How to use the risk of ruin calculator
The calculator has three tools, from the simplest question to the broadest.
Losing streaks gives the odds of a run of consecutive losses. Enter your win rate, the number of trades you plan (a year of trading, for example) and the streak length you want to test.
Drawdown from a streak shows what such a run does to the account. Enter the percentage of the current balance you risk per trade, the number of losses in a row and, optionally, your account balance.
Risk of ruin puts it all together:
- Enter the win rate and the payoff ratio (average win ÷ average loss), from your journal or backtest.
- Enter the risk per trade as a percentage of the current balance.
- Set the ruin level: the drawdown at which you would stop trading the strategy.
- Enter the number of trades to simulate and choose whether the drawdown counts from the starting balance or from the highest balance reached.
How it is calculated
Streak probabilities are exact. If An is the probability of no run of k losses in n trades:
Expected longest streak = Σk≥1 P(longest ≥ k)p is the win rate and q = 1 − p the loss rate. The subtracted term is the chance that the first run of k losses ends exactly at trade n: a win, then k losses, with no earlier run.
For the risk of ruin, each simulated trade multiplies the balance by (1 + b × r) on a win or (1 − r) on a loss, where b is the payoff ratio. An account is ruined when its drawdown reaches the ruin level, and it stops trading. The calculator runs 5,000 accounts with a fixed random seed, so the answer is the same every time you load the page.
Risk of ruin (unlimited trades) ≈ (1 − D)RD is the ruin level. R is the adjustment coefficient from ruin theory, solved numerically; the result is an upper bound. If the expected log growth p·ln(1 + b·r) + q·ln(1 − r) is zero or negative, ruin is certain over unlimited trades.
A worked example
Streaks. A system wins 45% of its trades. Over 200 trades the chance of at least one run of 8 losses or more is 53.04%, even though any particular stretch of eight trades is all losses only 0.84% of the time. The expected longest streak is 8.0 losses, and 95% of the time it is no longer than 12 losses.
Drawdown. Risking 1% of the balance per trade, a run of 8 losses cuts a $50,000 account by 7.73% to $46,137.23.
Ruin. With a 40% win rate, winners 1.8 times the size of losers and 2% risked per trade, 13.34% of simulated accounts hit a 30% drawdown within 500 trades. The closed form gives 13.57% for unlimited trades. The median account finishes +175.3% from where it started.
Losing streaks are longer than intuition says
People judge streaks by the odds of one stretch, but a trading record contains hundreds of overlapping stretches. With the same 45% win rate:
| Trades | Chance of 8+ losses in a row | Expected longest streak |
|---|---|---|
| 100 | 30.68% | 6.9 losses |
| 200 | 53.04% | 8.0 losses |
| 500 | 85.40% | 9.5 losses |
| 1,000 | 97.92% | 10.7 losses |
The longest streak grows with the logarithm of the number of trades, so it keeps creeping up as the record lengthens. A system with a sub-50% win rate will produce losing streaks that feel like a broken strategy; knowing the normal range in advance makes it easier to tell bad luck from a real breakdown.
Risk per trade is the lever you control
The win rate and payoff ratio come from the strategy. The risk per trade is a choice, and it moves the risk of ruin more than anything else. For the example system and a 30% ruin level over 500 trades:
| Risk per trade | Simulated risk of ruin | Closed form, unlimited | Median ending balance |
|---|---|---|---|
| 1% | 1.24% | 1.30% | +73.8% |
| 2% | 13.34% | 13.57% | +175.3% |
| 4% | 40.86% | 43.79% | +202.5% |
| 8% | 73.06% | 78.72% | -31.3% |
| 15% | 93.84% | 100.00% | -35.0% |
Higher risk raises the best outcomes but the chance of ruin rises much faster. At 15% the system still has a positive expectancy per trade, yet its typical growth is negative and ruin becomes certain over time. That is over-betting, the same effect the Kelly criterion calculator shows from the growth side.
Ruin from the start or from the peak
A 30% loss from the starting balance and a 30% fall from the highest balance are different events. Profitable accounts drift away from their starting point, so a drawdown from the start becomes less likely over time, while drawdowns from the peak keep coming. Measured from the peak, the example’s risk of ruin jumps from 13.34% to 65.92%; at 1% risk it is 5.82%. If your limit is a trailing one, as with many prop-firm rules, use the peak setting.
Assumptions and limitations
- Independent trades. Real results cluster: strategies have good and bad regimes, which makes long streaks more likely than this model says.
- Every win and every loss is the same size. Real trades vary, and occasional oversized losses (gaps through stops) raise the risk of ruin.
- A constant edge. The win rate and payoff ratio are taken as known. They are estimates, and an overestimated edge understates the risk.
- No costs or withdrawals, unless the payoff ratio already reflects them.
- Simulation error. 5,000 runs give an estimate within the ± range shown about 95% of the time. Ruined accounts stop trading, which pulls down the lower percentiles and the median when ruin is common.