Risk of ruin calculator
Simulate thousands of trade sequences with your win rate and R:R to see your real probability of blowing up the account and the worst losing streak.
By the TradingCalculator.Pro team · Updated on · About us
Start your 7-day free trial →Try position sizing free, no account needed →
Formula
Thousands of random simulations of your systemHow it works
A Monte Carlo simulation shuffles your system. It takes your hit rate, your risk/reward ratio and your risk per trade, generates thousands of random sequences with those properties and looks at what happened in each. Expectancy tells you whether the system wins in the long run; Monte Carlo tells you what can happen to you along the way.
Risk of ruin is the share of those sequences that hit your ruin threshold — usually losing half the account — before they end. It is not determined by your edge alone: it is determined by the edge combined with bet size. A good system risking too much goes broke; a mediocre one risking little survives long enough for the edge to show. That is why this calculation, not intuition, is what sets the risk percentage.
What you get
- Risk of ruin for your bet size
- Worst expected losing streak
- Validate if your system is sustainable
Worked example
A system with a 45% hit rate and a 1:2 ratio, simulated over 200 trades and 40,000 sequences, with ruin set at losing 50% of the account.
- Expectancy per trade: 0.45 × 2 − 0.55 = 0.35 R, a solid edge.
- Risking 2% per trade: risk of ruin 0.00%, median drawdown 18.3% and 29.2% in the worst 5% of cases.
- The same system at 5% per trade: risk of ruin 1.54%, median drawdown 40.5% and 59.1% in the worst 5%.
- With a thinner edge (40% hit rate, 1:1.5) at 2%: risk of ruin 6.13% and a 33.0% median drawdown.
The same system goes from never going broke to doing so in 1 of every 65 sequences purely by raising risk from 2% to 5%.
Notice what did NOT change: expectancy is 0.35 R in both cases. Raising risk does not improve the edge, it only widens the fan of outcomes in both directions — and ruin is absorbing: an account that reaches zero stops playing, however good the expectancy was.
Drawdown and ruin are not the same question
Ruin is the probability of falling so deep you do not come back; drawdown is how far you fall along the way. A system can never go ruined and still pass through declines nobody sits through. These figures come from 40,000 sequences of 200 trades each, with ruin fixed at losing half the account:
| Risk per trade | Win rate | R:R | Risk of ruin | Median drawdown | Worst 5% |
|---|---|---|---|---|---|
| 1% | 45% | 1:2 | 0.00% | 9.6% | 15.6% |
| 2% | 45% | 1:2 | 0.00% | 18.3% | 29.2% |
| 3% | 45% | 1:2 | 0.06% | 26.3% | 40.9% |
| 5% | 45% | 1:2 | 1.54% | 40.5% | 59.1% |
| 2% | 40% | 1:1.5 | 6.13% | 33.0% | 52.3% |
The first four rows are THE SAME SYSTEM with the same expectancy per trade (0.35R). The only thing that changes is the size of the bet, and with it the median decline goes from 9.6% to 40.5%. Raising risk does not improve the edge: it widens the fan in both directions, and ruin is absorbing.
The last row is the one to look at
Compare the second and the fifth. Both risk 2% per trade, and one never goes ruined while the other does 1 time in 16. The difference is not in the risk: it is in the edge. A system with a 45% win rate at 1:2 has an expectancy of 0.35R; one at 40% with 1:1.5 has 0.00R — it is exactly break-even.
That means a "safe" risk percentage does not exist in the abstract. 2% is prudent on a real edge and a slow bleed on a system that barely breaks even. The question that comes before "how much do I risk" is always "how much edge do I have", and that is measured with a record, not with an estimate.
There is also an effect the table does not show and that matters more than it looks: costs. Commissions and slippage come out of the gross edge before anything else, and in frequent trading they can turn the second row into the fifth without the win rate having moved a single point.
What a simulation does not know
Monte Carlo bounds what can happen given the properties you feed it, and it is worth knowing its three assumptions, because all three break in real life.
- It assumes trades are INDEPENDENT. They are not if your system opens several correlated positions at once: there the losses arrive together and the streaks are worse than the simulated ones.
- It assumes your edge does not change. A system that worked in one volatility regime can stop working in another, and the simulation will go on projecting the old edge.
- It assumes you execute the same way every time. After six losses in a row most people change something — the size, the stop, the entry — and from then on they are trading a different system from the one that was simulated.
None of this invalidates the tool: it makes it what it is, a way of sizing risk with the data you have, not a prediction. If the three assumptions make you uncomfortable, the right answer is to lower the risk, not to stop simulating.
How to use it
- Enter your inputs: the terms of the formula (capital, risk, prices and stop).
- The calculator applies the formula and returns the result in your instrument's units.
- Check the breakdown before trading: anything that cannot be computed is shown empty, never as zero.
Frequently asked questions
What is my risk of ruin at 2% risk per trade?
With a 45% hit rate and a 1:2 ratio, effectively zero over 200 trades, with an 18.3% median drawdown. With a thinner edge — 40% hit rate and 1:1.5 — that same 2% gives a 6.13% risk of ruin. The figure does not depend on risk alone: it depends on crossing risk with the edge.
What is the difference between risk of ruin and drawdown?
Drawdown is how far you fall from a peak; risk of ruin is the probability of falling so far you do not come back. One system can have an 18% median drawdown and a 0% risk of ruin, and another a similar drawdown while going broke once in sixteen runs: what separates them is the tail of the distribution, which is exactly what an average hides.
How many trades does the simulation need to mean anything?
At least 100 trades per sequence and several thousand sequences; here it is 200 and 40,000. With fewer trades chance dominates the result, and with few sequences the tail — the part that matters — never shows up. And your hit rate has to come from a real track record, not an optimistic estimate.
Why is a losing streak more likely than it looks?
Because there are many chances for it to happen. At a 45% hit rate, the probability of losing eight in a row on one specific attempt is 0.8%, but across 200 trades there are 193 possible attempts and the streak appears 0.73 times on average. At 2% per trade, those eight losses are 14.9% of the account.
Does Monte Carlo predict my result?
No: it brackets the range of possible outcomes given the properties you feed it. It does not know whether your edge will still exist next month, and it assumes trades are independent, which is untrue if your system clusters through correlation or if your own discipline cracks after several losses. It is a tool for sizing risk, not a crystal ball.