Sharpe ratio — what it really measures and where it breaks

The Sharpe ratio is the single most cited measure of risk-adjusted return in capital management, from pension funds to retail prop firms. The formula is brutally simple: return minus the risk-free rate, divided by the standard deviation of returns. The number tells you how much extra profit you collected per unit of volatility you were willing to live with. The wrinkle most beginners miss sits in the assumptions and the time window — and that is where most of this article focuses.

Where the Sharpe ratio came from and what it actually measures

William F. Sharpe published the first version in 1966 in his paper "Mutual Fund Performance" in the Journal of Business, where he called it the reward-to-variability ratio. He won the 1990 Nobel Memorial Prize in Economic Sciences for his work on the Capital Asset Pricing Model, and four years later revisited the measure in the Journal of Portfolio Management, settling on the modern name. The contemporary formula: portfolio return minus the risk-free rate (usually the yield on short Treasury bills), divided by the standard deviation of the excess returns.

The point is not to ask how much you earned but how much you earned above what a safe instrument would have paid, and how much the ride shook on the way. A strategy delivering 30 percent annually with a 10 percent standard deviation looks better through this lens than one delivering 50 percent with a 30 percent standard deviation, because it extracts more excess return per unit of volatility. Institutions quote the Sharpe ratio in their factsheets because it lets an analyst compare an equity fund with a bond fund, or a systematic futures programme with a balanced portfolio, even though the underlying return profiles look nothing alike.

The formula in plain prose and what to plug in

The numerator is the difference between the average annualised portfolio return and the risk-free rate. The denominator is the annualised standard deviation of the excess returns. In practice you compute it from daily or monthly results: take logarithmic returns, work out the mean and standard deviation, then scale to a year. The mean gets multiplied by the number of periods in a year (252 for daily, 12 for monthly), and the standard deviation by the square root of the same number. What comes out is a dimensionless figure such as a Sharpe of 1.4.

What goes into the risk-free leg? The yield on short government debt in the same currency as the strategy. For dollar strategies that is the three-month Treasury bill yield; for zloty portfolios, short Polish government bills or the WIBOR rate. In 2020, when the Federal Reserve held rates near zero, the denominator looked different from 2024, when the short end was paying close to 5 percent. The same strategy can show two different Sharpe ratios on opposite ends of the monetary cycle.

What actually counts as a good number

The thresholds most often repeated in the industry: a figure below 1.0 is weak. A value of roughly 1.0 to 1.5 is acceptable. The 1.5 to 2.0 band is considered solid. Above 2.0 is excellent. Above 3.0 you have reached institutional territory. The S&P 500 itself, measured over multi-decade windows, lands around 0.4 to 0.6, because equities carry meaningful volatility. The average hedge fund reports a Sharpe of about 0.8 to 1.0. Renaissance Technologies' Medallion fund has hovered around 2.5 for years — a number analysts treat as simply exceptional.

A retail trader showing a Sharpe of 2.0 on three years of myfxbook data has not, on its own, proved much. The window is short, the sample small, and the figure heavily dependent on the market regime. A six-month Sharpe of 3.0 collected during a clean trending phase may be a statistical artefact rather than evidence of a durable edge. Institutional analysts typically demand at least a 36-month window, ideally covering a full market cycle.

„The Sharpe Ratio is one of the most often-cited measures of investment performance. In part this is because it is so simple. In part it is because the measure directly indicates the trade-off between additional return and additional risk." — William F. Sharpe, The Sharpe Ratio, The Journal of Portfolio Management, 1994

A worked example in prose

Illustrative example. Two strategies tested over a three-year window. The first earns 24 percent a year with a standard deviation of 16 percent. The second earns 12 percent a year with a standard deviation of 5 percent. The risk-free rate is 4 percent. For the first, 20 percentage points of excess return divided by 16 of volatility lands a Sharpe of 1.25. For the second, 8 divided by 5 lands a Sharpe of 1.6. The second strategy, despite earning half as much, looks better through this lens because it squeezes more excess return out of each unit of volatility.

A practical intuition newer traders often miss. If you halve the position size in the first strategy, excess return falls to 10 percent, volatility to 8 percent, and the Sharpe stays at 1.25 — because the ratio is a relationship, not an absolute level. Leverage and position size, by themselves, do not change the Sharpe ratio. What changes it is the quality of the edge: whether the proportion between expected return and risk is fundamentally better.

Three known limitations worth keeping in mind

First problem. The ratio punishes volatility in both directions. A single 15 percent positive month inflates the standard deviation just as much as a deep loss would. Sharpe treats "good" and "bad" volatility as the same thing, which is precisely why Frank A. Sortino proposed an alternative in the mid-1990s in which the denominator counts only downside deviations. Our article on the Sortino ratio and the direct side-by-side comparison walk through how the fix is constructed.

Second problem. The formula assumes returns follow a normal distribution. Financial markets — especially emerging-market currencies and CFDs on commodities — exhibit fat tails. The Swiss franc move in January 2015, the March 2020 crash, the Turkish lira in 2018: outcomes so far from the mean that classical standard deviation seriously understates true tail risk. A high Sharpe computed over a quiet window can come undone in one week of crisis. That is not a calculation error but a built-in limitation of the assumptions.

Third problem. The figure depends heavily on sample length and market regime. A trend-following strategy shows a high Sharpe when currencies move cleanly and a much lower one in a sideways phase. A Sharpe over six months is statistically close to meaningless. This is why institutional scorecards report Sharpe alongside maximum drawdown, the Calmar ratio, and measures such as expected value per trade — no single parameter describes a strategy honestly.

What to do tomorrow

1. Open your spreadsheet of past trades or broker statement for the last twelve months and compute a Sharpe ratio on monthly data. Take the average monthly return, subtract the risk-free rate divided by twelve, divide by the monthly standard deviation, and multiply by the square root of twelve to get an annualised figure you can compare with standard benchmarks.
2. If the result lands below 1.0, do not change strategy yet. Check two things first: whether the sample is representative (six months is too little, only thirty-six months carries statistical weight) and whether the figure is being dragged down by a handful of extreme months that distort the standard deviation in both directions.
3. Compute the Sortino ratio over the same window in parallel. If the Sortino comes out materially higher than the Sharpe, your volatility is asymmetric: upside spikes, few deep drawdowns. Such a strategy may be better than Sharpe alone suggests, and the honest move is to publish both numbers together rather than the prettier one.
4. Add a column to your journal recording annualised return and monthly standard deviation, so Sharpe is computed automatically each month. After a year, compare the figure with the prior period — you care about the trend, not any single reading, because Sharpe on short windows is noisy.
5. Before sharing a Sharpe ratio with capital allocators or a prop-firm reviewer, make sure the window covers at least two years and includes both a quiet phase and a phase of elevated volatility. A Sharpe drawn from a single trend regime misleads both you and any potential investor reading the number.