Martingale is one of the oldest ideas in trading, and it survives because it works — right up until the moment it doesn't, and by then it's too late to matter.

What martingale actually is

The concept comes from 18th-century casino betting: after a loss, double your bet on the next round. In theory, one eventual win recovers every previous loss plus a small profit, and since you're guaranteed to win eventually, you're guaranteed to come out ahead.

Applied to trading, this usually looks like: after a losing trade, open the next position at a larger lot size (or add to the losing position at a worse price) — the idea being that when price eventually reverses, the larger position recovers everything at once.

Why it looks brilliant on a backtest

Martingale systems produce beautiful equity curves in backtests, because most of the time, price does eventually revert, and the strategy racks up a long string of small, steady wins. This is exactly what makes it so dangerous — the backtest isn't lying, it's just not showing you the one scenario that matters.

The math that eventually catches up

Every martingale system has a mathematical ceiling: your account size. Each loss doubles (or otherwise scales up) the position size, which means the required capital to survive N consecutive losses grows exponentially, not linearly. A system that survives 5 losses in a row comfortably might need 30 times the capital to survive 8 losses in a row — and financial markets absolutely do produce 8-loss streaks, especially during volatile, trending periods where "reversion to the mean" simply doesn't happen on the timeline the strategy needs it to.

When that streak finally comes — and on a long enough timeline, it always does — the position size required to recover is larger than the account can support. The result isn't a bad week. It's the account.

Why this matters even more on a prop firm account

On a personal account, a martingale blowup costs you your own capital. On a prop firm challenge or funded account, it costs you the account entirely — usually well before the position size gets anywhere near what it would take to actually blow a normal account, because daily and max loss limits are far tighter than a personal account's practical risk tolerance. A martingale system that might survive on a lightly-managed personal account will typically breach prop firm limits far faster.

What real risk management looks like instead

The alternative isn't complicated — it's just less exciting to market:

  • Fixed or risk-percentage-based sizing that doesn't change based on the outcome of the previous trade — or if it does adjust, it reduces risk after a loss rather than increasing it
  • A hard stop-loss on every single position, attached the moment the trade opens, not a mental stop or a "close if it gets bad enough" rule
  • A maximum loss-count circuit breaker — the system stops trading for the day after a defined number of losses, rather than trying to trade its way back to even
  • No averaging into a losing position at a worse price, ever, regardless of how convinced the system is that price will reverse

None of this produces the smooth, uninterrupted equity curve a martingale backtest can show you. It produces something more valuable: an account that's still open next month.

QMS Trading's Gold and Forex Editions are built explicitly without martingale, grid stacking, or loss-averaging of any kind — position sizing is risk-based and adapts down after a loss, never up. See the full risk management breakdown.