Investment Theory and Risk Management, + Website

Summary

The lessons learned from the recent credit crisis taught us that traditional risk measures like VaR, which were largely dependent on the assumption that returns are normally distributed, significantly underestimated portfolio risks. Risk managers are quickly adopting alternatives to VaR, such as expected tail loss (ETL) as well as alternatives to normal distributional assumptions such as extreme value theory (EVT). Highly volatile episodes in markets over the period from 2008 to 2010 also refocused attention on volatility and correlation methodologies. On a simple level, moving average volatility measures may be intuitively appealing but, as discussed earlier, these are biased estimates and we introduced GARCH as an efficient, mean-reverting measure of volatility and multivariate GARCH as a powerful method in modeling dynamic correlations.

Not all risk is adequately captured by returns volatility, and the credit crisis taught risk managers to look beyond returns to liquidity, leverage, and counterparty (credit) risk as well and we explored adjusting risk estimates in the presence of these sources of risk. Finally, we discussed the implications of endogenous risk in which risks are outcomes of underlying behavior rather than exogenous shocks (known unknowns). In the next chapter, we study the concept of systemic risk, which takes up the concept of endogenous risk.