Risk/reward ratio and win rate: table, expectancy and costs

7 min read

People often debate whether it matters more to win often or to win big when you win. Both matter at once, and there is a simple formula that combines them. Here you will find the break-even win rate table, worked expectancy examples and the factors that most distort the ratio in practice.

What the risk/reward ratio is

It is the distance to the target divided by the distance to the stop. If you buy USD/JPY at 150.00 with the stop at 149.60 (40 pips) and the target at 151.20 (120 pips), the ratio is 1:3: you would win three times what you risk. The ratio describes the planned trade, not the outcome.

Expressed in R, the stop is worth −1R and the target +3R. We will use that language throughout this guide, because it allows comparison independent of account size.

The break-even win rate: a table

To avoid losing money (before costs) you need to win at least 1 ÷ (1 + ratio), where the ratio is reward ÷ risk. Values: ratio 0.5 → 66.7%; 0.75 → 57.1%; 1 → 50%; 1.5 → 40%; 2 → 33.3%; 3 → 25%; 4 → 20%; 5 → 16.7%.

Notice the diminishing returns: going from 1:1 to 1:2 lowers the minimum by 16.7 points; going from 1:4 to 1:5 lowers it by only 3.3. Stretching the target works well at first and hardly at all later, while its cost, price reaching it less often, keeps growing.

Expectancy, with examples

Expectancy (in R) = win rate × ratio − (1 − win rate). With a 40% win rate and a 1:2 ratio: 0.4 × 2 − 0.6 = +0.20R per trade. With 35% and 1:2: 0.35 × 2 − 0.65 = +0.05R, almost zero. With 55% and a 1:0.8 ratio: 0.55 × 0.8 − 0.45 = −0.01R; a high win rate does not rescue a strategy with small wins. With 30% and 1:3: 0.3 × 3 − 0.7 = +0.20R.

When results are not uniform, use real averages. Over 40 trades: 18 winners averaging 85 USD and 22 losers averaging 52 USD. Gains 18 × 85 = 1,530; losses 22 × 52 = 1,144; net 386 USD, that is 9.65 USD per trade, or 0.19R if your average R was 52 USD. Over 200 trades at +0.2R, risking 50 USD, expected profit is 2,000 USD before costs, but with large swings along the way.

Costs: spread and commission distort the ratio

Costs add to your loss and subtract from your gain. Example on EUR/USD with a 1.2-pip spread, a 10-pip stop and a 20-pip target measured from the bid price on the chart (planned 1:2). You buy at the ask, 1.2 pips higher: if the stop is hit you lose 11.2 pips, and if the target is reached you gain 18.8. The real ratio is 18.8 ÷ 11.2 = 1.68, not 2.

Add a round-turn commission of 7 USD per lot (0.7 pips at 10 USD per pip) and the loss is 11.9 pips and the gain 18.1: a ratio of 1.52. The break-even win rate moves from the theoretical 33.3% to 39.7%. The same 1.9-pip cost on a 40-pip stop and 80-pip target gives a ratio of 1.86 and a minimum of 34.9%. Costs weigh more the smaller your stops are.

Partial exits: they change the shape of results

Closing part of the position along the way reduces variance, but it does not hand you extra expectancy. Say you have a 3R target, close half at +1R and move the stop on the rest to entry. Possible outcomes: stopped before 1R (−1R), reaching 1R and returning to entry (+0.5R), or reaching 3R (+2R in total, 0.5 × 1 + 0.5 × 3). With probabilities of 40%, 25% and 35%, expectancy is −0.4 + 0.125 + 0.7 = +0.425R.

Without the partial, with the same price behaviour, the 25% that return to entry would become −1R, and the 35% would win +3R: 0.35 × 3 − 0.65 = +0.40R. The difference is only +0.025R: the partial turns losses into small gains (+0.375R) but trims the best trades (−0.35R). If it helps you follow your plan, it is valid; but measure it with your data rather than assuming it is better.

Trailing stops: the ratio stops being fixed

With a trailing stop there is no fixed target, and therefore no planned ratio: the average win depends on how long the trend lasts. It is typical of trend-following systems, with few wins and large gains. Example: 35% win rate, an average win of +2.8R and an average loss of −1R: 0.35 × 2.8 − 0.65 = +0.33R.

In that case the relevant figure is the realized ratio: average win in R divided by average loss. Measure it after at least several dozen trades, and do not expect it to resemble the one you pictured.

Why a high R:R does not guarantee profit

The further away the target, the less often price gets there. A 1:5 ratio with a 12% win rate has expectancy 0.12 × 5 − 0.88 = −0.28R, a losing one, because the minimum is 16.7%. Imagine three targets on the same entry, with hypothetical win rates: at 1R, 55%; at 2R, 38%; at 3R, 28%. The expectancies are +0.10R, +0.14R and +0.12R: the best is not the one with the highest ratio.

The ratio is a consequence of market structure, not a free parameter. Tuning it until it “looks right” without checking the effect on your win rate is an elegant way of fooling yourself.

How big a sample you need to trust your win rate

A 40% win rate observed over 50 trades has a standard error of √(0.4 × 0.6 ÷ 50) ≈ 6.9 points; the approximate 95% interval is about ±13.6 points, that is, between 26% and 54%. At a 1:2 ratio, whose minimum is 33.3%, you cannot tell a profitable strategy from a losing one with that sample. With 200 trades, the margin shrinks to about ±6.8 points.

The practical consequence: do not change ratio or strategy after 15 trades, and do not trust a backtest with few trades.

A full example with costs, start to finish

A 10,000 USD account, 1% risk (100 USD). You want to buy EUR/USD with a 25-pip stop and a 50-pip target (planned 1:2), a 1.0-pip spread and a 6 USD round-turn commission per lot, that is 0.6 pips. Total cost is 1.6 pips. Effective loss: 25 + 1.6 = 26.6 pips; effective gain: 50 − 1.6 = 48.4 pips; real ratio 48.4 ÷ 26.6 = 1.82.

The lot is calculated from the effective loss, not the stop you drew: 100 ÷ (26.6 × 10) = 0.376, so 0.37 lots. If the stop is hit you lose 0.37 × 26.6 × 10 = 98.42 USD; if the target is reached you gain 0.37 × 48.4 × 10 = 179.08 USD, that is +1.79R. The break-even win rate is no longer 33.3% but 1 ÷ (1 + 1.82) = 35.5%. A small difference on one trade, and a large one accumulated over hundreds.

Common mistakes with the ratio

Do not force the ratio: it is tempting to push the target out so the quotient “looks good”, but a target price rarely reaches lowers your win rate. Nor should you tighten the stop until normal market noise triggers it often: you would improve the ratio on paper and worsen the win rate in practice.

Other mistakes: mixing trades with very different ratios in one statistic without converting them to R, calculating the win rate without counting trades closed at breakeven, and moving the stop to entry too early, which turns many trades that would have won into scratch exits.

Summary: using the ratio without fooling yourself

First, place stop and target at chart levels, and calculate the ratio afterwards. Second, include spread and commission in both loss and gain. Third, calculate the break-even win rate from the net ratio, not the theoretical one.

Fourth, record your trades in R and compute your real expectancy. Fifth, do not judge on fewer than several dozen trades. Sixth, if you use partial exits or trailing stops, compare with and without them using your own data. You can try your numbers with the risk/reward calculator.

Educational content, not financial advice. Risk disclosure

Related tool

Risk/reward calculatorMeasure the R:R ratio, the win rate you need and your mathematical expectancy.

More guides