Performance and statistics

Trading Expectancy Calculator

Expectancy is an estimate, not a measurement. Enter your numbers and the sample they came from, and the page shows both the figure and how much it is allowed to be wrong.

Your numbers

Money or R — just keep both fields in the same unit.
Enter it as a positive number.

How sure you can be

How many trades those averages came from. This is what turns a single number into an honest range.
Expectancy per trade
Over 100 trades
Win / loss size ratio
Break-even win rate
True win rate, 95% rangeenter sample size
Expectancy, 95% rangeenter sample size

The range uses the standard error of a proportion, so it reflects uncertainty in the win rate only — the two averages carry their own error on top. Treat it as a floor on how unsure you are, not a ceiling.

FreeNo signupNothing leaves your browser

The formula, and what it is really saying

expectancy = (win rate × average win) − (loss rate × average loss)

The output is the amount an average trade is worth. Positive means the method pays; negative means it costs, no matter how good the winners feel.

It also gives the break-even win rate for your particular win and loss sizes:

break-even win rate = average loss ÷ (average win + average loss) × 100

If your average win is 320 and your average loss is 200, break-even sits at 38.5%. Below that the method loses money however disciplined you are, and no amount of screen time changes the arithmetic.

The part almost every calculator leaves out

Expectancy is computed from a sample, so it is an estimate, and estimates have error bars. Nearly every tool that returns this number returns it as if it were measured.

The uncertainty in the win rate follows the standard error of a proportion:

SE = √( p(1 − p) / n )          95% range ≈ p ± 1.96 × SE

Put a real case through it. Forty-five percent win rate, average win 320, average loss 200 — expectancy is +34 per trade, which looks like a business. Now say those numbers came from 30 trades. The 95% range on the win rate is roughly 27% to 63%, which puts expectancy somewhere between −59 and +127. The honest reading is not "I make 34 a trade". It is "I cannot yet tell whether I make money."

At 200 trades the same win rate narrows to about 38%–52%, and expectancy to roughly −2 to +70. Still wide, and still worth knowing before you scale size.

Trades 95% range on a 45% win rate
30 27% – 63%
60 32% – 58%
100 35% – 55%
200 38% – 52%
500 41% – 49%

The range narrows with the square root of the sample, which is why it shrinks so slowly: to halve the uncertainty you need four times the trades. There is no shortcut, and a confident number produced from a short sample is not confidence, it is arithmetic performed on noise.

Two levers, and only two

When expectancy is negative there are exactly two things that can move it, and naming them prevents a lot of wasted effort:

  1. Win more often — better selection, fewer marginal setups, avoiding the sessions or instruments where your record is worst.
  2. Change the sizes — bigger winners relative to losers, which usually means holding to plan rather than closing early, and cutting at the stop rather than past it.

Everything else people try — new indicators, a different platform, trading more — acts on one of these two or on neither. Most acts on neither. If a change cannot be described as "this raises my win rate" or "this improves my win-to-loss size ratio", it is not a fix, and the calculator will say so when the numbers come back the same.

Why the averages hide as much as they show

Two accounts can produce identical expectancy and behave nothing alike. One earns +30 per trade steadily; the other earns it through a single enormous win among many small losses. The average is the same, the survivability is not — the second account spends most of its life in drawdown, and drawdown is where rules get abandoned.

So read expectancy with the distribution beside it, not instead of it. Grading trades in R-multiples makes the shape visible: a healthy record has several winners above 2R, not one above 8R. And the depth of the drawdowns the shape produces is what decides whether the expectancy ever gets collected, because recovery is not symmetrical.

FAQ

What is expectancy in trading?

The average amount a single trade is expected to return, calculated as win rate times average win minus loss rate times average loss. It answers the only question that matters about a method — does repeating it make money — and it is the reason a 30% win rate can be excellent and a 70% win rate can be ruinous.

How many trades do I need before my expectancy means anything?

More than most people assume. At 30 trades the 95% range on a 45% win rate spans roughly 27% to 63%, wide enough that expectancy can be positive or negative. Around 100 trades the estimate becomes usable for decisions, and around 200 it becomes reasonably stable. Because the range narrows with the square root of the sample, quadrupling the trades only halves the uncertainty.

Should I calculate expectancy in money or in R?

In R if you want to judge the method, in money if you want to judge the account. R removes position-size differences, so it shows whether the setups are good; money includes them, so it shows what actually happened to your balance. When the two disagree, the sizing is inconsistent, and that is worth knowing on its own.

Can expectancy be positive and the account still lose money?

Yes, in three ways. The sample may be too small for the positive figure to be real. Position sizing may be inconsistent, so the losses land on larger positions than the wins. Or a drawdown may end the account before the average has time to assert itself — which is the specific risk on prop firm accounts, where a hard limit closes the account regardless of what the long-run expectation says.

Does expectancy account for costs?

Only if your average win and average loss already include commission, spread and swap. Compute them from your closed trades net of costs, not from your plan. On short-hold strategies the difference is large enough to move expectancy across zero, which is why a method that looks profitable on paper can lose money in the account without anything else going wrong.

This is the plan. What did you actually do?

A calculator tells you the size you should have taken. It cannot tell you the size you took at 2pm after two losers, or how often your stop moved once price went against you. Drop in a statement from MT4/MT5, a broker CSV or a crypto export and see the answer for your own last 90 trades.

Import my trades — free