Furthermore, joint distributions estimated over periods that do not include panics will underestimate correlations between asset returns during panics. Under these circumstances, fear and hence disengagement on the part of investors holding net long positions often lead to simultaneous declines in the values of private obligations, as investors no longer materially differentiate among degrees of risk and liquidity, and to increases in the values of riskless government securities. Consequently, the benefits of portfolio diversification will tend to be overestimated when the rare panic periods are not taken into account.

... At a minimum, risk managers need to stress test the assumptions underlying their models and consider portfolio dynamics under a variety of alternative scenarios. The outcome of this process may well be the recommendation to set aside somewhat higher contingency resources--reserves or capital--to cover the losses that will inevitably emerge from time to time when investors suffer a loss of confidence. These reserves will appear almost all the time to be a suboptimal use of capital, but so do fire insurance premiums--until there is a fire.

Probability distributions estimated largely, or exclusively, over cycles that do not include periods of panic will underestimate the likelihood of extreme price movements because they fail to capture a secondary peak at the extreme negative tail that reflects the probability of occurrence of a panic. Furthermore, joint distributions estimated over periods that do not include panics will underestimate correlations between asset returns during panics. Under these circumstances, fear and disengagement on the part of investors holding net long positions often lead to simultaneous declines in the values of private obligations, as investors no longer realistically differentiate among degrees of risk and liquidity, and to increases in the values of riskless government securities. Consequently, the benefits of portfolio diversification will tend to be overestimated when the rare panic periods are not taken into account.

Another theme that emerges is liquidity risk. A fundamental assumption of many risk management procedures is the ability to get out of a position or to hedge it. Events in Asia demonstrated once again that assumptions about liquidity in normal markets rarely hold in more volatile ones. This argues both for a reassessment of the assumptions themselves and for more careful and fundamental thinking about liquidity risk in risk management procedures.