Measuring and modelling risk

This paper will examine some commonly adopted approaches to the measurement of risk in finance and the various shortcomings implicit in the underpinnings of these approaches: early views on the nature of risk and uncertainty (Hume, Bernoulli, Knight, Keynes and Ramsey); the adoption of a mean variance decision choice criteria as a central foundation in financial economics and its accompanying limitations; the various approaches in financial econometrics to modelling volatility (ARCH, GARCH, stochastic volatility, realised volatility and attempts to capture 'tail risk'); the measurement of risk implicit in applications of option pricing models and implied volatility (in particular the VIX index); the Basel Agreements and convention of modelling risk in a value at risk (VaR) framework; and the attractions of conditional value at risk (CVaR) as an alternative metric. I shall conclude with a consideration of the shortcomings of these various approaches when faced with a system wide shock as recently experienced in the global financial crisis.