Sharpe ratio

The Sharpe ratio is the mean return of a series divided by the standard deviation of that series, so it measures how much result was obtained per unit of dispersion. In trade-level analysis it is usually computed per trade and left unannualised.

Senzoukria · Glossary · Updated September 2026

Definition

Take a series of returns. Subtract a reference return if one applies, compute the mean, then divide by the standard deviation of the same series. The result is dimensionless: a strategy with a large average gain and a large dispersion can score the same as a strategy with a small average gain and a small dispersion.

For discretionary and intraday futures trading the series is often the list of trade results in ticks, dollars or R, and the reference return is zero. This gives a Sharpe per trade. Annualised Sharpe ratios multiply the per-period value by the square root of the number of periods in a year, which assumes the periods are independent and identically distributed.

Per trade versus annualised

  • A per-trade Sharpe compares strategies with the same trade frequency. It does not compare a scalper with a swing trader.
  • Annualising one session assumes the year looks like that session. Over a short sample the annualised figure is mostly an artefact of the multiplier.
  • The ratio penalises upside and downside dispersion equally. The Sortino ratio divides by downside dispersion only.
  • A few large winners inflate the mean and the dispersion at the same time, so the ratio can stay modest even when the equity curve looks impressive.

In Senzoukria

The Replay performance panel of the desktop reports Sharpe / trade with the hint 'Mean / standard deviation. NOT annualised: annualising one session would assume the year looks like it.' Sortino / trade sits next to it. The same panel warns when the sample is too small for the ratios to mean anything and shows them because they are exact, not because they are significant.

In the backtest panel, Sharpe per trade is one of the three objectives the walk-forward optimiser can maximise, alongside expectancy per trade and SQN. A Rolling Sharpe chart measured over the last N session days shows where a whole-period figure would hide a strategy that stopped working halfway through.

Common mistakes

  • Comparing an annualised Sharpe from a fund report with a per-trade Sharpe from a backtest.
  • Reading a Sharpe ratio on a handful of trades. The standard deviation of a small sample is itself unstable.
  • Ignoring selection: the highest Sharpe among many configurations is inflated by the number of trials. See the deflated Sharpe ratio.
  • Using the Sharpe ratio as a drawdown measure. It says nothing about the order of the returns.

This page in other languages

Frequently asked questions

Can the Sharpe ratio be negative?
Yes. When the mean return is below the reference return the numerator is negative and so is the ratio. A negative Sharpe still carries information about dispersion, but comparisons between two negative values are hard to interpret because a higher dispersion makes the ratio less negative.
Why does the desktop refuse to annualise the Sharpe ratio?
Annualising multiplies a per-period value by the square root of the number of periods per year and assumes each period resembles the sample. One replay session or one short backtest does not represent a year of trading. The desktop shows the per-trade figure and lets you decide whether a longer sample justifies a projection.

Keep reading

Strategy researchTest a trading strategy, from observation to challenge rulesDefine an observation, question it with the copilot, write a rule and explore costs, drawdown and fictional challenge limits in an interactive example.Quant · Start hereQuant trading for beginners: from idea to backtestDefine a testable rule, choose suitable data and evaluate later periods after costs.Order flow · HubOrder flow trading: the complete guideThe hub. What order flow is, what it shows — aggression, delta, imbalance, absorption, liquidity — and the tools that read it, each linked to a full guide.