Measuring Risk

Volatility: the standard deviation of returns

Standard deviation treats upward and downward deviations the same way and implicitly assumes that returns are distributed fairly symmetrically. In reality, markets often experience extreme events more frequently than a normal distribution suggests, so volatility should be supplemented with other measures.

Beta: sensitivity to the market

Beta measures only systematic risk. For example, if the covariance between a stock's returns and the market's returns is 0.018 and the market variance is 0.015, then β = 0.018 ÷ 0.015 = 1.2. On average, if the market falls by 10%, the systematic component of the stock's return would be about −12%, plus any company-specific moves.

• β above 1: the asset is more sensitive than the market (for example, cyclical sectors).
• β between 0 and 1: the asset is less sensitive than the market (for example, utilities).
• Estimated beta depends on the period, data frequency and index chosen; for infrequently traded stocks, the estimate can be unreliable.

Maximum drawdown

An illustrative portfolio starts at MDL 1,000,000, rises to a peak of MDL 1,250,000, then falls to MDL 950,000 and recovers to MDL 1,100,000. Maximum drawdown = (950,000 − 1,250,000) ÷ 1,250,000 = −300,000 ÷ 1,250,000 = −24%. Getting back to the peak from the low requires a gain of 1,250,000 ÷ 950,000 − 1 ≈ 31.6%.

Risk-adjusted return: the Sharpe ratio

Although portfolio B has the higher return, A delivers slightly more excess return over the risk-free rate per unit of volatility. The Sharpe ratio thus lets you compare strategies with different levels of risk.