TL;DR
Hedging, grid, and martingale bans target three distinct mechanics: opposing positions on the same or correlated instrument, a ladder of orders that adds to a losing position automatically, and sizing up after a loss to recover it faster. A firm doesn't need to know your strategy's name to flag it — the violation is read off the trade log's sizing and sequencing, and it can void a payout even on an account that finished green.
These three bans get lumped together in most rulebooks under one paragraph, and traders read that paragraph the way they read a cookie consent banner — skim, accept, move on. That's the mistake. Each ban is written to catch a different mechanic, and a strategy can violate one without violating the others, or trip a rule you didn't know applied because the EA you're running does something the vendor never called by that name.
What each ban actually restricts
"Hedging," "grid," and "martingale" are trader slang before they're rule language. Firms translate them into something a compliance reviewer or an automated log-checker can actually test for — and that translation is where the ambiguity lives. A hedge to you might be an opposing position on the same symbol; to the firm's terms, it might also mean an opposing position on a correlated pair, or an opposing position across two of your accounts at the same firm.
| Rule | What it actually restricts | Common way traders trip it without meaning to |
|---|---|---|
| Hedging ban | Opposing exposure on the same or correlated instrument, inside one account or across accounts at the same firm | A manual 'insurance' trade against an open position, or two EAs on the same symbol pulling in opposite directions without you coordinating them |
| Grid ban | A ladder of pending orders at fixed price intervals that adds losing legs automatically as price moves against the first entry | A DCA-style or grid EA installed without checking the terms, or a manual scale-in habit that the firm's log review reads as a grid pattern |
| Martingale ban | Increasing position size after a loss to recover it in fewer trades, regardless of what the strategy calls itself | Any sizing model where the next lot size is a function of the prior loss — including systems that never use the word 'martingale' anywhere in their documentation |
Why firms write these three specifically
All three mechanics share a property firms care about: they can produce a smooth, low-drawdown equity curve right up until they don't. A grid can hold a losing position quietly for a long stretch and then take a single large loss that busts the account in one move. A martingale sizing scheme does the same thing by design — it's built to eventually recover a string of losses in one trade, and 'eventually' includes the outcome where it doesn't. Hedging is different in kind: it's less about blow-up risk and more about a firm not wanting to pay out on a position that's structurally guaranteed not to lose, which some firms treat as gaming the profit target rather than trading it.
placeholder
Read the actual clause, not the heading. Some firms define a hedging violation as opposing positions on the identical symbol only. Others extend it to a defined correlation group — EURUSD and GBPUSD, or gold and silver — where holding opposite directional exposure on both counts as a breach even though the instruments aren't the same. A strategy that's clean at one firm can be a violation at another purely because of how wide that correlation definition is written. If you run more than one system, or more than one account, this is the clause to reread before you assume you're fine.
- Read the exact wording of the rule, not just the heading — 'hedging' often means opposing exposure within the same account, and the scope of what counts as 'opposing' varies by firm.
- Check whether the rule applies per-symbol or per-correlated-group — some firms treat correlated pairs as a single exposure for this purpose, some don't say either way.
- Look at how your EA or manual routine sizes the next trade after a loss — if lot size is a function of the prior loss, that's the pattern a martingale clause is written to catch, whatever the strategy calls itself.
- Ask whether the firm evaluates this at execution or retroactively from your trade log — most of these are caught during a payout review of the full trade history, not blocked live.
- Test the rule against your actual last challenge or funded account's trade history, not a demo run under different conditions — sizing patterns that look fine in isolation can look different across a full trade log.
Grid and martingale: named because they're detectable, not because they're rare
A grid EA doesn't need to be labelled 'grid' to get flagged — a reviewer or an automated log parser can see it in the pattern: pending orders at regular intervals, lot sizes that don't change with market conditions, a position count that only grows in one direction until it reverses all at once. Same with martingale sizing — it shows up as a lot-size series that steps up cleanly after a loss and resets after a win. Some retail EAs sold as 'grid' or 'DCA' strategies, including ones commonly discussed under names like the grid-style systems listed in general strategy catalogs, are built on exactly this mechanic and don't always disclose it plainly in their marketing.
The detail traders miss: the firm doesn't have to prove intent. It's reading a sequence of trades and lot sizes off a log. If the pattern matches the definition in the terms, the account gets flagged whether the trader meant to run a grid or just liked scaling into positions manually and never called it that.
Detection is retroactive — that's the part the tooling gap actually hits
This is where most of these violations actually get caught: not live, but during a payout review, when someone pulls the full trade history and looks at the shape of it. A pre-trade check can catch a hard, well-defined breach in real time — daily loss limit, max position size, a symbol restriction. Sizing-pattern violations like grid and martingale are softer to define at the point of the order; they're a pattern across many trades, not a single number a system can reject on the spot. That doesn't mean tooling is useless here — it means the useful tooling is different: checking the terms before you deploy a strategy, not just checking each order as it fires.
Where to actually check before you run anything
The terms are the terms — there's no substitute for reading the specific firm's wording on hedging, grid, and martingale before deploying a strategy that touches any of the three. What's worth doing before that: mapping out how your strategy actually sizes trades and whether it ever opens opposing exposure, so you know which clause to go looking for instead of skimming the whole document hoping the relevant line jumps out.
Citations
FAQ
- Does every prop firm ban hedging, grid, and martingale strategies?
- No, and where they do, the wording differs enough that a strategy allowed at one firm can be a clear breach at another. There's no industry-standard definition of any of the three — check a specific firm's own terms rather than assuming a general rule applies.
- Is scaling into a losing position always a martingale violation?
- Not necessarily — the distinguishing feature firms look for is whether position size increases as a function of a prior loss. Adding to a position based on a technical signal, unrelated to whether the last trade lost, is a different pattern than a martingale clause is usually written to catch, though a reviewer's read of your log is what actually decides it.
- Can a hedging ban apply across two accounts at two different firms?
- Generally no — most hedging clauses are scoped to a single firm's own account or accounts, not exposure held anywhere else in the market. But some firms do restrict hedging across multiple accounts you hold with them specifically, so check whether the clause names 'this account' or 'accounts held with us.'
- How do firms actually detect grid trading in a trade log?
- By pattern, not by label: a run of pending orders placed at regular price intervals, lot sizes that don't vary with conditions, and a position count that only grows before a single large exit. It's read off the sequence and sizing, which is why it's typically caught during a review of the full history rather than blocked at any single order.
- Does a hedging ban usually include correlated instruments, or only the same symbol?
- It depends entirely on the firm's wording — some restrict opposing positions on the identical instrument only, others extend the definition to a stated correlation group. This is one of the most commonly misread lines in a rulebook because the scope isn't always obvious from the heading alone.
- Will a pre-trade risk check stop me from accidentally running a martingale-style EA?
- It can flag hard, well-defined limits like a daily loss cap or an oversized position at the moment you submit an order. Sizing-pattern violations like martingale or grid are read across many trades, not one order, so the more useful safeguard is checking a strategy's sizing logic against the firm's terms before you deploy it, not relying on a single pre-trade block.
How we maintain accuracy
Reviewed by Ryan Tran, Strategy Lead, Glitch Executor. Every quantitative claim cites a primary source; firm-rule values come from the firm-rule registry audited quarterly in this repo. No paid placements, no fabricated reviews.
Post last reviewed . Tier 1 surface last reviewed .

Written by Ryan Tran
Strategy Lead · Glitch ExecutorWrites on prop-firm rule modelling, backtest correctness, and why most "passed challenge" stories don't reproduce.
More from Ryan
