Elements of Financial Risk Management, 2nd Edition

6. Nonnormal Distributions

This chapter develops the final piece of the univariate risk model, namely conditional nonnormality in portfolio returns. Returns are not normally distributed. The tails of return distributions are typically much fatter than the tails of the normal distribution, and return distributions are often more peaked around zero than the normal distribution. From a risk management perspective, fat tails, which are driven by relatively few but very extreme observations, are of most interest. Dynamic volatility models will capture part of the fatness in the distribution tails but for most assets some tail risk remains. This chapter suggests distributions that can adequately capture the probability of large negative and positive ...