Comparing Risk-Adjusted Returns: Start with How Risk Is Measured
A higher-return strategy is not necessarily better relative to risk. Learn Sharpe ratio calculation requirements, pitfalls with negative ratios, and how to assess drawdowns and benchmark performance.
Risk-adjusted return evaluates returns together with the risks taken to earn them. The Sharpe ratio is a common measure that divides excess return above the risk-free rate by return volatility. Even identical figures are difficult to compare directly when measurement periods, costs, or return frequencies differ.
Match the periods in the numerator and denominator first
Dividing average monthly excess return directly by annual volatility mixes incompatible units. Align the observation intervals in the underlying data and check how annualization was performed. Conventional square-root annualization can be inaccurate when returns exhibit strong dependence over time. Choose a risk-free reference appropriate to the investment currency and measurement period as well. Do not apply the reference rate for a dollar strategy unchanged to a won strategy, or mix returns before and after costs.
A high Sharpe ratio can miss important loss patterns
Standard deviation includes both upward and downward fluctuations. A strategy that earns frequent small gains but occasionally suffers very large losses can look attractive if no major event appears in the sample. Infrequently traded assets whose prices are rarely updated can also have understated volatility. Rather than ranking strategies solely by Sharpe ratio, also check maximum drawdown, liquidity, leverage, and extreme-loss scenarios.
Interpret negative ratios and relative performance separately
When excess return is negative, greater risk moves the ratio closer to zero, making the number alone appear less unfavorable. Simply ranking a higher Sharpe ratio as always better can therefore mislead. The information ratio, which assesses performance relative to a benchmark, uses benchmark-relative excess returns and their volatility rather than returns above a risk-free rate, so it answers a different question. First define the purpose of the comparison, then check long-term consistency and missing data.
Also check whether start and end dates were selected to make performance look good. Dividing the same strategy's record into several periods and checking whether the ratio holds up reveals dependence on one market environment. Live trading and simulated calculations differ in costs, order feasibility, and omitted assets. Even if displayed side by side, they should not be treated as equivalent evidence. The scope and limitations of the calculation data matter more than decimal places in a high ratio.
Worked example
Hypothetical example: On the same annual basis, A has an arithmetic average return of 10%, a risk-free rate of 2%, and volatility of 16%, giving a simple Sharpe ratio of (10−2)÷16=0.5. B returns 8% with the same risk-free rate and volatility of 8%, giving 0.75. A earns a higher return, but B delivers more excess return per unit of risk under these assumptions. If the strategies differ in drawdowns and trading costs, this figure alone cannot determine the final choice. Do not casually substitute a compound annual growth rate into the numerator either.
What to check
- Have you matched the periods for average excess return and volatility?
- Are the currency and treatment of costs consistent?
- Have you considered the ranking pitfall with negative excess returns?
- Have you also checked MDD and liquidity risk?
Sources and primary materials
This is educational material about research methods. For an actual company decision, verify the filings and primary sources for the relevant period.