Win Rate

What is Win Rate?

Win Rate, sometimes called hit rate or batting average, counts the fraction of return periods that finished in the green. In a backtest of equity portfolios, the natural period is a calendar month, so Win Rate is usually reported as the share of monthly portfolio returns greater than zero. A strategy with 60 percent Win Rate had positive returns in 60 percent of the months observed and negative or flat returns in the other 40 percent.

For broad equity strategies, monthly Win Rate typically sits in the 55 to 65 percent range over multi-decade windows. The S&P 500 itself lands close to 62 percent, a useful baseline. Concentrated factor strategies often fall a bit below the index, sometimes 50 to 58 percent, because they take more idiosyncratic risk and have larger month-to-month swings. Trend-following and momentum strategies famously run lower Win Rates (often near 40 to 50 percent) but earn their living from the few months when they catch a big move. The takeaway is that a high Win Rate is one signature of a strategy, and a low Win Rate is another, neither one inherently better.

The most important point about Win Rate is what it does not measure: magnitude. A strategy can win 80 percent of months with small gains and then give it all back in one terrible month, producing a beautiful Win Rate and a poor compounded return. Conversely, a strategy can win only 45 percent of months but post a respectable long-run return if the winning months are systematically larger than the losing months. Win Rate and average payoff per win are two halves of the same coin, and a Win Rate read in isolation is genuinely misleading.

Formula

The denominator is the total count of return observations across the backtest window. The numerator is the subset of those observations where the period return came in strictly above zero. Periods of exactly zero are conventionally excluded from both the winning and losing tallies, though for equity portfolios this almost never matters because a monthly return of precisely zero is vanishingly rare. The result is expressed as a percentage between 0 and 100.

The choice of period length matters more than people expect. Returns are mean-reverting toward a positive drift over time, so a longer averaging window mechanically raises the share of periods above zero. Daily Win Rates for equity strategies typically sit near 53 to 55 percent, monthly Win Rates near 60 to 65 percent, and annual Win Rates well above 70 percent. Whenever a Win Rate is quoted, the period it was computed on should be quoted with it. Comparing a monthly Win Rate to a daily Win Rate is a category error, even for the same strategy.

Why Win Rate Matters in Backtesting

Win Rate is a consistency signal. Behavioral research has shown repeatedly that investors tolerate a series of small losses far worse than a single large one of equal cumulative magnitude. A strategy that loses 1 percent in nine months out of ten before posting one giant winner has the same expected return as a strategy that loses big once a decade, but the lived experience is completely different. Win Rate quantifies the texture of that experience: how often did the equity curve tick up versus tick down? For an investor evaluating whether they can actually stay invested in a strategy, that texture matters as much as the headline return.

Win Rate is also useful for diagnosing what kind of strategy you are looking at. Mean-reversion strategies (buying weakness, selling strength) tend to have high Win Rates and small average wins, with rare large losses when the reversion fails. Trend-following strategies have low Win Rates and large average wins, with frequent small losses when the trend never materializes. Buy-and-hold equity sits in the middle. Reading Win Rate alongside the average win and average loss is a fast way to triangulate the underlying behavior of a strategy without diving into the trade-by-trade detail.

For risk-adjusted comparisons, Win Rate complements but does not replace ratios like Sharpe and Sortino. Two strategies with identical Sharpe ratios can have very different Win Rates: one with a 70 percent Win Rate and modestly asymmetric losses, another with a 50 percent Win Rate and modestly asymmetric gains. Both have the same dispersion around the mean, but a long-horizon investor would experience them very differently. Sharpe captures dispersion. Win Rate captures cadence. They are answering different questions.

How SledgeKey Implements Win Rate

Win Rate appears on the backtest results page as a percentage figure alongside the other performance metrics. It is computed from the simulated monthly portfolio returns, derived from end-of-month portfolio values net of any transaction costs that the backtest already deducted along the way. Each monthly observation is classified as a win if it is strictly above zero and as a non-win otherwise. The number of wins is divided by the total number of observed months and expressed as a percentage.

The benchmark's Win Rate is computed the same way over the identical window and the same monthly frequency. A side-by-side reading is the most useful thing to do with the pair. A strategy that posted a 64 percent monthly Win Rate against a benchmark Win Rate of 62 percent in the same window has roughly index-like cadence, which tells you the strategy is not getting its edge from being right more often than the market, but from when it is right or how much it wins by. A strategy at 70 percent against the index's 62 percent is fundamentally a higher-frequency winner, which signals a different underlying behavior even if returns are similar.

The Win Rate figure travels with the rest of the headline metrics across the results page and the saved-backtest history. When viewing two strategies side by side, comparing their Win Rates against each other and against the benchmark, ideally in the same window, is a fast way to spot whether the strategies are behaving similarly under the hood or whether they are getting to similar returns through very different paths.

Common Pitfalls

The largest pitfall is treating Win Rate as a proxy for profitability. Win Rate has no relationship to expected return on its own. A strategy that wins 1 percent in 90 percent of months and loses 12 percent in the other 10 percent has a 90 percent Win Rate and a negative expected return. A strategy that wins 8 percent in 40 percent of months and loses 2 percent in the other 60 percent has a 40 percent Win Rate and a strongly positive expected return. The right way to read Win Rate is always in tandem with the size of the wins and losses, never alone.

A second pitfall is treating Win Rate as a stable property of a strategy when it depends heavily on the period length. The same strategy can have a 53 percent Win Rate at daily frequency, 62 percent at monthly, and 78 percent at annual. None of those numbers are wrong, but they are not interchangeable. Comparisons across periods must hold the period length constant. When a strategy's marketing material quotes a 90 percent Win Rate without specifying the period, that omission is almost always intentional, because moving to a longer period mechanically inflates the figure.

A third pitfall is forgetting that Win Rate is sample-dependent in a different way than ratio metrics. Win Rate is robust to a few outlier months in the sense that one big winner does not change the count of winning months at all. But it is sensitive to small return shifts near zero. A strategy whose returns are skewed just slightly toward small positive numbers can post a Win Rate well above 50 percent even when the average return is barely positive. That sensitivity is part of why Win Rate should always be read alongside the magnitude of average wins and losses, never in isolation.

Related Concepts