a pension fund is able to commit part of its funds to be tied up on a long-term basis this should not be an insurmountable problem.
Another area that has attracted much attention recently is hedge funds. We see a reluctance from institutional investors to invest in hedge funds in some countries, but less so in others. However, although institutions may not be rushing to invest in hedge funds, the publicity around this area is raising awareness of more sophisticated investment instruments and techniques.
Even in countries like the UK, where pension fund trustees are showing considerable reluctance to invest in hedge funds, their investment strategies are becoming more sophisticated for different reasons. As bonds are becoming a more important part of pension fund portfolios, in an environment of fast maturing liabilities and increasingly short-term funding measures, trustees are looking at ways to add more value to their bond portfolios from active management. This is forcing them to look at a wider range of bond asset classes than before, and increasingly sophisticated tools, eg swaps.
Having developed appetite for more sophisticated approaches in bond investment, these could be applied to the equity portfolio too. If you allow an equity manager to short sell within an equity portfolio that is benchmarked against an index it removes the constraint imposed by the benchmark. Then the only problem becomes setting the objective and parameters of the mandate in a way that ensures the manager takes on an appropriate amount of risk for achieving their objective. Once an investor has bought into the idea of allowing a fund manager to short sell it opens up a whole range of possibilities. Allowing fund managers to use the full range of derivatives as well as conventional financial instruments allows development of the concept of ‘portable alpha’. A good fund manager can add value or ‘alpha’ in the area where they have skill, and this can then be transplanted on to the underlying investment return of any benchmark (for example, a bond-related liability benchmark that you choose). You could even have several diversified added value (alpha) processes adding on value to the same underlying benchmark return.
Of course, investment managers need to become more sophisticated too. Not all good ‘long-only’ investment managers can easily translate their skill to a ‘long-short’ environment. Fund management firms making this transition may need to learn new skills and also get a better understanding of managing the risks of more sophisticated financial instruments under a wide range of economic conditions.
Alvar Chambers is principal and actuary, investment, with Aon Consulting in London
No comments yet