Grab a piece of the small pie

The Alternative Investment Management Association Limited (AIMA) Journal

December 2007

By Fatima Vawda

The case for hedge funds is a strong one. Many investors, including increasing numbers of institutional investors, have opted for hedge funds as a diversification element in their traditional portfolios. The upside of broader mandates, strategy diversification, risk reduction. Downside protection and yield enhancement offered by hedge funds, has been too attractive to ignore. Research has shown that the appropriate addition of a hedge fund component increases the Sharpe ratio of a fund; that is, it increases the ratio of excess returns per unit of risk. In the Sharpe ratio the unit of risk measured is the standard deviation.

So where is the downside? One important theme, raised by international hedge fund researcher Harry Kat and others is that hedge funds may reduce standard deviation but at the expense of skewness (and in particular, negative skewness). A distribution which is negatively skewed is one which has a higher-than-normal probability of large negative returns. Clearly this is a kind of risk that matters to us. Large losses pose very difficult challenges for the investor and the portfolio manager. Two things compound this problem. One is that the most popular measure of risk, the standard deviation (as used in the Sharpe ratio), is insensitive to skewness. Two funds can have the same standard deviation but one of the funds can have a much greater probability of large losses. Another issue that Kat also showed is that many hedge funds display co-skewness with typical investment portfolios. This is a serious problem. It means that your hedge fund, chosen as a diversifier of risk, may crash at exactly the same time as the rest of your portfolio does. Here is an example. Suppose you add an equity market-neutral hedge fund to your portfolio. Such funds will, under most circumstances, provide a very useful addition, adding return but displaying a low correlation with the rest of your long-only equity portfolio. That is very attractive. But suppose that your hedge fund works entirely by going long illiquid, small-cap stocks and shorting liquid large caps. Historical evidence shows that such funds take a beating at times of market crashes because of the subsequent flight-to-quality as investors ditch small illiquids for the safety of the big battalions. Something like this happened after the 1998 crash. Although hedge funds were not much in evidence then, there were many small-company funds which had been star performers prior to the crash but which struggled for many years thereafter, until relief arrived in the form of falling interest rates in the early 2000s. So, what to do? As is so often the case, the answer is simple. Be aware. Be intelligent. Be aware. With any investment, you should know under what circumstances that investment will underperform. In particular, you should know what factors your investment is exposed to. In the example above, a good risk analyst would easily identify the exposure to liquidity and size factors in the portfolio. Good fund of hedge funds managers will ensure that undesired factor exposures are minimised at the overall fund level. Be intelligent. Construct your hedge fund portfolio so that it has the properties that you require. Researchers at the French institute Edhec have shown clearly that hedge funds differ in the degree of co-skewness they exhibit with various traditional portfolios. Certain styles of hedge funds make very good diversifiers of long equity portfolios, under all market conditions. Some do not. A good fund of funds manager will ensure a high proportion of such “rain-proof umbrellas” in your hedge fund portfolio and make sure that the overall portfolio has just the right characteristics to optimise your total portfolio. Where the optimisation is quantitative, the use of a risk measure sensitive to skewness, such as VaR or expected tail loss, will ensure that skewness is minimised. If the optimisation is carried out at the level of the total portfolio (i.e. traditional and alternative together) the co-skewness effects will also be dealt with.