Seemingly a stupid question and one many professionals would be able to answer in many ways. These professionals would use advanced mathematics, quantitative modelling techniques and they would rely on statistics and data mining of historical numbers to back up their answers. While this may explain why we don’t find many risk managers in client facing roles, what we would really like to know is, how good are all of these models in predicting how much money we can lose?
The simple answer is, it depends. Clearly, any risk model can only give you predictions made from data obtained in the past and will, therefore, not be able to tell you what is going to happen in the future. However, a good risk model can give you the parameters of any potential occurrence within a specified level of confidence. The gold standard of most risk systems used in the financial services industry are so called Value-at-Risk (VAR) models. They model a random number of possible outcomes and then calculate the expected loss within varying sets of outcomes.
A lot of people also look at liquidity as a measure of risk and look at the time it would take to get out of their investments. Of course, this risk would only apply if you actually had to sell when things go bad and, as such, only looking at liquidity may be an inefficient way to invest, as we are potentially losing out on better opportunities with better risk and reward profiles.
There is also the volatility of a given instrument, as measured by the extent of the historical up and down moves in price or value. But again, if you intend to hold the investment over a long time horizon, this measure becomes almost entirely meaningless.
So how should we really look at risk? For one, it is vital to know how much one can lose when everything goes wrong and, more importantly, can one live with that risk? This is often overlooked in all the excitement of putting on a trade: it can also go wrong, horribly wrong. The financial crisis is a prime example of when things that were never deemed possible, all of a sudden became painful realities.
The other way to look at risk is to realise that there is no good or bad risk – there is only unrewarded risk. In other words, you can have a lot of risk in an investment, but if you can also make huge amounts of money, it becomes a great trade. Whereas, a low risk investment may not offer any upside, in which case this becomes a terrible trade.
The fact is, in today’s world of zero interest rates, you have to take risk to generate returns. At Blu, we categorise risks according to their relative behaviour and the underlying drivers of return (premiums). From there, we diversify across as many different risks as possible from the entire investment spectrum, whilst making sure that we put on the best risk/reward trade per category, regardless of asset class or product type.
Ultimately, no risk model, whether based on VAR, liquidity, volatility or any other measure, will help you predict when things go wrong. It is therefore essential to be honest with oneself in order to prepare for such an eventuality.