Losing Streak Probability
The exact probability of hitting a run of N consecutive losses, given your win rate.
| K | Probability | 1 in |
|---|---|---|
| 1 | 100.00% | 1.0 |
| 2 | 100.00% | 1.0 |
| 3 | 100.00% | 1.0 |
| 4 | 99.93% | 1.0 |
| 5 | 96.59% | 1.0 |
| 6 | 80.09% | 1.2 |
| 7 | 54.37% | 1.8 |
| 8 | 31.97% | 3.1 |
| 9 | 17.32% | 5.8 |
| 10 | 8.99% | 11.1 |
| 11 | 4.57% | 21.9 |
| 12 | 2.30% | 43.6 |
| 13 | 1.15% | 87.1 |
| 14 | 0.57% | 174.7 |
| 15 | 0.28% | 350.9 |
| 16 | 0.14% | 705.1 |
| 17 | 0.0706% | > 1000 |
| 18 | 0.0351% | > 1000 |
| 19 | 0.0175% | > 1000 |
| 20 | 0.0087% | > 1000 |
| 21 | 0.0043% | > 1000 |
| 22 | 0.0021% | > 1000 |
| 23 | 0.0011% | > 1000 |
| 24 | 0.0005% | > 1000 |
| 25 | 0.0003% | > 1000 |
Normal losing streaks are longer than you think
A 50% win-rate strategy over 500 trades has about a 50% chance of hitting a run of 9 or more losses, and a 10% chance of hitting 13 or more. Both are statistically normal, not evidence that something has broken.
This is why most working traders cap position size well below Kelly: the long streaks that happen just from the laws of large numbers can wipe out an over-sized account before the edge has time to show.
Common questions
What is the probability of a long losing streak?
For N independent trades with a win rate p, the probability of hitting a run of at least K consecutive losses is roughly 1 - exp(-N(1-p) * (1-p)^(K-1)). This calculator computes it exactly via a Markov chain, so there is no small-N approximation error. A 50% win-rate strategy over 500 trades has about a 50% chance of hitting a streak of 9 or more losses; most traders have not internalised numbers this large.
What is the expected longest losing streak?
Approximately log(N(1-p)) / log(1/(1-p)), where N is total trades and p is the win rate. A 55% win-rate system over 1,000 trades has an expected longest losing streak of around 8. That is the baseline you should plan around, not the worst streak you have personally observed.
Does a long losing streak mean my edge is broken?
Not necessarily. The streaks that will happen just from the laws of large numbers are far longer than most traders expect, and a 10-loss run in a 55% win-rate system is statistically normal, not evidence of a dead edge. The honest test is whether your realised average trade still sits inside the confidence interval your backtest predicted - which is what the Monte Carlo Simulator checks.
Are trading losses independent?
Approximately, for strategies that take unrelated setups across time. But when positions overlap - correlated FX pairs taken simultaneously, or an indices portfolio hit by one macro event - the effective independence is much lower and the clustered losing days are worse than this calculator shows. If your trades are correlated, treat the numbers here as a floor, not a ceiling.
How does this relate to risk of ruin?
Directly. A long losing streak at a given risk-per-trade is what blows the account. The Risk of Ruin Calculator asks the dollar question - what is the probability I lose X% of my capital. This page asks the sequence question that drives it: how many losses in a row should I plan for.
Where we use this
Every published strategy on this site discloses its worst observed losing streak and the Monte Carlo distribution it was drawn from.
Get the rulebooks: exact rules, win rate, worst streak and net expectancy for every strategy we've tested. Free by email.