Monte Carlo Analysis Simulator

How the Monte Carlo simulator works

A backtest gives you one history. That history is one sample of luck: the same strategy with the same odds could have drawn its wins and losses in another order and ended somewhere else. Monte Carlo analysis plays the odds forward many times and shows the spread.

This tool draws 2,000 paths. On every path each trade is a win with the chance you enter and a loss otherwise. A win adds the amount at risk times the payoff ratio, and a loss takes off the amount at risk. The random numbers come from a seeded generator (mulberry32, seed 987654321), so the output is the same every time. Then it sorts the 2,000 final balances, and separately the 2,000 maximum drawdowns, and reads the percentiles.

A percentile is a position in a sorted list. The 5th percentile final balance is the value that 5 percent of the paths finished at or below. The 95th is the value 95 percent finished at or below. The 50th is the median.

A worked example

Take a made-up strategy: $10,000, a 40 percent win rate, a payoff ratio of 1.6, 2 percent of the balance risked per trade, 250 trades. The average result is 0.4 x 1.6 - 0.6 = 0.04 times the amount risked, a small positive edge. The median final balance across the 2,000 paths is $11,269. The 5th percentile is $5,754 and the 95th is $20,957. Same odds, same size, and the endings run from $5,754 to $20,957, which is luck alone. 37.2% of the paths ended below the starting $10,000.

The drawdowns tell a second story. The median path fell 32.8% from a peak at some point, and the 95th percentile path fell 54.3%. A person who could not sit through a 54.3% fall would have stopped on about one path in twenty, whether or not the strategy was sound.

Now switch the sizing to a fixed $200 per trade, which is 2 percent of the starting balance only at the start. The median final balance becomes $12,000, the 5th percentile $5,240 and the 95th percentile $18,760. Fixed sizing does not compound, so the lucky end is lower, and it does not shrink after losses, so the unlucky end can fall further in relative terms.

A fixed check: with a 0 percent win rate and 10 trades at 10 percent, the balance is 10,000 x 0.9 ^ 10 = $3,487, and the drawdown is 65.1 percent every time.

The same made-up strategy under two sizing rules. Calculated by us on 27 Oct 2026 from the tool above, exact for this seed. The rows of the drawdown column are a separate sorted list from the balance column.

What it shows and what it cannot

What it shows

• Path luck. How much the ending and the worst fall vary when nothing but the order of wins and losses changes.
• The drawdown range. The depth to plan for is not the median fall. It is a bad but ordinary case, such as the 95th percentile.
• The effect of size. Raising the risk percent widens everything: the lucky end, the unlucky end and the drawdowns.

What it cannot show

• A change of regime. Markets change. A strategy tuned to one period can lose its edge, and then the odds you typed in are not the odds you are running.
• A bot's tail. A grid or martingale bot wins small and often, then takes one very large loss that its history may not contain. Typing a high win rate and a small payoff ratio into a model of independent trades hides exactly that event. Check it with the risk of ruin calculator and by reading the bot's margin.
• Odds that were fitted. If the win rate and payoff came from an optimised backtest, they are probably too good. See overfitting.
• Dependence between trades. Real losses cluster. This tool draws each trade on its own.

Resampling your own trades

Many Monte Carlo tools do not draw wins and losses from a win rate. They take your list of real trades and reshuffle the order, or draw from the list with replacement. That keeps your real trade sizes, including unusually large ones, and removes the assumption of a single average. This page uses fixed odds because they are simple to enter and to test. The reshuffle approach is better when you have a long, honest list of trades, and it still cannot invent a loss your history never contained.

Reading the drawdown numbers

For the drawdown that matters to your account in dollars, see the drawdown calculator, which converts a percent fall to what is left and the gain needed to recover.