An 80% Win Rate Can Still Lose Money.

A strategy wins eight trades out of ten. The trader still loses money. Nothing about that sentence is contradictory: win rate describes how often a result is positive, not how large the positives and negatives are.

Before searching for a higher hit rate, ask whether the current method earns enough when right to pay for being wrong—and for the cost of trading.

Start with the arithmetic

For a simplified two-outcome model, expected profit per trade equals the probability of winning multiplied by the average win, minus the probability of losing multiplied by the average loss. CME Group’s explanation of trading expectancy sets out that relationship.

For practical recordkeeping, express results in R, where 1R is the initial planned monetary risk on that trade. If the initial risk is $100, a $50 gain is +0.5R and a $200 loss is -2R. A loss larger than 1R is possible: gaps, slippage, execution failures or moving a stop can defeat the original plan.

Keep the accounting consistent. Either calculate wins and losses net of every cost, or calculate gross expectancy and subtract average costs separately. Doing both deducts the same fees twice.

Two fictional strategies

The following figures are arithmetic illustrations, not backtest results or forecasts. Assume average round-trip costs of 0.08R per trade, including the spread, fees and modeled slippage.

Strategy A’s calculation is 0.80 × 0.35 minus 0.20 × 2.00: -0.12R before costs. Strategy B’s is 0.45 × 1.80 minus 0.55 × 1.00: +0.26R. The lower-win-rate illustration has the better expectancy, but neither set of assumed numbers establishes a real edge.

The table also hides the path. A positive average can coexist with uncomfortable losing streaks and deep drawdowns. Position sizing determines whether a trader survives that path; the average alone does not.

Planned reward is not realized reward

A drawn 3:1 target is not evidence that average winners reach 3R. Early exits, partial profits, missed fills and rules changed during a trade alter the realized distribution.

Likewise, a chart stop 1% away is not a guarantee of a 1% fill. For an illiquid market or a fast move, the difference between a planned exit and an executable exit can dominate the result. Include funding when it actually applies; do not assume every short scalp pays it or every overnight position faces the same schedule.

A journal that can answer the question

For each trade, retain the initial risk, actual entry and exit fills, fees, applicable funding, net result and reason for exit. Preserve the original plan rather than rewriting it after seeing the outcome.

In Basis, use the position/risk tools to specify the plan and the journal to record what happened. The position-size calculator can help with sizing arithmetic. It cannot establish that the setup deserves a trade. If a journal field does not cover a cost, put it in the notes and include it in your net calculation.

Separate materially different setups. Mixing trend entries, range fades and impulsive trades into one average can make an unprofitable behavior appear supported by a profitable one. But avoid slicing a small sample into dozens of flattering subgroups; that can manufacture apparent winners by chance.

Test the result, not the headline

Use data not involved in choosing the rules. Check whether the result survives less favorable costs, slower entries and periods unlike the one that inspired the strategy. Report sample size and drawdown alongside expectancy. Twenty good trades are observations, not a dependable probability estimate.

The right question is not “Can I make the win rate higher?” It is “After realistic costs, does this precisely defined behavior produce a repeatable result on data I did not optimize it against?” Sometimes the honest answer is not yet. That is more useful than an impressive percentage.

Method note: the examples assume fixed average outcomes and costs only to demonstrate the calculation. They do not model compounding, changing risk, liquidation or the full distribution of returns. Reviewed September 16, 2026. Educational material, not personalized investment advice.