Free tool
Losing streak calculator.
Five losses in a row feels like a broken strategy. At a 50% win rate it happens somewhere in 81% of samples of 100 trades. Put in your own numbers and see which streaks are normal, and how much of the account the one to plan for would take.
What are the odds of this streak?
The chance of at least one run that long, anywhere in the sample.
What streak should I plan for?
The run with no more than a 5% chance of being beaten, and what it costs.
Reference
Losing streaks over 200 trades. The run to plan for has no more than a 5% chance of being beaten. The drawdown is that run at 1% risk per trade.
| Win rate | Average longest run | Run to plan for | Drawdown at 1% risk |
|---|---|---|---|
| 35% | 10.7 | 16 | 14.9% |
| 40% | 9.2 | 14 | 13.1% |
| 45% | 8.0 | 12 | 11.4% |
| 50% | 7.0 | 10 | 9.6% |
| 55% | 6.1 | 9 | 8.6% |
| 60% | 5.4 | 8 | 7.7% |
| 65% | 4.7 | 7 | 6.8% |
The method
How the numbers are computed.
The odds are exact, not simulated. After each trade the current losing run either grows by one or resets to zero, and the calculator follows the probability of every run length across the whole sample. This is the recursion for the longest run in a sequence of trials described by Mark Schilling in 1990, and it produces the same figures as the tables inhow many losing trades in a row is normal.
It assumes your trades are independent and that your win rate is the true one. Neither is quite true: a win rate from a backtest is an estimate, and you can check how loose it is with the sample size calculator. If a prop firm challenge is the reason you are asking, size your risk from the daily and maximum loss limits before the first trade.
FAQ
Frequently asked questions.
Exactly, with a recursion over the length of the current losing run: after each trade, the run either grows by one loss or resets on a win, and the calculation tracks the chance of every run length until one reaches the streak you asked about. It assumes each trade is independent. The common shortcut, multiplying the number of starting points by the chance of k losses, overstates the odds because overlapping stretches share trades.
The shortest losing run that has no more than a 5% chance of being beaten in your sample. The average longest run is exceeded about as often as not, so it is the wrong number to size risk against. At a 45% win rate over 200 trades, the average longest run is about 8, and the run to plan for is 12.
At a fixed percentage of the current balance, each loss takes that percentage of what is left, so k losses at r% remove 1 minus (1 minus r) to the power k. Twelve losses at 1% take 11.4% of the account; at 0.5% they take 5.8%. Open positions measured on equity can add to that before a trade closes.
Not on its own. If the calculator gives a streak that long more than a 5% chance in your sample, it is ordinary variance. Below 5%, read when a losing streak is a warning sign before you change anything.
See the streaks your own strategy produces. On paid plans, BacktestFlow's Stability section tracks streaks, drawdown and recovery from the trades you replay, and Monte Carlo resamples your own R multiples into 1,000 equity paths.
Free planNo credit cardRuns in your browser