Expectancy in trading is the average amount a method makes or loses per trade once wins and losses are weighed by how often they happen. It answers the question a win rate cannot: if you keep taking these trades exactly as planned, do they add up to a gain or a loss?
Quick answer: what is expectancy in trading?
Expectancy is the average result per trade: (win rate x average win) minus (loss rate x average loss). Measured in R, where 1R is the amount risked, a result above zero means positive expectancy, an edge on paper. It only means something over a large sample of trades taken by the same rules, after costs.
What expectancy actually measures
Expectancy is the trader’s name for expected value. Probability courses define expected value as each possible result multiplied by its probability and added up. MIT’s 18.05 notes call it a weighted average of the possible values, and use a dice game where you win $1,000 on a roll of 2 and lose $100 otherwise.
That game shows the point well. The win is ten times the loss, yet the expected result is about minus $69 per play, because the win comes so rarely. A big average win means nothing without the rate at which it arrives, and a high win rate means nothing without the size of the losses. Our post on why win percentages are irrelevant on their own makes the same argument from the win-rate side.
The trading expectancy formula, in R and in pounds
The formula has four inputs, all taken from closed trades:
- Win rate: winning trades divided by all trades.
- Loss rate: losing trades divided by all trades (1 minus the win rate, if you count break-even trades separately or not at all).
- Average win: total won divided by the number of winners.
- Average loss: total lost divided by the number of losers, written as a positive number.
Expectancy = (win rate x average win) – (loss rate x average loss)
You can run it in pounds or in R. R is the amount you planned to lose if the stop was hit, so a trade risking £50 that makes £100 is +2R. Planning that relationship before entry is covered in risk reward ratio explained; expectancy is what the trades actually delivered afterwards.
| Factor | In R | In pounds |
|---|---|---|
| Compares methods fairly | Yes, size is stripped out | Only at the same stake |
| Survives changes in stake | Yes | No, bigger stakes inflate it |
| Shows real money per trade | Needs converting | Yes |
| Easy to add costs | Convert costs to R first | Yes, subtract directly |
| Best use | Judging the method | Budgeting for costs and drawdowns |
Worked example: expectancy from 50 trades
Example (hypothetical): on a £5,000 account, a trader risks £50 per trade, so 1R is £50, or 1 percent of the account. Their journal holds 50 trades taken by one written rule set: 20 winners averaging 2.2R (£110) and 30 losers averaging 1R (£50).
The win rate is 20 out of 50, or 0.4, so the loss rate is 0.6. In R, expectancy is (0.4 x 2.2R) – (0.6 x 1R) = 0.88R – 0.60R = +0.28R per trade.
In pounds, it is (0.4 x £110) – (0.6 x £50) = £44 – £30 = +£14 per trade. As a check, the winners made 20 x £110 = £2,200 and the losers cost 30 x £50 = £1,500. The net £700 divided by 50 trades is £14, and £14 divided by the £50 risk is 0.28R, so both routes agree.
Then take off the costs
If those results were logged from chart prices, they leave out what trading actually costs. Suppose the spread, the odd slip past the stop and overnight funding average £5 per trade. That is 0.1R, so expectancy falls to £9, or +0.18R. Our guides to slippage in trading and overnight funding charges show where those pounds come from.
Then test how fragile it is
Now suppose one of the 20 winners was a 6R trade. Remove it and the other 19 winners total 38R against 30R of losses: 8R over 49 trades, or about +0.16R before costs. With costs of 0.1R, the edge almost disappears. One trade carried a large part of the result, which is the usual weakness of small samples.
How to measure positive expectancy in your own trading
The formula is easy; the honest data is the hard part. In our experience, this order keeps the number meaningful:
- Fix the rules first. Expectancy belongs to one rule set. Mixing trades from different setups gives an average of nothing in particular.
- Collect the sample. Use a rules-based test, as in how to backtest a trading strategy without fooling yourself, then a forward test.
- Record every trade in R. A trading journal for beginners with an R column makes the calculation a few cell formulas.
- Calculate after costs. Use the fills you actually got, not the prices you planned.
- Look for concentration. Remove the best one or two trades and recalculate. If the edge vanishes, you have a hope rather than a measurement.
- Keep measuring. Recalculate on a rolling block of trades; markets change and so can a method’s results.
How many trades is enough? There is no single number, and anyone quoting one as a rule is guessing. We want at least several dozen trades, and ideally 100 or more, before we trust the sign of the result, let alone its size. Turning an edge into a stake size is the subject of the Kelly criterion in trading.
What expectancy does not tell you
Expectancy is an average, and averages hide the path. Two methods with the same +0.2R can feel completely different: one wins often in small amounts, the other loses ten times in a row before one large winner.
That path decides whether you survive long enough for the average to show up. Losing streaks are normal even with a positive edge, and how deep they cut depends on risk per trade. Our article on the risk of ruin in trading covers the account-level odds, and how much to risk per trade covers the sizing side.
Expectancy also says nothing about the future on its own. It describes the trades you took in the conditions you took them. Leveraged products such as spread bets and CFDs can turn a small negative expectancy, or one bad gap, into losses well beyond the plan.
Common mistakes
- Judging by the win rate. A 70 percent win rate with an average loss of 1.3R and an average win of 0.5R works out at -0.04R per trade.
- Averaging pounds across different stakes. A few large trades dominate; use R.
- Leaving out costs. A thin edge before costs can be negative after them.
- Trusting ten trades. A short run tells you very little either way, as a losing streak and drawdown often proves.
- Changing rules mid-sample. Each change starts a new sample.
Frequently asked questions
What is a good expectancy in trading?
Any expectancy reliably above zero after costs, over a large sample, is a starting point. There is no published benchmark worth relying on; a small positive figure from 200 trades beats a large one from 15.
Can a low win rate have positive expectancy?
Yes. In the worked example, a 40 percent win rate produced +0.28R because the average winner was 2.2 times the average loser. The break-even point for that payoff is a win rate of about 31 percent.
Should I calculate expectancy in R or in pounds?
Use R to judge and compare methods, because it ignores stake size. Convert to pounds when you budget for costs and drawdowns.
Can positive expectancy still lose money?
Yes, over any short run. A positive average still allows long losing streaks, and costs, slippage or changing conditions can turn a thin edge negative.
Next steps
Take your last block of trades from one setup, convert each to R and run the formula, before and after costs. Then see how the result fits a full rules-based risk model and the wider process in how to develop your own trading system.
Building a rules-based risk model is part of our price action trading course, and the basics section is free to start.
This article is for education only and is not financial advice. Trading leveraged products such as CFDs and spread bets carries a high risk of losing money.
