Risk of ruin: how many losses in a row can an account take?
The Nuqta Markets team · 28 October 2026 · 7 min read · How we write
In short
- After n losses at r% each, the account is (1 − r)ⁿ of where it started.
- Streaks that feel rare are normal over 100 trades: at a 40% win rate, five losses in a row are close to certain.
- Risk per trade moves risk of ruin far more than small changes in win rate.
Risk of ruin is the probability that an account falls to a level from which the trader stops, before the rules have had time to work. It sounds abstract, but most of it comes down to two numbers anyone can control or measure: the risk per trade and the length of the losing streaks that the rules produce.
What a streak does at each risk level
With a fixed percentage r risked per trade, n losses in a row leave the account at (1 − r)ⁿ of where it started.
| Risk per trade | After 5 losses | After 10 losses | After 20 losses |
|---|---|---|---|
| 1% | 95.1% | 90.4% | 81.8% |
| 2% | 90.4% | 81.7% | 66.8% |
| 5% | 77.4% | 59.9% | 35.8% |
| 10% | 59.0% | 34.9% | 12.2% |
Illustrative calculation, not an expectation of results.
The way back is always steeper than the way down: a gain of d ÷ (1 − d) is needed to recover a drawdown d.
| Drawdown | Gain needed to recover |
|---|---|
| 10% | 11.1% |
| 20% | 25% |
| 30% | 42.9% |
| 50% | 100% |
How likely long streaks are
For independent trades, the chance of at least one losing streak of a given length in 100 trades is higher than most people guess. At a 40% win rate there is about a 98% chance of five or more losses in a row somewhere in those 100 trades, about 49% for eight and about 20% for ten. The maths behind win rates and streaks is in expectancy and win rate.
A hypothetical simulation
To see risk of ruin in one table, two hypothetical rule sets were simulated 100,000 times each over 200 trades. Ruin here means equity falling to 50% of the start at any point.
- Rules A: 45% win rate, wins of 1.5R, losses of 1R. Expectancy +0.125R before costs.
- Rules B: 40% win rate, wins of 1.4R, losses of 1R. Expectancy −0.04R before costs.
| Risk per trade | Rules A: chance of ruin | Rules B: chance of ruin |
|---|---|---|
| 1% | under 0.1% | under 0.1% |
| 2% | about 0.2% | about 12% |
| 3% | about 2% | about 37% |
| 5% | about 14% | about 70% |
| 10% | about 50% | about 93% |
Illustrative calculation, not an expectation of results. Hypothetical inputs, simulated; costs, gaps and changing market conditions are left out, and real results can be worse.
Three things stand out. First, the risk per trade moves the ruin probability far more than the difference between the two rule sets does at low risk. Second, a positive expectancy does not protect an account at 10% per trade: rules A halve it in about half of the runs. Third, the low figure for rules B at 1% is not good news: with negative expectancy the account still drifts down, roughly 8R over 200 trades, it just takes longer than 200 trades to reach the 50% line.
Using the numbers
A plan can start from the streak instead of the target. Decide the drawdown that would make you stop, look at the longest streak the backtest showed, add a margin for worse, and choose the risk per trade so that streak stays inside the limit. The lot size calculator then turns that percentage into a lot for each stop, and the risk-reward calculator checks that the wins in the plan are large enough for its win rate.
In short
A streak at r% leaves (1 − r)ⁿ; recovering from d needs d ÷ (1 − d); long streaks are normal over 100 trades; and risk per trade is the lever that moves ruin the most. For the other checks a position size should pass, see the position sizing checklist.
Questions people also ask
Does ruin mean losing the whole account?
Not necessarily. Many traders define ruin as the drawdown they would not continue past, such as 50%, because recovering from it needs a 100% gain. The definition changes the probability, so it should be stated.
Is a 10-trade losing streak a sign the strategy is broken?
Not by itself. At a 40% win rate there is roughly a 1 in 5 chance of at least one streak of 10 losses somewhere in 100 trades. Comparing the streak with the backtest's own streaks is more informative.
Can a positive expectancy still go broke?
Yes, if the risk per trade is large enough. In the hypothetical simulation in this article, the same positive-expectancy rules halved the account in about half of the runs at 10% per trade.
Take the quiz and find your trader profile.
Take the quizTrading financial markets carries a high level of risk and can result in the loss of all the money you invest. Nuqta Markets content is for education only and is not investment advice. Past results do not guarantee future results.