Introduction to Risk Models*
This chapter provides an introduction to risk models. Modern risk management is position-based. This is more forward-looking than return-based information. Position-based risk measures are more informative because they can be used to manage the portfolio, which involves changing the positions.
Part Four of this book focuses primarily on market risk models. Ideally, risk should be measured at the top level of the portfolio or institution. This has led to a push toward risk measures that are comparable across different types of risk. One such summary measure is value at risk (VAR). VAR is a statistical measure of total portfolio risk, taken as the worst loss at a specified confidence level over the horizon. More generally, risk managers should evaluate the entire distribution of profits and losses. In addition, the analysis should be complemented by stress-testing, which identifies potential losses under extreme market conditions that may not show up in the recent history.
Section 12.1 gives a brief overview of financial market risks. Section 12.2 describes the broad components of a VAR system. Section 12.3 then shows how to compute VAR for a simple portfolio exposed to one risk factor only. It also discusses caveats, or pitfalls to be aware of when interpreting VAR numbers. Section 12.4 turns to the choice of VAR parameters, that is, the confidence level and horizon. Next, Section 12.5 shows how to implement stress tests. Finally, Section ...