Martingale: the arithmetic of ruin
Posted: Wed Oct 28, 2026 4:21 am
Double the position after every loss, and the first win recovers everything plus a profit. It's the oldest betting system, and it has a perfect record on paper, until the sequence of losses exceeds the capital.
The maths is simple. If the starting size is 0.01 lots and the strategy doubles after each loss, ten consecutive losses require a position of 5.12 lots, and total capital at risk grows to 10.23 lots-worth of losses. A ten-loss streak in a 50% strategy occurs about once in every thousand sequences, and over several thousand trades, it's nearly certain.
Brokers limit position sizes and margin, so the sequence often breaks before the theoretical limit. The result is a large loss, wiping out many small wins.
Variants like soft martingale, multiplying by 1.5 or adding fixed increments, slow the explosion without removing it.
The attraction is that the system produces a high win rate and steady growth for a long time. That's what makes it so dangerous. The curve looks great until it doesn't.
Casino operators love martingale players, since their winning streaks are small and frequent, while their losses are big and catastrophic, and the house limit on bets exists precisely to cap the doubling. Brokers' margin rules serve a similar purpose.
A simulation with 10,000 sequences, using the strategy's actual win rate, shows how often the account is wiped out before reaching a target. For most realistic win rates and account sizes, the answer is uncomfortably often.
How long would a martingale survive on the strategy's actual loss streak distribution, and what capital would it need to survive the worst case in the data?
The maths is simple. If the starting size is 0.01 lots and the strategy doubles after each loss, ten consecutive losses require a position of 5.12 lots, and total capital at risk grows to 10.23 lots-worth of losses. A ten-loss streak in a 50% strategy occurs about once in every thousand sequences, and over several thousand trades, it's nearly certain.
Brokers limit position sizes and margin, so the sequence often breaks before the theoretical limit. The result is a large loss, wiping out many small wins.
Variants like soft martingale, multiplying by 1.5 or adding fixed increments, slow the explosion without removing it.
The attraction is that the system produces a high win rate and steady growth for a long time. That's what makes it so dangerous. The curve looks great until it doesn't.
Casino operators love martingale players, since their winning streaks are small and frequent, while their losses are big and catastrophic, and the house limit on bets exists precisely to cap the doubling. Brokers' margin rules serve a similar purpose.
A simulation with 10,000 sequences, using the strategy's actual win rate, shows how often the account is wiped out before reaching a target. For most realistic win rates and account sizes, the answer is uncomfortably often.
How long would a martingale survive on the strategy's actual loss streak distribution, and what capital would it need to survive the worst case in the data?