Expectancy: how to tell whether a strategy is actually working
One number tells you what an average trade is worth. How to calculate expectancy from your own results, and why win rate on its own tells you nothing.
Expectancy is what your average trade is worth. It combines how often you win with how much you win and lose, and it produces one number: the amount you can expect to make or lose per trade over a large sample. If that number is positive, the strategy is worth trading. If it is negative, no amount of discipline, position sizing, or psychology will save it.
This is the only performance question that actually matters, and it is the one most traders never calculate.
The formula
Expectancy = (win rate x average win) minus (loss rate x average loss)
That is it. Four inputs, all of which you already have if you keep records.
Worked example. Out of 100 trades you won 40 and lost 60. Your average win was $300 and your average loss was $100.
- Winning side: 0.40 x $300 = $120
- Losing side: 0.60 x $100 = $60
- Expectancy: $120 minus $60 = $60 per trade
So across 100 trades, this strategy is worth roughly $6,000, despite losing 60 percent of the time. Every individual trade is more likely to lose than win, and the strategy still makes money. Both of those statements are true at once, and holding them together is most of what separates a professional from an enthusiast.
Why is win rate alone a vanity metric?
Two strategies, both with results that sound completely different:
| Strategy A | Strategy B | |
|---|---|---|
| Win rate | 70% | 35% |
| Average win | $100 | $400 |
| Average loss | $250 | $100 |
| Expectancy | minus $5 | plus $75 |
Strategy A wins seven times out of ten and loses money. Strategy B loses nearly two trades out of three and is excellent. If you judged either one by its win rate you would keep exactly the wrong one.
Strategy A is not a made-up example either. It is the natural result of taking profits early because they feel good and letting losses run because closing them makes the loss real. That behaviour produces a high win rate and a negative expectancy, and it is probably the most common way that otherwise sensible people lose money.
Expectancy in R, which is more useful
Dollar figures make it hard to compare periods, because your account size changes. So express everything in R, where 1R is the amount you risked on the trade.
A trade that hits a target three times your risk is +3R. A trade that hits your stop is minus 1R. A trade you closed early for half your risk is +0.5R.
Now expectancy becomes a clean number like “0.4R per trade”, which means the average trade returns 40 percent of what you risked. That number is comparable across accounts, across instruments, and across years. It is also the number that lets you answer a genuinely important question: how much do I need to trade to make this worthwhile?
If your expectancy is 0.4R and you risk one percent per trade, each trade is worth 0.4 percent of your account on average. A hundred trades a year is 40 percent before compounding, in theory, assuming nothing changes and you follow the rules exactly. Both of those assumptions are heroic, which is why the number is a planning tool rather than a promise.
The sample size problem
Expectancy calculated on twenty trades is close to meaningless. Random variation dominates. You need enough trades that one or two outliers cannot swing the whole picture, and in practice that means at least a hundred before you take the number seriously, preferably more.
This is inconvenient for a part-time trader taking two or three trades a week, because a hundred trades is most of a year. There are two honest responses. Collect the data anyway and be patient, or generate a larger sample from historical data, which is what backtesting is for.
What is not an honest response is declaring a strategy good after twelve trades because it felt good.
Where expectancy quietly leaks
The number you calculate from your live results is almost always worse than the one from your testing. The gap is where the real lessons are:
Costs. Spread, commission, and slippage come straight off the top of every trade. On small targets they are a large percentage of the edge.
Early exits. Closing winners before target lowers the average win. This single habit destroys more expectancy than any other, because it attacks the larger of the two terms in the formula.
Widened stops. Moving a stop to avoid a loss raises the average loss. Same damage, other side of the equation.
Trades that were not in the plan. Setups taken out of boredom or frustration have their own, much worse, expectancy. They get mixed into your results and drag the average down while looking like part of the same strategy.
The fix for all four is the same: tag every trade in your journal as either “in the plan” or “not in the plan”, then calculate expectancy separately for each group. The first time you do this is usually uncomfortable, and it is the single most valuable hour you will spend on your own trading. The columns to log are in why your trading journal beats your next strategy.
What does a realistic expectancy look like?
I will not give you a target to aim for, because a number floating free of context is exactly the kind of thing this site exists to avoid, and because anyone quoting an impressive expectancy without a verified track record is selling something.
What I will say is that positive and modest is the goal. An expectancy of 0.2R to 0.4R, held consistently across a large sample and traded with sensible size, is a real edge. People do not usually fail because their expectancy is 0.3R instead of 0.9R. They fail because it is negative and they never checked, or because it was positive and they abandoned it during a drawdown that the number had already told them to expect.
How this connects to everything else
Expectancy sits underneath every other decision:
- Your risk to reward ratio sets the size of the average win relative to the average loss. See the risk to reward ratio.
- Your stop placement determines what 1R actually is. See where to place a stop loss.
- Your position sizing turns R into money without changing the underlying edge. See position sizing for small accounts.
- Your drawdowns are the natural variance around a positive expectancy. See what a normal losing streak looks like.
Change any one of those and the expectancy moves. That is the whole system in one sentence.
Calculate yours this week, even if the sample is small and the answer is unflattering. A negative number found early is cheap. Found late, it is expensive.
Key takeaways
- Expectancy is what the average trade is worth: win rate times average win, minus loss rate times average loss.
- It is the only performance question that matters. Positive means the strategy is worth trading, negative means no amount of discipline will save it.
- Win rate alone is a vanity metric. A 70 percent strategy can lose money while a 35 percent one makes it.
- Express results in R, where 1R is what you risked, so the number is comparable across accounts and years.
- Calculate it separately for trades that followed the plan and trades that did not. The gap is usually where the real problem is.
The full method, with the review process I run against these figures, is in The 9-to-5 Trader.
Educational only, not financial advice. Trading carries a real risk of loss.
The whole method, in one place
These notes are pieces of the system in The 9-to-5 Trader. Start with the free tools, or read the book.