Backtest win rate is the share of a ticker’s past BUY entries that closed with any positive return by the horizon end. If eight of ten past BUY entries finished even slightly green, the win rate is 80% — regardless of how big those gains were.
Precision counts only the BUY signals that cleared a real return threshold (default +5%); win rate counts any gain above zero, however small. So win rate is almost always the higher, softer number of the two. A name can show a high win rate and a much lower precision — lots of tiny green closes, but few that reached a meaningful target. Reading win rate on its own can flatter a signal that rarely delivers a move worth trading.
Like precision, win rate is a per-ticker stat measured on a small history — often only a handful of past entries — so a high reading is suggestive, not proof, and a run of small wins can flip with a single large loss. A backtest is a rewind of what already happened, not a promise of what comes next; live results routinely differ. Always read win rate alongside average return, which captures the size of the wins and losses that win rate ignores.
Win rate and average return together are the raw material for the Kelly position-sizing calculation: how often you win and how much you win or lose per trade. A high win rate with a poor average return (many small wins, occasional large losses) can still call for a small position, which is why neither number is read in isolation.
Win rate counts any positive return, however small; precision counts only the BUY signals that cleared a real threshold (default +5%). Win rate is therefore usually the higher, softer number — a name can win often yet rarely reach a meaningful target.
Yes. A string of tiny green closes gives a high win rate but little tradeable edge, and a single large loss can undo many small wins. Read win rate alongside average return, which captures the size of wins and losses.
Treat it cautiously. It is measured per ticker on a small sample, so a high reading is suggestive, not proof, and a backtest is not a guarantee of live results.
Browse all S&P 500 tickers to see this metric applied to individual companies.
Educational research only — not investment advice.