03 June 2009

Published in: Capital - models, Risk governance, Market risks

Make proper allowances for risk interactions

Individual risks like volatility aren't the problem, argues Milliman's Neil Cantle. The way they combine is.

The CRO Forum paper Calibration principles for the Solvency II standard formula (see News, "CRO Forum stresses importance of implied volatility") recognizes that in the current crisis a very significant factor is volatility risk. They obviously want to make sure that people are thinking about this and taking it a lot more seriously.

I support this objective but I think that the problem is that doing things by risk components will never actually work. It's not the individual factors per se that are the problem. It's the way they combine to create unexpected outcomes. Risk is very complex and you have to look at how different risk characteristics interact.

Right now, there's an economic crisis which everyone's reacting to. The products and risk transfer mechanisms that were used prior to the crisis have caused these sensitivities to the equity markets, which is why volatility has become important.

Once people react to the current markets, however, the product designers will create new features and risk transfer mechanisms that work better in this sort of economic environment. People will then find that another factor becomes one of the key risk drivers rather than volatility.

If you look back, solvency regulation was not such a bad fit with the way that companies operated in the past. But the problem was that the environment very quickly moved on and the solvency regime was no longer able to cope with the dynamic and adaptive nature of the modern financial services market. The danger is that we now do the modern equivalent of that first attempt at solvency regulation and latch on to what we think currently drives the risk exposure: you build that in to the legislative framework and everybody manages to that and feels comfortable. Then all of a sudden something else pops up and actually that was the thing you should have been looking at.

A better way of accounting for risk is the own risk and solvency assessment (ORSA): you have to start looking at how risks actually work in your organization. Within the ORSA, if your organization has a sensitivity to volatility, for example, then you absolutely have to include it and understand how it works and what it could do to your particular organization. The ORSA is probably the most powerful part of Solvency II, and the part that people are least prepared for.

I think in the end people are going to have to move towards tools that are capable of making proper allowances for the risk interactions. Rather than simply trying to look at the correlated outcomes of risk processes, these tools recognize that the observed outcomes of risk processes emerge over time through complex interactions and that there are patterns to be found which help you to know what is going on early in the evolution. It's about making people remember that it is not just about looking for individual risk factors but the way risk factors interact to create emerging risks which causes the unexpected outcomes that surprise everybody.

Neil Cantle is Principal and Consulting Actuary, life practice, at Milliman in London

Back to top