What the Sharpe ratio measures (and what it doesn't)
The Sharpe ratio answers one question: for every unit of volatility you took on, how much did you earn above the risk-free rate? It is the most widely used risk-adjusted return metric in finance — you will find it in fund selection, portfolio construction, backtest comparison, and the CFA curriculum.
It is not a measure of raw return: two strategies that both return 15% a year can have very different Sharpes if one got there on a much wilder ride. And it is not a guarantee — a high Sharpe on three years of data says little about year four. What it is, is a fair way to compare strategies with different risk profiles, so that a smooth 10% and a roller-coaster 15% stop looking like the same bet.
The Sharpe ratio formula
Sharpe = (Rp − Rf) / σ — the portfolio's average return minus the risk-free rate, divided by the portfolio's volatility (the standard deviation of its returns).
- Collect the period returns — daily, weekly, or monthly, but stay consistent.
- Average them to get Rp, and compute their standard deviation to get σ.
- Pick a risk-free rate (Rf) matched to your currency and horizon — typically the short government bond yield, e.g. the 3-month T-bill for USD.
- Subtract and divide: (Rp − Rf) / σ.
If your returns are quoted per year, the result is already an annualized Sharpe. If they are per day, week, or month, multiply by the square root of the periods in a year to annualize: daily × √252 (≈15.87), weekly × √52 (≈7.21), monthly × √12 (≈3.46).
Worked examples
- Annual return 12%, volatility 15%, risk-free 4% → (12 − 4) / 15 = 0.53 — "acceptable" territory.
- Annual return 15%, volatility 10%, risk-free 4% → (15 − 4) / 10 = 1.10 — "good". Similar raw return, half the risk, more than double the Sharpe.
- Daily returns averaging +0.05% with 1.0% daily volatility, risk-free ignored: per-period Sharpe 0.05, annualized 0.05 × √252 ≈ 0.79.
- Annual return 2%, risk-free 4%, volatility 10% → (2 − 4) / 10 = −0.20: a negative Sharpe means the strategy underperformed risk-free cash after taking on the volatility.
How to interpret Sharpe ratio values
Common reference bands, on the annualized ratio:
- Below 0.5 — poor. The volatility is not being compensated.
- 0.5–1.0 — acceptable. Where many diversified equity funds sit in an ordinary year.
- 1.0–2.0 — good. A strong result for an actively managed fund.
- 2.0 and above — excellent. Sustained over several years, it is rare; a 3+ on a long history almost always signals a data problem, survivorship bias, or a leverage illusion rather than skill.
For context, the US equity market itself typically carries a Sharpe around 0.4–0.6 in long-horizon annualized terms. A fund that posts 1+ consistently across a full market cycle is doing something meaningful; a fund advertising 4.5 "over the last 18 months" deserves skepticism before enthusiasm.
Mistakes and limitations that make Sharpes misleading
- Volatility is not the same as risk. The Sharpe divides by total volatility, so upside swings count against you too. A strategy with occasional huge winners (momentum, event-driven) can look worse than it is, while one with smooth, capped returns looks better. If big upside is a feature, not a bug, look at the Sortino ratio as well.
- Total risk vs. systematic risk. For a single stock, most volatility is diversifiable. Judging it by Sharpe penalizes risk you could have eliminated by holding a basket; the Treynor ratio (per unit of beta) or the information ratio (versus a benchmark) is often the fairer yardstick there.
- Short histories. Volatility estimates need long samples to stabilize. A Sharpe computed on eight weeks of daily data is mostly noise — evaluate over at least a year, ideally three to five years including a downturn.
- Leverage inflates both sides. Levering a strategy multiplies return and volatility roughly in proportion, so the Sharpe barely moves. A "3.0" Sharpe that comes from 5× leverage is the same 0.6 Sharpe underneath, just amplified — and it can collapse just as fast.
- Mismatched risk-free rate. Use a rate in the same currency and over the same horizon as your returns. The choice shifts the number a bit, but rarely changes the ranking of comparable strategies.
Calculate your Sharpe ratio
The Sharpe Ratio Calculator runs the full computation in the browser. Two modes:
- Series mode — paste period returns, one per line (a trailing % is fine; at least 2 values). The tool computes the mean and standard deviation, uses a 0% risk-free rate, and annualizes assuming daily returns (× √252). You get the per-period and annualized Sharpe plus the data point count.
- Summary mode — enter the annual average return, annual volatility, and risk-free rate directly for a one-step annualized Sharpe.
The result is rated against the standard bands (below 0.5 poor, 0.5–1.0 acceptable, 1.0–2.0 good, 2.0+ excellent), and everything stays client-side: nothing you paste is uploaded. For a broader risk-adjusted picture — profit factor, drawdown, expectancy — the Backtest Metrics tool computes the full metric set from a trade or return history.
Sharpe vs. the other risk-adjusted metrics
- Sortino ratio — (Rp − Rf) divided by downside deviation only. It ignores upside volatility, so it is usually at least as high as the Sharpe for the same strategy, and higher when big upside swings are part of the story.
- Calmar ratio — annualized return ÷ maximum drawdown. It captures the pain of the worst loss rather than everyday wobble; a good companion to the Sharpe (see the maximum drawdown guide).
- Treynor ratio — (Rp − Rf) divided by beta. It rewards only systematic risk, the part that cannot be diversified away; the right yardstick for a single holding inside a diversified portfolio.
- Information ratio — active return divided by tracking error. It answers "was the manager skilled relative to the benchmark?", a different question from "is the strategy efficient on its own?".
A practical habit: read the Sharpe for overall risk efficiency, the Calmar for worst-case pain, and the information ratio when the real question is manager skill versus a benchmark.