Drawdown and expectancy

Expectancy is the average amount a trader can expect to win or lose per trade, over a large number of trades, given the strategy's win rate and the size of average wins versus average losses. The formula is: expectancy = (win rate × average win) − (loss rate × average loss). A strategy that wins 40% of the time with an average win of $300 and an average loss of $150 has an expectancy of (0.40 × $300) − (0.60 × $150) = $120 − $90 = $30 per trade. Over 100 trades, this strategy would be expected to net approximately $3,000, before costs, even though it loses more often than it wins.

Expectancy is more informative than win rate alone because win rate says nothing about the size of wins relative to losses, and it is more informative than average profit per winning trade alone because it doesn't account for how often losses occur. A trader fixated purely on win rate might reject a genuinely excellent 35%-win-rate, high-reward strategy in favour of an 80%-win-rate strategy with tiny wins and occasional catastrophic losses that has a far worse, even negative, expectancy. Expectancy forces the trader to look at the whole picture in one number.

It's important that expectancy be calculated from a large enough and honest enough sample to be meaningful — ideally 100+ trades, using real (or carefully simulated) results including costs, not a curated subset of 'good' trades. A trader who calculates expectancy from their best 20 trades of the year is not measuring their strategy's expectancy; they are measuring survivorship bias. Expectancy should also be recalculated periodically as new trades accumulate, since a strategy's true win rate and average win/loss sizes only become reliably estimable with a sufficient sample size, and market conditions can shift the numbers over time.

One of the hardest lessons for newer traders to internalise is that a genuinely profitable, positive-expectancy strategy will still produce losing streaks and drawdowns, sometimes severe ones, purely from ordinary statistical variance — not because the strategy has stopped working. This is because expectancy describes an average outcome over a large number of trades, and any individual sequence of trades is subject to the randomness inherent in probability, the same way a fair coin can land on the same side eight times in a row without the coin being unfair.

The probability of a losing streak of a given length can be estimated (for independent trades) as the loss rate raised to the power of the streak length. For a strategy with a 40% win rate (60% loss rate), the probability of five consecutive losses is 0.6^5 = 0.0778, or about 7.8% — meaning across a year of active trading, a trader taking several hundred trades should expect to see five-in-a-row losing streaks multiple times, not as a rare anomaly but as a routine feature of the strategy's normal behaviour. Streaks of seven or eight losses (0.6^7 ≈ 2.8%, 0.6^8 ≈ 1.7%) become statistically rare but still entirely plausible over a long enough trading history.

This matters enormously for psychology and for account survival rules, because a trader who does not expect these streaks will interpret a normal, statistically predictable losing run as evidence the strategy is broken, and may abandon a genuinely good system at exactly the point it is due to mean-revert, or may panic-size into revenge trades that turn a normal drawdown into an account-ending one. Understanding the expected frequency and depth of drawdowns in advance, through this kind of streak analysis, converts a terrifying surprise into an anticipated, planned-for event.

The mathematical relationship between a percentage loss and the percentage gain required to recover it is asymmetric and non-linear, and this asymmetry is one of the most underappreciated dangers in trading. If an account loses 10%, it needs a gain of 11.1% (10 ÷ 90) on the remaining capital to return to breakeven. If it loses 20%, it needs 25% (20 ÷ 80) to recover. If it loses 50%, it needs a 100% gain (50 ÷ 50) just to get back to where it started. If it loses 80%, it needs a 400% gain. The formula is: recovery percentage required = loss percentage ÷ (1 − loss percentage).

This asymmetry accelerates dramatically as losses deepen, which is precisely why risk management exists to prevent deep drawdowns in the first place rather than relying on 'trading your way out' of them. A trader who has lost 50% of their account cannot simply resume their normal strategy and expect to be whole again in a similar timeframe to how they lost it; they need to generate returns twice as large, in percentage terms, as the loss that put them there, which for most realistic strategies takes considerably longer than the drawdown itself took to occur, and often tempts traders into taking larger, riskier positions specifically to 'catch up' — a behaviour that frequently deepens the hole further.

This is the core mathematical justification for keeping risk per trade small and enforcing maximum drawdown limits: the goal is never to have to face this recovery asymmetry at severe levels in the first place. A trading system that limits maximum realistic drawdown to 10-15% keeps the required recovery gain in a very manageable 11-18% range; a system that allows drawdowns to reach 50% or more is exposing the account to a mathematical recovery challenge that many trading strategies, even good ones, will struggle to overcome within a reasonable time, if ever.

Not every drawdown has the same cause, and a mature trader learns to distinguish between drawdowns produced by normal statistical variance in a sound, unchanged strategy, and drawdowns produced by a breakdown in the trader's own process — rule violations, sizing errors, trading a strategy outside its tested conditions, or emotional decision-making. These require completely different responses: a variance-driven drawdown within the expected range calls for patience and continued disciplined execution, while a process-failure drawdown calls for an immediate halt and review.

A useful diagnostic is to review, trade by trade, whether each losing trade during the drawdown was executed exactly according to the strategy's rules (correct setup, correct sizing, correct stop placement, no impulsive entries or exits) or whether rules were bent or broken. If the losses occurred within a well-executed, rule-following sequence, and the loss frequency and depth are consistent with what streak probability analysis predicts for the strategy's known win rate, this is very likely normal variance. If, on review, several of the losing trades involved oversized positions, moved stops, impulsive entries outside the strategy's criteria, or trading during a state of emotional agitation, the drawdown is at least partly a process failure, and continuing to trade without addressing the process issue risks compounding the damage.

This distinction should be made through a written trade-by-trade review, not through a gut feeling in the moment, because the emotional state induced by a drawdown itself tends to distort self-assessment — traders in a drawdown often either excuse their own rule violations ('the market was unusual') or, in the opposite direction, catastrophise a perfectly normal variance-driven drawdown as proof the whole approach is wrong. A structured review process, ideally involving a trading journal with entries logged before the emotional outcome was known, is the most reliable way to make this call honestly.

Because drawdowns are mathematically inevitable for any active trading strategy, a serious trader builds a pre-committed response plan for what happens at defined drawdown thresholds, decided in advance while calm, rather than improvised during the stress of an actual drawdown. A typical structure might specify: at a 5% drawdown from the equity high-water mark, reduce risk per trade by half and increase the scrutiny of trade selection; at an 8% drawdown, pause new trades entirely and conduct a full process review comparing recent trades against the strategy's rules; and at a 10% drawdown (or whatever threshold a funded account's own maximum loss limit requires a buffer against), stop trading entirely until a formal reassessment is complete.

This kind of tiered response is especially critical for traders operating under funded account programmes, which typically impose hard maximum drawdown limits that result in disqualification if breached. A trader's personal drawdown response thresholds should sit meaningfully inside the programme's hard limits, leaving room to respond and de-risk before an automatic breach occurs — waiting until the programme's own limit is nearly reached before taking any protective action leaves no margin for error and effectively guarantees that any further adverse variance ends the account.

The plan should also specify what re-entry looks like: what conditions must be met (a certain number of paper or minimal-size trades executed flawlessly, a specific review completed, a specific rest period) before risk is restored to normal levels. Without an explicit re-entry condition, traders in a drawdown often either return to full risk prematurely, right when confidence is lowest and errors are most likely, or remain overly cautious indefinitely, missing legitimate opportunities during a strategy's normal recovery phase.