Risk of Ruin Calculator
Enter your win rate, average win and loss, risk per trade and the drawdown you call ruin. A 10,000-run simulation and the classic formula show the chance of reaching it, with the typical and bad-case drawdown.
- Classic formula (no time limit)
- —
- Expectancy per trade
- —
- Payoff ratio
- —
- Losses in a row to reach the level
- —
- Median max drawdown
- —
- 95th percentile max drawdown
- —
- Median ending balance
- —
- Runs ending below the start
- —
Risk per trade compared (same win rate, R and ruin level)
| Risk per trade | Classic formula | Simulated (your trades) | Median max drawdown | 95th percentile |
|---|
What risk of ruin means
Risk of ruin is the chance that a run of losses takes your account down to a level you cannot accept, here called the ruin level: for example 20% below the starting balance, or a prop firm's maximum loss. It depends on four things: your win rate, the size of your average win and loss, your risk per trade and how deep the ruin level is. This calculator gives two answers. A Monte Carlo simulation plays 10,000 random sequences of your trades over the number of trades you choose. The classic formula gives the chance with no time limit.
How to use it
- Enter your win rate and average win and loss in R from a journal or backtest after costs. If you know your payoff ratio (average win ÷ average loss), enter it as the average win and leave the average loss at 1.
- Enter your risk per trade and the drawdown you count as ruin.
- Choose how many trades to simulate, whether risk is fixed in dollars or a percentage of the current balance, and whether the drawdown is measured from the start or from the highest balance so far.
- Compare the rows of the table: the same method at different risk per trade.
The method
- Expectancy = win rate × average win − loss rate × average loss (in R)
- Classic formula: with 1:1 payoff and an edge A = win rate − loss rate, the chance of ever losing U risk units is ((1 − A) ÷ (1 + A))U, where U = ruin level ÷ risk per trade. For other payoffs the calculator solves p·z−win + q·zloss = 1 and uses z−U, the standard generalisation. It measures from the starting balance, has no time limit and is 100% when expectancy is zero or negative.
- Simulation: 10,000 runs of N trades. Each trade wins with your win rate. A win adds average win × risk; a loss takes average loss × risk. A run counts as ruined if the balance touches the ruin level at any point. The maximum drawdown of each run (the largest fall from a high, as a % of that high) gives the median and the 95th percentile.
Assumptions. Every trade is independent of the last, the win rate and R stay fixed, every loss is exactly the average loss and every win exactly the average win, and risk per trade never changes. Real trading breaks all of these: losses cluster, stops slip on gold news, and traders change size after a streak. Treat the result as a lower estimate of the real risk and a way to compare settings, not a forecast.
Worked examples
Example 1: the classic coin-flip case. Win rate 50%, 1R wins and 1R losses: the edge is zero, so with no time limit the chance of losing 20% is 100%. Within 250 trades at 2% risk (10 losses net to reach −20%), the exact chance is 52.8%. Any method without an edge reaches the level eventually; a short horizon only hides it. At a 55% win rate the edge is A = 0.10, and the classic formula gives ((1 − 0.1) ÷ (1 + 0.1))10 = 13.4%.
Example 2: a 2R method at 1%, 2% and 4% risk. Win rate 45%, 2R wins, 1R losses, ruin = −20% from the start, fixed dollar risk. Expectancy = 0.45 × 2 − 0.55 = +0.35R. At 1% risk the level is 20 losses away: classic formula 0.12%, exact chance within 250 trades 0.12%. At 2% it is 10 losses away: 3.4% and 3.4%. At 4% only 5 losses: 18.5% and 18.5%. Same method, same edge: the risk per trade alone multiplies the danger.
Example 3: reading the simulation. Enter the example 2 numbers at 2% risk. The simulated chance sits close to the exact figure above; it differs by a fraction of a percentage point because 10,000 runs are a sample. The median and 95th percentile maximum drawdown show what an ordinary run and a bad run look like. Switch the drawdown to be measured from the highest balance: the chance rises, because a drawdown can now start after the account has grown.
Common mistakes
- Using a win rate from a small sample. Twenty trades can show almost any win rate. Check your edge first with the expectancy calculator.
- Treating a low number as safe. The model assumes independent trades and perfect stops. Real losses come in clusters, especially when a market regime changes.
- Raising risk after a losing streak. That is the fastest way to turn a small risk of ruin into a large one; see why martingale is dangerous on gold.
- Forgetting the time limit. With no edge, ruin is only a matter of time: a short simulation can look comfortable.
Learn more
See how deep a losing streak goes and what recovery takes with the drawdown recovery calculator, read how to handle a gold losing streak and the risk of ruin glossary entry. For a funded account, the prop challenge calculator and the prop drawdown tracker apply the same idea to a challenge's limits.
Before you plug a strategy into any calculator, check how it actually behaved on gold: our gold trading strategies, tested and gold indicator settings, tested hubs show real XAUUSD results with costs, including the losing streaks.
Frequently asked questions
What is risk of ruin in trading?
The chance that losses take your account down to a level you cannot accept, such as a 20% drawdown or a prop firm’s maximum loss. It depends on win rate, average win and loss, risk per trade and how deep the ruin level is.
What is the risk of ruin formula?
For 1:1 payoff, the classic formula is ((1 − A) ÷ (1 + A))^U, where A is win rate minus loss rate and U is the number of risk units to the ruin level. With other payoffs, solve p·z^−win + q·z^loss = 1 and take z^−U. It assumes independent trades, fixed risk and no time limit.
Why use a Monte Carlo simulation?
It works for any payoff, for a fixed number of trades, for risk taken as a % of the current balance and for drawdowns measured from the peak, where a simple formula does not exist. It also shows the spread of outcomes, such as the median and 95th percentile drawdown.
What is a good risk of ruin?
There is no standard figure. Lower is better, and because the model ignores clustered losses, slippage and changes in your edge, the real risk is usually higher than the number shown. Compare risk settings with each other rather than chasing one target number.
How does risk per trade change risk of ruin?
Strongly. Halving risk per trade doubles the number of losses needed to reach the ruin level, and the chance falls far faster than by half. A 2R method with a 45% win rate goes from about 0.1% at 1% risk to 3.4% at 2% and 18.5% at 4% risk for a 20% ruin level, with no time limit.
Why does the result not change between visits?
The simulation uses a fixed random seed, so the same inputs always produce the same 10,000 runs. That makes results repeatable and comparable when you change one input at a time.
Educational tool only. Not financial advice and not a signal. Check your own broker's contract specification, margin rules and costs. Past data does not predict future prices. See our risk disclaimer.