What these clauses are for
A funded account trades on the firm's pricing, often in a simulated environment. Anything that extracts money from the mechanics of that environment rather than from the market is a problem for the firm, and every rulebook prohibits it.
Most of the list is uncontroversial and would not describe anything you do. The risk sits in two places: the breadth of the wording, and the handful of prohibitions that catch ordinary trading by accident. Those are what this guide is about.
Copy trading across accounts
Banned almost everywhere, including between accounts you own yourself. The concern is a trader running opposite positions across two accounts so that one passes whatever the market does — the firm has then sold two evaluations and funded a coin flip.
The accidental version is more common than the deliberate one: you hold accounts at the same firm and trade both manually with the same strategy at the same times. In the account history that is indistinguishable from copy trading. If you run several accounts, expect to explain it, and read managing several funded accounts before you open the second one.
The same clause usually covers trading someone else's account or having yours traded for you, which matters if you were considering sharing an account with a friend. Firms detect it through IP addresses, device fingerprints and fill-time correlation, and it is one of the few violations they pursue actively rather than at review.
Latency and pricing arbitrage
Latency arbitrage — exploiting a delayed feed against a faster one — is prohibited universally and is a standard reason for voiding profit. So is arbitraging between the firm's price and another venue.
Very few retail traders do this deliberately. The relevance is that the clause is usually written broadly enough to cover any strategy whose profit comes from execution rather than direction, which can extend to some high-frequency scalping that has nothing to do with feeds. If your average trade lasts seconds and your edge is measured in ticks, this clause is closer to you than you think.
Tick scalping and minimum hold times
Some firms impose a minimum time in trade — commonly a few seconds — or prohibit strategies whose average hold falls below a threshold. The stated reason is that very short holds profit from spread mechanics rather than from movement.
If you scalp, find the number. A strategy with a four-second average hold is fine at a firm with no minimum and voided at one with a ten-second rule, and you will not discover the difference until a payout is reviewed. Ask whether the rule applies to the average or to individual trades, because those are very different constraints on the same strategy.
Hedging
Hedging within one account is usually permitted. Hedging across accounts, or across firms, is usually prohibited — again because it converts a coin flip into a guaranteed pass on one of them.
Correlated hedging — long one index, short a closely related one — sits in a grey area at many firms. It is rarely named explicitly and is easily caught by a catch-all if the correlation is high enough to look deliberate. If it is part of your approach, ask before you rely on it.
Automation, EAs and bots
The most common question, and the answer is better than most traders expect: automation is permitted at the large majority of firms. You may run an expert advisor, a script or a bridge to your own code, provided the strategy it runs would be legal if you clicked it manually.
The exceptions are worth knowing precisely:
- Latency-sensitive or high-frequency bots fall under the arbitrage clause regardless of how they are written.
- Commercially sold EAs are banned at some firms specifically because many traders run the same one, which produces correlated positions across hundreds of accounts and creates exactly the exposure the firm cannot hedge.
- Any tool that manipulates the platform rather than trading through it — modifying feeds, injecting fills — is universally prohibited and treated as fraud rather than a rule breach.
- A minority of firms ban all automation in the evaluation while permitting it once funded, or vice versa. It is asymmetric often enough to be worth checking rather than assuming.
If your edge is automated, get the permission in writing with the strategy described in a sentence. "Do you allow EAs" invites a yes that will not protect you; "do you allow an EA that trades one instrument on a 15-minute chart with an average hold of two hours" invites an answer you can rely on.
Martingale, grid and averaging down
Rarely prohibited outright, and frequently discouraged in wording that becomes a problem later. A grid strategy that opens fifteen positions against a trend does not violate a named rule at most firms — but it produces exactly the equity profile that triggers a manual review, and it interacts terribly with an equity-measured drawdown that counts unrealised loss.
Where it does appear as a rule it is usually phrased as a limit on total exposure or on positions in the same instrument, rather than by name. That is the wording to search for.
Gap and news exploitation
Deliberately holding into a scheduled release or a weekend to capture a gap is prohibited at many firms even where news trading generally is not. The distinction is intent, which the firm infers from your position timing — a position opened four minutes before a rate decision reads differently from one opened three days earlier.
Group and coordinated trading
A clause that has grown more common: trading the same signals as a group, whether from a paid signal service, a community, or several friends running one plan. The firm's exposure is the same as with copy trading — many accounts, one position — and the detection is the same too.
This one catches people who have done nothing wrong by any ordinary standard. Following a public signal channel is not cheating in any moral sense, but it produces correlated fills across accounts and is treated accordingly. If you trade someone else's calls, assume it is visible.
How firms actually detect this
Understanding the detection explains why some clauses are enforced and others are not.
What is checked automatically: fill timestamps against a news calendar, hold times, position correlation across accounts on the same IP or device, and the profit distribution used by the consistency review. What is checked manually, and only when a payout is large or a pattern looks unusual: strategy characteristics, the catch-all, and anything requiring judgement.
The practical implication is uncomfortable but useful: small payouts are rarely scrutinised, and the first large one usually is. Traders who have withdrawn four times without a question and then hit a review on the fifth have not been treated inconsistently — they have crossed the threshold where a human looks.
The catch-all clause
Nearly every rulebook contains a sentence permitting the firm to void profit from trading it considers abusive, manipulative or inconsistent with the spirit of the programme. This is the clause a refusal will cite, and no amount of compliance with the specific rules removes it.
You cannot negotiate it away, but you can weigh it. A firm that pairs a broad catch-all with fully quantified specific rules is behaving reasonably — the catch-all is there for cases nobody anticipated. A firm whose specific rules are also vague has effectively reserved total discretion, and that is the single most reliable warning sign in how to spot a bad firm.
If a strategy is central to your edge
Ask support in writing, describe the strategy plainly in two or three sentences, and keep the reply. Include the instrument, the average hold, the number of positions you may have open, and whether it is automated.
It is not a guarantee — support cannot bind the risk desk at every firm — but a written answer describing your actual strategy is the single most useful document to have if a payout review ever questions it. It also tells you something before you pay: a firm that answers a specific question with a specific answer within a day is a different counterparty from one that replies with a link to the terms.