Sharpe Ratio

The Sharpe ratio is a measure of risk-adjusted return, developed by economist William Sharpe. It is calculated by taking an investment's return, subtracting the risk-free rate to isolate the return earned for taking risk, and dividing that excess return by the investment's standard deviation, its volatility. The result is the amount of excess return earned per unit of risk, and a higher number is better: it means the investment delivered more reward for each unit of bounciness it put its owner through.

Its value is comparison. Raw return alone cannot say whether a fund that returned 12% was better managed than one that returned 9%, because the first may have taken on far more risk to get there. By dividing excess return by volatility, the Sharpe ratio puts both on a common scale and asks which one paid its investors more for the risk they actually bore. Two funds with identical returns can have very different Sharpe ratios, and the steadier one wins.

The Sharpe ratio's greatest strength and its central flaw are the same choice: it defines risk as standard deviation. That makes it clean and comparable, but standard deviation treats an unexpectedly good year exactly like an unexpectedly bad one, so the Sharpe ratio penalizes upside volatility as if it were a hazard. An investment that occasionally delivers a huge positive surprise is scored as riskier, and therefore worse per unit of risk, than a duller one, even though few investors mind large gains. Measures such as the Sortino ratio were designed to answer this by dividing by downside deviation only, counting bad surprises but not good ones.

Three further cautions bear on how the number is read. First, it is entirely backward-looking: it summarizes a past period and depends heavily on which period is chosen, so a strategy can look excellent over one stretch and poor over another. Second, it assumes returns are roughly bell-shaped, which understates the danger of strategies whose losses are rare but severe (selling insurance-like risk is the classic case), so such strategies can post a high Sharpe ratio right up until the tail event arrives. Third, the ratio has no absolute scale that means "good" in every context; it is most useful as a relative measure, comparing similar investments over the same period against the same risk-free rate, rather than as a verdict on a single fund in isolation.

The choice of risk-free rate matters more than it looks. Subtracting a higher risk-free rate lowers every excess return and compresses the ratios, so comparisons are only fair when the same rate is used for every investment being ranked. And because the denominator is volatility, leverage cuts both ways: borrowing to amplify a position raises both the excess return and the standard deviation, so it does not reliably improve the Sharpe ratio the way a naive look at higher returns might suggest.

Take an investment's return, subtract the risk-free rate, and divide by its standard deviation. Suppose a fund returned 10% over a year, the risk-free rate was 4%, and the fund's standard deviation was 12%. The excess return is 10% − 4% = 6%, and the Sharpe ratio is 6 ÷ 12 = 0.5.

Now compare a second fund that returned the same 10% but with a standard deviation of only 8%. Its excess return is still 6%, but the ratio is 6 ÷ 8 = 0.75. Same return, less volatility, higher Sharpe ratio, so the second fund delivered its result more efficiently per unit of risk.

A third fund returned a headline-grabbing 16% but with a standard deviation of 28%. Its excess return is 16% − 4% = 12%, and its ratio is 12 ÷ 28 ≈ 0.43, the lowest of the three despite the highest raw return. That is the whole point of the measure: it refuses to reward a return that was bought with an outsized amount of risk. All figures are illustrative.

Pros

• It reduces the tradeoff between return and risk to a single comparable number, so investments of different volatility can be ranked fairly.
• It is widely reported and simple to compute from return, the risk-free rate, and standard deviation.
• It exposes returns that were achieved only by taking on large amounts of volatility.

Cons

• It defines risk as standard deviation, so it penalizes upside swings as if they were losses.
• It is backward-looking and highly sensitive to the period measured.
• It understates strategies with rare but severe losses, which can show a high ratio until the tail event hits.
• It has no universal "good" threshold and is only meaningful as a relative comparison over the same period and risk-free rate.