Martingale rules on prop firm accounts: why the pattern, not the name, is banned

Martingale is the one trading method that is almost universally banned on prop firm accounts, and the interesting thing is not that it is banned — it is why, and what the ban actually covers. A trader who thinks "I do not run a martingale" can still trip the rule, because what firms prohibit is a pattern, not a name.

What martingale actually is, in account terms

In its pure form, martingale doubles the stake after each loss, so a single eventual win recovers all prior losses plus one unit. The strategy has a well-known property: it has a very high probability of a small gain and a small probability of a total loss. Over a finite account with a drawdown cap, the small probability is the only one that matters, because it is the one that ends the account.

The mechanism is not complicated to see:

TradeStake (units)Cumulative loss if it loses
111
223
347
4815
51631
63263

Six losses in a row, which happens regularly in any honest sequence, requires you to stake 32 units to recover 63 — and the drawdown consumed by that sequence is 63 units, against a rulebook that permits a fixed fraction of the account. The "high win rate" is bought by a loss distribution whose tail is the entire account.

Why the ban exists

Firms ban it for the same reason they cap drawdown: a martingale converts a bounded loss into an unbounded one. On the firm's side, the expected value of the account is not the issue — the variance is. A strategy with a 95% chance of a small gain and a 5% chance of a total loss is, from the firm's perspective, a machine for manufacturing the exact event (a hard breach) that their whole rulebook exists to bound.

There is also a fairness argument the rules pages make explicit: a martingale that recovers its losses is, in effect, borrowing from the drawdown budget to produce a gain, and the risk it takes is not proportional to the reward.

What the rule actually covers

The ban is usually written as a prohibition on increasing position size after a loss, or more generally on a sizing pattern that increases risk as the account's drawdown deepens. That is broader than "doubling". It catches:

  • Fixed multiples — doubling, tripling, or any fixed factor after a loss.
  • Add-to-loser averaging — adding to a losing position to lower the average entry. Economically the same bet as martingale, and often named separately.
  • Recovery scaling — increasing size not after a loss but after a string of losses, to "get it back".
  • Grid and hedging-martingale systems — the classic grid EA, which places progressively larger orders against the move.

A trader who sizes risk as a function of remaining drawdown will sometimes increase size after a loss for benign reasons (more room relative to a fixed stop), and firms distinguish that from martingale by the direction of the risk curve: does your risk increase as the account's safety decreases? If yes, it is martingale-shaped regardless of the arithmetic.

The anti-martingale, and its own trap

The "correct" alternative is anti-martingale: increase size after wins, decrease after losses. That matches the intuition that you should press an edge only when it is working. It also has a trap of its own on a prop account — because after a win, your drawdown floor may be trailing upward, so the "more room to press" is illusory. Pressing after a win on a trailing drawdown account increases both your size and your distance from a floor that just moved.

The robust answer on a capped account is neither: constant risk, sized from the remaining budget, not from the last result. Doing nothing in response to a win or a loss is, mechanically, the strategy with the smallest tail.

The detection, and the payout risk

Like hedging, martingale is detected retrospectively. The firm's systems see a rising size series correlated with a falling equity, and the flag is raised when a payout is requested. The trader who passed with a martingale may find the profit declined, and the account's history is the evidence. There is no version of this rule that only fires during the evaluation.

What to do

  1. Fix your risk per trade as a fraction of remaining room, and if you must vary it, vary it down when room shrinks, never up.
  2. Never add to a losing position on an account with a drawdown cap. The averaging trade is the martingale trade.
  3. Treat "recovery scaling" as the same rule. The label does not matter; the risk curve does.
  4. Check your EA for grid logic. Many commercial grid systems are martingale-shaped under the hood even when the marketing says otherwise.

The rule is not really about the name. It is about whether your risk grows as your account's safety shrinks — and on a product whose entire design is a hard floor, the answer has to be no. The firm-by-firm wording, and which systems fall inside the ban, is in the martingale rules breakdown.

Comments

Post a Comment

Popular posts from this blog

What a prop firm challenge actually costs — the full arithmetic

The number on the challenge page is the fee. It is also the smallest part of what a challenge costs, and the two bigger parts are almost never shown next to it. This is a full accounting of the three costs, with the arithmetic worked out end to end, so you can price an attempt before you buy one rather than after the third reset. Every numeric result below comes from a simulator I wrote and ran to produce this page, with a fixed random seed so the figures are reproducible exactly. The code, the assumptions it makes, and the places it is likely to be wrong are all stated, because a cost figure you cannot audit is not much better than the sales page it came from. Cost one: the evaluation fee This is the advertised price and the only one most traders budget for. It scales with account size: a $100,000 evaluation typically costs several hundred dollars, and a $25,000 one a fraction of that. What that money buys is a single attempt at the evaluation, and on most firms it is not refund...
Read more

Why Most Prop Firm Evaluations Fail on Rules, Not Strategy

A trader can hold a genuinely profitable edge and still wash out of a prop firm evaluation. Across more than 1,400 managed evaluations at FTMO, FundedNext, Apex Trader Funding, Topstep and E8 Markets, one pattern repeats: the overwhelming majority of failures are rule breaches, not bad analysis. That distinction matters. A losing trade is a market outcome you can study and fix. A breached daily loss limit is an administrative failure — one position held too long — and it ends an evaluation instantly, regardless of how good the thesis was. Trailing drawdowns are the most misunderstood rule in the industry Treating a trailing drawdown as static is the single most expensive mistake in prop trading. At firms using trailing drawdowns, the loss buffer rises as your equity peaks. Unrealised profit has already moved your floor upward, so giving it back is doubly expensive. Consider a $100,000 futures evaluation with a $3,000 trailing drawdown. Run equity to $101,500 in open prof...
Read more