Stage 3: Execution and Risk · Module 8: Risk Management and Position Sizing · Lesson 6 of 9 · Back to course
An “edge” just means a method that, over many trades, makes more than it loses after costs. Expectancy is the number that tells you whether you might have one. It’s the average result per trade, measured in R.
This lesson shows you the formula, how to work it out from your journal, and – just as important – how to avoid fooling yourself with too few trades.
What you’ll learn
- The expectancy formula in plain English
- How to calculate it from your journal in a spreadsheet
- Why sample size matters so much
- What makes expectancy look better than it really is
- What to do with the number once you have it
The formula
Expectancy = (win rate × average win) − (loss rate × average loss)
Measure the wins and losses in R (see Lesson 8.5) and the answer is in R per trade. A positive number means that, in the trades you’ve recorded, the method made money on average. Zero or negative means it didn’t.
An even simpler way, if you log every trade in R: expectancy = total R ÷ number of trades.
A worked example
Illustrative journal: 60 trades, each risking 1R = £40, results after costs.
- Wins: 24 (40%), average +1.9R
- Losses: 36 (60%), average −0.95R (some losses were cut before the stop)
Expectancy = (0.40 × 1.9) − (0.60 × 0.95) = 0.76 − 0.57 = +0.19R per trade.
At £40 a trade, that’s about £7.60 per trade on average, and +11.4R (about £456) across the 60 trades.
That looks encouraging. The next section explains why you shouldn’t trust it yet.
Sample size: the part most people skip
Trading results are noisy. A handful of lucky or unlucky trades can swing your average a long way. A rough way to see how much is to work out the standard error of your average:
Standard error ≈ standard deviation of your R results ÷ √(number of trades)
In the example, the R results have a standard deviation of about 1.4R. With 60 trades:
- Standard error ≈ 1.4 ÷ √60 ≈ 0.18R
- A rough 95% range is the average ± 2 standard errors: 0.19 ± 0.36, or roughly −0.17R to +0.55R.
In plain English: 60 trades can’t tell you whether this method has an edge. The true figure could easily be negative.
With the same results over 200 trades, the standard error drops to about 0.1R, and the rough range becomes about −0.01R to +0.39R – better, but still not certain.
This is a simplified statistical rule of thumb. It assumes trades are independent and conditions don’t change, which isn’t fully true in markets.
In a spreadsheet
If column B holds your R results:
| What | Formula |
|---|---|
| Number of trades | =COUNT(B:B) |
| Expectancy (R per trade) | =AVERAGE(B:B) |
| Win rate | =COUNTIF(B:B,">0")/COUNT(B:B) |
| Standard error | =STDEV(B:B)/SQRT(COUNT(B:B)) |
| Rough low end | =AVERAGE(B:B)-2*STDEV(B:B)/SQRT(COUNT(B:B)) |
If the rough low end is below zero, treat your edge as unproven.
Things that make expectancy look better than it is
- Leaving out costs. Always record results after spread, commission and financing.
- Skipping “bad” trades in your journal, or excluding them as “not really my setup”.
- Demo fills. Demo accounts can fill more kindly than live markets, especially around news.
- Backtest tuning. If you adjusted the rules until the backtest looked good, you may have fitted the past rather than found an edge – see curve fitting.
- One market regime. Results from a strong trending month may not hold in a choppy one.
- Changing the rules midway. If the method changed, the earlier trades don’t belong in the same sample.
What to do with the number
If it’s negative: that’s useful information, not failure. Look at where it’s going wrong. Are losses bigger than 1R (stops moved or gaps)? Are wins smaller than planned (closing early)? Is one setup or one session dragging the average down? Break your journal down by setup and time of day.
If it’s positive but unproven: keep trading small and keep recording. Don’t increase risk just because the first 30 or 60 trades look good.
If it’s positive over a large sample: that’s the best evidence you can get, but it’s still evidence about the past. Keep monitoring it on a rolling basis, for example the last 50 or 100 trades, and decide in advance what figure would make you stop and review.
Expectancy also links straight back to risk. Even a real edge produces long losing streaks, which is why the risk sizes in Lesson 8.2 and the loss limits in Lesson 8.9 still matter.
Key takeaways
- Expectancy = (win rate × average win) − (loss rate × average loss), in R.
- Or simply total R ÷ number of trades.
- Small samples are very noisy. 60 trades usually can’t prove an edge.
- Record results after costs and include every trade.
- A positive past expectancy is evidence, not a guarantee. Keep monitoring it.
Quick check
- Win rate 50%, average win 1.5R, average loss 1R. What’s the expectancy?
- You’ve logged 20 trades with +0.4R expectancy. Should you double your risk?
- Name two things that can make expectancy look better than it really is.
Answers: 1) 0.75 − 0.5 = +0.25R. 2) No – 20 trades is far too few to tell. 3) Any two of: leaving out costs, skipping trades, demo fills, backtest tuning, one market regime, changing the rules.
FAQs
How many trades do I need? There’s no magic number, but 100 is a sensible minimum to start forming a view, and more is better. Use the standard error check rather than a fixed count.
Can I combine demo and live results? Keep them separate. Live results include real fills, real costs and real emotions, so they’re the ones that matter most.
Is expectancy the same as profit factor? They’re related. Profit factor is total wins ÷ total losses (above 1 means profitable). Expectancy tells you the average per trade, which is easier to link to risk per trade.
Disclaimer: This page is general education, not financial or investment advice or a personal recommendation, and it doesn’t take your circumstances into account. Examples are hypothetical, for illustration only, and not an invitation to buy or sell. Learning a strategy doesn’t guarantee profits. Past performance is not a reliable guide to future results. Futures, spread bets and CFDs are complex and carry a high risk of losing money quickly because of leverage, and most retail traders lose money. With a real futures account, losses can exceed your deposit. Only trade with money you can afford to lose, and speak to an independent FCA-authorised adviser if you’re unsure. Read our full risk warning.