Martingale and Prop Firm Rules

Martingale is the most reliably appealing idea in trading and the most reliable way to lose an evaluation. It produces a smooth, profitable-looking equity curve almost all of the time, and then gives the entire account back in a single sequence. Firms do not ban it because it is unusual — they ban it because the risk it creates is structurally incompatible with an account that has a hard loss limit.

What martingale actually does

The rule is simple: after a loss, double the size. The purpose is that one win recovers every previous loss plus the original target. A trader running it wins small amounts very often and loses one very large amount eventually.

Grid trading is a close relative. Rather than scaling geometrically, it adds positions at fixed intervals as price moves against the original entry, building a larger position at an average price closer to the market. The intent is the same — avoid realising the loss by improving the average — and the exposure profile is the same.

The arithmetic, laid out

Start from a position risking 1% of the account and doubling after each loss. The individual trade risk grows like this:

A typical evaluation permits a 10% loss. That is crossed during the fourth trade of the sequence. Five consecutive losses — an ordinary event for most strategies, and close to certain over a few hundred trades — removes more than half the account. The method does not need a disaster to fail; it needs a normal losing run.

Why firms write it into the prohibited list

Rulebooks usually group martingale, grid and related sizing under a prohibited-strategies clause, and the reasoning is consistent:

• It is incompatible with a fixed drawdown. The firm's risk limit is a hard floor. A method that accepts an unbounded loss sequence will eventually meet it.
• It hides the true risk while the curve looks good. Monitoring that sees a smooth equity line is monitoring that cannot see the position size doubling behind it.
• It transfers an unbounded liability. On a funded account the size can exceed anything the fee accounts for, which is not a risk a firm accepts from an unknown trader.
• It behaves like gambling, not trading. The method's outcome is decided by the arrival of a streak rather than by edge, and firms generally define what they want as demonstrable skill.

Why it can look like it works

This is the part that makes the discussion necessary. Martingale genuinely does produce long runs of smooth profit. A trader can run it for weeks, see a rising curve, and reasonably conclude that the method works and that the warnings are theoretical.

Two things are being mistaken for each other. The first is frequency of winning runs, which martingale maximises. The second is survival of the whole period, which is what an evaluation requires and what the method does not provide. A strategy that is right 95% of the time and loses everything 5% of the time is not a good strategy on an account whose failure is permanent, because the 5% is the only outcome that matters.

The legitimate version of the same instinct

The instinct underneath martingale — recover quickly after a loss — is worth examining rather than suppressing. It usually points at one of two things:

1. An oversized position in the first place

   If a loss feels worth doubling to recover, the original size was too large for the account's rules. The fix is at entry, not at the loss. Sizing against the daily drawdown is the mechanism.

2. A strategy without a real edge being propped up

   If a system needs recovery sizing to be profitable, the underlying entries and exits are not the edge — the sizing is, and it is a losing bet. Back-test the entries alone and see what remains.

The correct response to a losing streak is the opposite of martingale: reduce size, or stand aside until the conditions that suit the strategy return. Both make the bad case smaller, which is what a drawdown-limited account requires.

If the pattern is already established and you recognise it, the honest question is not how to tune the sizing but whether the discipline is in place — revenge trading is the same failure with a different name.

Frequently asked questions

The FAQ block covers what martingale is, why firms ban it, why it can appear to work, how it differs from grid trading, the alternative, and how few consecutive losses it takes to break an account. Related: hedging rules, copy trading rules and daily drawdown explained.