Quantitative Risk Management: A Practical Guide to Financial Risk, + Website

8.2 Comments Regarding Quantitative Risk Measures

I want to highlight some points that I think are not sufficiently elaborated in standard treatments of quantitative risk measures, VaR in particular, or that are misunderstood by many users. I am not criticizing or rejecting VaR. There are many critiques of VaR, many which are not justified. Some commentators have said that it is useless and even a fraud. In my experience, views usually fall at one of two extremes:

1. Pro-VaR: It is the silver bullet that answers all risk-measurement questions.

2. Anti-VaR: It is at best useless, more often outright misleading or worse.

As often happens, the truth is closer to a synthesis of the two views: VaR can provide useful information but has definite limitations. When properly understood and appropriately applied VaR provides information and insight, but when VaR is misapplied or misunderstood it can certainly be misleading.

Standard Trading Conditions versus Extreme Events

There are two related but somewhat divergent uses of summary risk measures such as VaR and volatility:

1. To standardize, aggregate, and analyze risk across disparate assets (or securities, trades, portfolios) under standard or usual trading conditions.

2. To measure tail risk or extreme events.

Risk measurement theory and texts usually focus on extreme events and VaR, but risk measurement in practice focuses as much on standard trading conditions and summary measures other than VaR. Paying heed to risk under standard ...