Trading Is a Game of Probability, Not Certainty
Every trade has a range of possible outcomes. The trader’s job is not to know exactly what happens next. Instead, the job is to build a repeatable decision process with a favourable distribution of outcomes.
Probability describes what is likely to happen over many trades. It does not guarantee the outcome of the next trade.
Win rate
The percentage of trades that finish with a profit.
Average reward
The average size of a winning trade, measured in cash, points or R.
Average loss
The average size of a losing trade, including real execution costs.
What Is Trading Expectancy?
Expectancy estimates the average amount a system makes or loses per trade. It combines the probability of winning with the size of wins and losses.
The formula
E = (P(win) × Average win) − (P(loss) × Average loss)
A simple expectancy calculation
Assume a system wins 60% of its trades. The average winning trade is £200 and the average losing trade is £100.
Worked result
Over a large enough sample, the system has an estimated expectancy of £80 per trade before fees, slippage and changes in market behaviour.
Win Rate Alone Does Not Define a Winning System
| System | Win rate | Reward-to-risk | Expectancy |
|---|---|---|---|
| System A | 80% | 1:1 | +0.60R |
| System B | 40% | 3:1 | +0.80R |
System B wins less often but has the stronger expectancy. Therefore, traders should stop treating win rate as the main scorecard.
The three-part model
A robust system needs positive expectancy, controlled variance and repeatable execution. Remove any one of those and the strategy becomes difficult or impossible to trade consistently.
Variance, Losing Streaks and Drawdown
Positive expectancy does not produce a smooth stream of profits. Randomness creates clusters of wins and losses, even when the underlying edge remains intact.
Variance measures how far individual outcomes move away from the average. High-variance strategies can be profitable over time while still producing severe short-term drawdowns.
Why losing streaks matter
A strategy with a 60% win rate will still experience several consecutive losses. Therefore, risk must be sized for the losing sequence that will eventually occur, not for the average trade.
Drawdown is not proof that the edge is gone
However, drawdown must stay within the range implied by testing. A drawdown that is materially deeper or longer than expected may signal a regime change, execution problem or model failure.
The Kelly Criterion and Position Sizing
The Kelly Criterion connects edge with capital allocation. In simplified form, it estimates the fraction of capital that maximises long-term growth.
For a system with a 60% win rate and a 1:1 reward-to-risk ratio, full Kelly suggests 20% of capital. That is mathematically aggressive and psychologically unrealistic for most traders.
Consequently, professional traders often use half-Kelly, quarter-Kelly or a much smaller fixed-fractional risk limit.
Positive Expectancy Is Not Enough Without Survival
A profitable method can still destroy an account when risk per trade is too large. Risk of ruin rises sharply when leverage increases, drawdowns cluster or the true expectancy is weaker than the backtest suggests.
- Reduce the percentage of capital risked per trade.
- Model realistic losing streaks and slippage.
- Use Monte Carlo analysis to randomise trade order.
- Set maximum drawdown and stop-trading rules.
- Do not scale until live behaviour confirms the model.
A Renko Futures Expectancy Example
Consider an illustrative ES futures strategy with a 55% win rate, a $500 average win and a $300 average loss.
Expectancy calculation
Ten trades per week would imply a theoretical annual expectancy of $72,800 per contract before fees, slippage, missed trades and edge decay. This is an illustration, not a forecast.
A Practical Probability-Based Trading Framework
- Define precise entry, exit and risk rules.
- Collect a meaningful sample of historical and forward trades.
- Calculate win rate, average win, average loss and expectancy.
- Measure variance, drawdown and losing streaks.
- Run Monte Carlo simulations.
- Select a conservative position-sizing method.
- Forward test at micro size.
- Scale only when live behaviour remains consistent with the model.
Think in distributions, not individual outcomes
The professional trader does not need the next trade to win. The professional needs the process to maintain positive expectancy across a large sample.
That changes everything. It reduces emotional attachment to individual trades and forces the trader to focus on edge, execution and survival.
I do not need to win every trade. I need to execute my edge every time.
Put the Maths Into Practice with the $200 Renko Trading Starter Pack
Build your futures trading platform foundation through Stephen Hoad’s tracked NinjaTrader route and claim the Renko Trading Starter Pack. No funding or live trade is required for this bonus route.
Further official risk education
For additional position-sizing context, see CME Group’s 2% risk rule guide.
Trading expectancy: frequently asked questions
What is a good trading expectancy?
Any reliably positive expectancy after costs is potentially useful. Its quality also depends on drawdown, variance, trade frequency and robustness.
Can a 40% win-rate strategy be profitable?
Yes. It can be highly profitable when average winners are sufficiently larger than average losers.
How many trades are needed to measure expectancy?
More data is generally better. One hundred trades is a useful starting point, but different regimes and instruments should also be tested.
Does positive expectancy guarantee profit?
No. The edge can change, execution can differ from testing and excessive position size can create ruin before the expectancy emerges.
English