Definition

Win rate is the number of profitable completed trades divided by the number of completed trades in a sample. It describes how often a trade wins under a stated accounting convention. It does not say how much the winners earn, how much the losers lose or whether the portfolio compounds profitably.

Formula and interpretation

win rate = profitable completed trades / all completed trades

State how breakeven trades, partial exits and costs are handled. A position closed in several pieces can count differently across reports. Open positions are generally excluded from a completed-trade win rate, so unrealized losses may be absent from the statistic.

A useful companion is the payoff ratio, average win divided by the absolute average loss. In a simplified trade-level model, expected profit is the win probability times average win, minus the loss probability times average loss, before any omitted costs. This makes clear why a higher win rate need not indicate a better strategy.

A U.S. backtest example

The U.S. mean-reversion study records 107 trades and about 6.4x annual turnover over its May 2016 to October 2026 window. Its headline CAGR is 51.9% with a -27.8% maximum drawdown. These portfolio statistics do not determine its win rate. To calculate that figure, inspect the completed-trade records and their cost convention.

That distinction matters because the study holds leveraged ETFs under a regime rule. A trade count does not describe position size, time invested or leverage. Copying a portfolio CAGR into an expected profit per trade would mix different units and conceal the compounding path.

How to inspect a strategy

Open the report's trade list and check a sample of entries and exits. Separate winning frequency from average win, average loss, largest loss and time held. Check whether a small number of large winners account for most of the return. Read the monthly portfolio returns as a second view: a trade win rate and the fraction of positive months are different statistics.

Use the backtesting guide to create a report and the risk metric calculators to examine the portfolio path. If you ask an AI assistant to explain the result, request the exact denominator and treatment of open or breakeven trades.

Common pitfalls

Costs can turn a small gross winner into a net loser. Many small wins can be offset by a large tail loss. A low-frequency strategy may have too few completed trades to support a precise probability estimate. Related trades in the same market regime are also not independent observations.

Frequently asked questions

What is a good win rate?

There is no universal threshold. It depends on payoff sizes, costs, concentration, holding time and uncertainty in the sample.

Is win rate the same as accuracy?

Only if “accuracy” has been defined using the same trade outcome and sample. A model's directional forecast accuracy, a completed-trade win rate and a profitable-month ratio use different observations.

Read CAGR, slippage, turnover and the quant glossary before comparing a strategy using one ratio.

For education and research only. This is not investment advice. Historical performance does not guarantee future results.

Last updated: 2026-10. Author: FinLab Quant Research.