Selectors need a new approach to risk

This essay appeared in Citywire Selector on December 18, 2009.

‘Risk’ is everywhere but impossible to pinpoint. For investors, the term is often used in a generalised manner to mean ‘something to be exposed to when times are perceived as good to make money, and to avoid when times are bad and exposure is likely to lead to losses.’

This definition is clearly an over-simplification – there are in fact many types of risks, all intricately intertwined. In our industry, there is not a single term that is used so broadly, that is so important and is so perplexing. Liquidity, deflation, inflation, credit, market, concentration, operational, style-drift, volatility are all types of risk to be managed, controlled, profiled, bought, ‘VaRed’, exposed to, sold and/or avoided.

Selectors have a broad variety of risks with which they must contend. Identifying the right ones to focus on in a given situation is the real challenge. In some cases, risks are clear – no selector should be surprised that a fund with a 35% position in the bonds of a single (very illiquid) emerging-market country will experience significant downside when markets hit a bump.

In other cases, risks are more difficult to identify and manage – particularly when multiple funds are combined. Across several or more funds, the potential risks of concentration across stock, industry, sector or country must be considered across the entire exposure.

Specific style biases and investment themes often show up repeatedly across diverse ranges of funds – particularly towards the peak of bull markets. Further, the rapidly increasing levels of sophistication in investment techniques and instruments in funds lead to a whole new range of risks to be assessed.

What from one perspective appears of limited risk and being well under control can quickly evolve to be a major danger. Think about how many popular funds during the extended bull-run of this decade had only positive flows month after month – managing liquidity was only a matter of finding somewhere to put the money coming in the door. I, however, am the kind of guy that takes note of the closest exit in a cinema.

Without an adequate methodology for managing and understanding risk, there is really no basis for making an investment at all. But over-reliance on VaR models, credit ratings, style analysis, risk/reward ratios and other such tools, in the assumption that a ‘universal’ method for containing and monitoring risk has been discovered, actually has a way of increasing overall risk levels.

By focusing on and containing a given set of risks, others are often overlooked and dismissed as insignificant. In one instance several years ago, during a due diligence meeting with the risk management team of a prominent fund company, I realised the danger inherent in the over-reliance on such models. The team was very excited about the progress they had made in ‘controlling’ the risk in their fund range by use of VaR modelling. We went into depth on the analysis of one particular fixed-income strategy. The composition of the fund was complicated and contained a significant number of less-liquid and non-market traded instruments for which mark-to-market pricing was non-existent or at best highly unreliable. When I asked about how these were accounted for in their VaR models, I was told that they were excluded.

The risk that really hurts is the one you don’t see coming. This is the reason for the ubiquitous ‘Look right’ signs painted on roadways in London – ‘right hand side’ visitors have modelled risk to be coming from the other direction and step off the curb to cross the road… screech, boom!