A 68% win rate can still hide a drawdown large enough to threaten an account. Review the size and sequence of losses, the risk per trade, and the maximum drawdown before treating any win-rate headline as evidence of a usable strategy.
At 4:18 p.m. on a Friday, Eli was sitting in his parked car outside a grocery store in Chicago, one hand on a paper receipt and the other refreshing a spreadsheet on his phone. The strategy summary looked reassuring: 68% wins across its test. He had spent the week telling himself that two bad trades were simply the price of a system that won more often than it lost.
Then he filtered the results by equity curve.
Near the middle of the test, nine losses appeared close together. The individual entries had looked manageable when he first reviewed them. Taken as a run, they pushed the account from its prior high into a decline he had not noticed in the headline numbers. If a similar sequence happened with real money and his intended position size, the account could lose enough that his next decisions would be driven by recovery pressure rather than the plan.
The bad ending was no longer an abstract red number. Eli could keep trusting the percentage, increase size after a few winners, and discover the drawdown only when it arrived in his live account.
He closed the spreadsheet and went back to the trade list.
A win rate leaves out the shape of the losses
Win rate answers one narrow question: how often did this set of trades close positive? It does not show how much each winner made, how much each loser cost, or how losses clustered.
A strategy can win frequently with small gains and give back far more on occasional losses. It can also produce a tolerable average result while forcing a trader through a drawdown they cannot realistically hold. The headline percentage does not tell you which version you are looking at.
Maximum drawdown measures the largest decline from a prior account peak to a later low. It gives the losing stretch a shape. That matters because traders experience strategy risk as a sequence, not as a tidy average at the end of a backtest.
For an illustration, a trader risking 2% per position faces a very different account path from one risking 0.5%, even if both use the same entries and report the same win rate. The first trader has less room for normal variation. A short losing run can turn a review process into a decision made under stress.
A better review begins with questions the headline cannot answer:
- What was the largest peak-to-trough decline?
- How many losses occurred in the worst stretch?
- Was position size fixed, or did it change after wins and losses?
- Did one oversized loss erase several ordinary winners?
- Would the drawdown still be tolerable after slippage, fees, or a missed exit?
Those questions move the review from “How often did it win?” to “Could I survive the way it loses?”
The buried drawdown needs its own review
Eli rebuilt his spreadsheet around the painful section rather than the favorable summary. He marked each trade’s setup, market context, position size, planned risk, actual outcome, and any deviation from the original rule. He added a running equity line so the drawdown could not disappear inside a column of percentages.
The pattern was clearer by the time the grocery store lights came on. The nine losses were not random evidence that the method was broken. Several appeared during the same market condition, and two losses were larger because the original risk limit had not been applied consistently in the test.
That distinction matters. A drawdown can come from ordinary strategy variation, poor sizing, loose exits, or rules that only look clear after the chart has finished forming. Each calls for a different response. Reducing size may help with survivability. Rewriting an undefined exit rule may help with consistency. Neither response should be based on a win rate alone.
This is also where backtests need restraint. A clean historical result can understate the friction of live execution. Slippage can change an exit. A small sample can make a short favorable stretch look reliable. Repeatedly tuning rules to fit old data can produce an answer tailored to history rather than a plan that holds up on unseen data. This review of backtesting overfitting explores why that gap deserves attention.
Survivability sets the position size
Eli’s next step was deliberately less exciting. He chose a maximum loss per trade he could see plainly, set a daily loss limit, and treated consecutive losses as information to review rather than a cue to press harder.
Position size connects the strategy to the account. Before approving a trade, calculate the distance between entry and stop, then choose a quantity that keeps the planned dollar risk within the limit. If the quantity is too small to make the trade practical, pass on the trade. The setup does not become safer because the trader wants it to fit.
The same discipline applies after a loss. A queued signal can look compelling when the previous trade was frustrating, especially late in the day or after a strategy has spent weeks producing more winners than losers. That is when a visible approval step earns its place. Pause, check the current exposure and loss limits, and decide whether the next trade still fits the rules. A fourth alert after three losses deserves a different review from an isolated setup on a quiet morning.
Build a record that can challenge the headline
By the following Friday, Eli still kept win rate in his journal. It had a place, alongside average win, average loss, maximum drawdown, largest single loss, consecutive losses, and rule deviations.
The difference was the order of attention. He started each review with the period that made the strategy hardest to hold, then checked whether his position size and approval rules could contain it. The 68% figure remained on the page. It no longer got to make the decision by itself.
Educational content, not financial advice.
Comments
No comments yet.