Sykes Wilford, Professor and Hipp chair of finance, The Citadel, Charleston, South Carolina recently gave a presentation to the CFA Socity in Cayman, talking about the history of risk management. Asset managers, individual investors and anyone interested in the markets and the current crisis were encouraged to attend. Georgina Loxton from the CFA Society in Cayman describes the presentation.
The history of risk management is entwined with past market crises – it has been developed in response to losses. Each time we have had a financial blow-up, such as LTCM, the Asian financial crisis or Black Monday in 1987, risk managers have looked at what happened and asked themselves ‘what did we miss?’
In response to the loss and the inadequacy of their models they added another variable or another layer of complexity to the model. However, what many risk managers have failed to appreciate is that tools are just tools and they are not the solution. The art is in knowing how to use the tools, in asking the right question and in taking the right perspective. A portfolio manager has a very different risk perspective from a fund of funds manager, or from an investor, or CIO or board member.
Wilford spoke about the problems that a fund of funds manager faces. Risk management should involve charting the historical monthly portfolio returns, say over the past five years. This gives the manager a return series on which he can plot a best-fit normal distribution. The manager needs to analyse, not only the deviations in returns, but also the correlation of the returns and the shape of the return series, specifically the series’ skewness or kurtosis. Skewness measures whether a distribution has relatively few high values (positive skewness) or whether a distribution has relatively few low values (negative skewness). The kurtosis of a distribution measures whether the distribution is peaked or flat relative to a normal distribution. A higher kurtosis means that the distribution has longer and flatter tails meaning that more of the variance is due to infrequent extreme deviations, as opposed to frequent modestly-sized deviations.
A fund of funds manager then needs to ask a question such as ‘what is the most extreme event that could affect my investment?’ Perhaps it is correlations going to one, or correlations moving negative. He can look at the distributions of performance for all his investments in order to identify how his portfolios are likely to behave in such a crisis. As with all risk management, the most important thing is to apply the correct technique to the situation.
Risk management now involves many complex models. Unfortunately most of the people using the models and interpreting the results do not understand the assumptions behind the model, and this renders a lot of the work useless. Risk managers need to be reminded of what the ‘art’ actually is – it is using unruly data to get clear answers for individual portfolios. The art is in asking the RIGHT question – the data do not lie, but the answers can.