Sharpe Ratio: Return Per Unit of Risk Explained

Key Takeaways

• Sharpe ratio divides excess return over the risk-free rate by the standard deviation of those excess returns; a reading of 1.0 or above is generally considered reasonable for an equity strategy.
• Annualising a Sharpe ratio requires multiplying by the square root of the number of periods, not by the number itself, a common error that inflates daily figures by a factor of 15.
• Option-selling and short-volatility strategies can print high Sharpe ratios for years and then hand back a decade of gains in a single month, because standard deviation understates their real risk.
• A 36-month Sharpe of 1.0 carries a wide confidence interval; statistical significance requires several years of data before skill can be separated from luck.

What It Is

William F. Sharpe introduced the ratio in 1966 as part of his work on mutual fund performance, then revised it in 1994 in his paper The Sharpe Ratio in the Journal of Portfolio Management. The 1994 revision generalises the measure to any two-asset comparison and formalises it as the expected differential return per unit of differential volatility.

In practice, the Sharpe ratio compares a portfolio's excess return over a benchmark, usually a risk-free rate, against the standard deviation of that excess return. Higher is better. A higher number means the portfolio delivered more reward per unit of bumpiness.

The Intuition

Return alone tells an incomplete story. A portfolio that earned 15 percent could have done so calmly or could have whipsawed investors through a 40 percent drawdown along the way. The Sharpe ratio is the simplest attempt to put those two experiences on the same axis.

The construct is a reward-to-variability ratio. The numerator is what you got paid above a riskless alternative. The denominator is how much your returns jumped around. If the numerator is large relative to the denominator, the strategy compensated you well for the risk you took. If the numerator is small relative to the denominator, you got paid little for a lot of turbulence.

Sharpe's 1994 paper is careful to frame the ratio as the expected value of a ratio of differentials, not a guarantee about realised outcomes. The distinction between ex-ante (forward-looking) and ex-post (historical) Sharpe ratios matters and is often blurred in practice.

How It Works

The standard formula is:

Sharpe = (Rp - Rf) / sigma_p

Where:

Rp      = portfolio return over the measurement period
Rf      = risk-free rate over the same period (often 3-month T-bills)
sigma_p = standard deviation of the portfolio's excess return (Rp - Rf)

The 1966 original used total return volatility in the denominator. The 1994 revision uses the volatility of the excess return series, which is the statistically cleaner choice when the risk-free rate itself moves. For most equity portfolios the two are close, but for fixed income or highly leveraged strategies the distinction matters.

To annualise from a higher-frequency estimate, you scale the mean by the period count and the standard deviation by the square root of the period count:

annualised Sharpe = sqrt(N) * (mean periodic excess return / stdev periodic excess return)

Where N is 252 for daily trading data, 52 for weekly, or 12 for monthly. The sqrt(N) scaling assumes returns are independent and identically distributed, which is an approximation.

A rough reading guide used in the industry: below 1 is weak, 1 to 2 is reasonable, 2 to 3 is very good, and above 3 is excellent but rare and warrants scrutiny. These are conventions, not laws.

Worked Example

A long-only US equity portfolio has the following monthly record over the last year.

• Average monthly return: 1.2 percent
• Average monthly risk-free rate (T-bill): 0.35 percent
• Standard deviation of monthly excess returns: 3.1 percent

Monthly Sharpe:

(1.2% - 0.35%) / 3.1% = 0.85% / 3.1% = 0.274

Annualised:

0.274 * sqrt(12) = 0.274 * 3.464 = 0.95

The portfolio produced a Sharpe of about 0.95 over the year. That is broadly in line with long-run US equity indices, which historically sit around 0.4 to 0.5 on a long sample but can run much higher in a single bull year. A Sharpe of 0.95 is not evidence of skill on one year of data alone, because standard errors on a 12-month Sharpe estimate are large.

Frequently Asked Questions

Q: What is the Sharpe ratio in simple terms? The Sharpe ratio tells you how much extra return you earned for each unit of volatility you accepted. A ratio of 1.0 means you earned 1% above the risk-free rate for every 1% of standard deviation, considered a solid result for a long-only equity strategy.

Q: How does the Sharpe ratio affect investment decisions? Investors use it to compare strategies that differ in both return and volatility on a common scale. A fund with a 12% return and 15% standard deviation (Sharpe ~0.5) may be less attractive than one with 9% return and 5% standard deviation (Sharpe ~1.0), depending on the context.

Q: What is a real-world example of the Sharpe ratio? A US equity portfolio with a 1.2% average monthly return, 0.35% monthly risk-free rate, and 3.1% monthly standard deviation produces a monthly Sharpe of 0.274. Annualised via sqrt(12), that is 0.95, broadly in line with a typical equity bull-market year.

Q: How can investors avoid misusing the Sharpe ratio? Always pair it with maximum drawdown. A strategy collecting option premiums can show a Sharpe of 3 for years until a crash wipes out the account, because volatility undercounts the true downside risk. Use Sortino or conditional VaR alongside Sharpe for skewed strategies.

Q: How is the Sharpe ratio different from the Sortino ratio? The Sharpe ratio uses total return standard deviation (up and down moves) in the denominator. The Sortino ratio uses only downside deviation, returns below the minimum acceptable return. For strategies with strong positive skew, Sortino is higher than Sharpe; for symmetric returns, they are similar.

Common Mistakes

1. Using a backward-looking Sharpe as a forecast. Historical Sharpe is a point estimate with real uncertainty. For a 36-month sample, the 95 percent confidence interval around a realised Sharpe of 1.0 can easily span from 0.3 to 1.7. Treating the past number as the expected future value consistently overstates what you will earn.

2. Applying Sharpe to non-normal distributions. The ratio implicitly assumes returns are roughly symmetric and that standard deviation captures the risk that matters. Hedge fund strategies, option-selling strategies, and anything with fat tails or strong skewness violate that assumption. A strategy that collects small premiums and occasionally blows up (selling out-of-the-money puts, for example) can print a Sharpe of 3 for years and then hand back a decade of gains in one month. Volatility under-measures the real risk.

3. Annualising by multiplying by N instead of sqrt(N). Volatility scales with the square root of time under the i.i.d. assumption, not linearly. A daily Sharpe of 0.05 annualises to 0.05 * sqrt(252) = 0.79, not 0.05 * 252 = 12.6. The linear error produces absurd numbers that occasionally make it into marketing decks.

4. Comparing Sharpe across asset classes with very different distributional properties. A bond fund's Sharpe, a long-equity Sharpe, and a managed-futures Sharpe are not directly comparable. The bond fund's returns are smoother but can have sharp left tails; the managed-futures program often has positive skew. Pick comparisons within a peer group.

5. Ignoring the risk-free rate choice. In a zero-rate environment the numerator is close to the raw return and the ratio looks fine. When short rates rise, the same strategy's Sharpe falls mechanically even if nothing about its behaviour changed. Always state the Rf source and tenor.

6. Treating volatility as the same thing as risk. Sharpe uses standard deviation as a proxy for risk. For a retail investor the real risk is more often loss of capital or deep drawdown. Pair Sharpe with maximum drawdown and the Sortino ratio before drawing conclusions.

Sources

1. Sharpe, W.F. (1994). "The Sharpe Ratio." The Journal of Portfolio Management, 21(1), 49-58. Author's page: https://web.stanford.edu/~wfsharpe/art/sr/sr.htm
2. The Journal of Portfolio Management. "The Sharpe Ratio" (1994 archive). https://jpm.pm-research.com/content/21/1/49
3. The Hedge Fund Journal. "Sharpe Ratio." https://thehedgefundjournal.com/sharpe-ratio/
4. QuantStart. "Sharpe Ratio for Algorithmic Trading Performance Measurement." https://www.quantstart.com/articles/Sharpe-Ratio-for-Algorithmic-Trading-Performance-Measurement/

Disclaimer

This article is educational content only and is not financial advice. Nothing here is a recommendation to buy, sell, or hold any security. Consult a licensed advisor before making investment decisions.