FTMO EA Rules: Request Limits and the $400k Cap

Most of what gets written about FTMO and automated trading is wrong in the same direction. Search it and you'll find pages implying approval processes, source code submissions and restricted strategy types, usually on sites paid a commission when you buy a challenge.

FTMO's own position is simpler than that, and it's published. Your trading style is up to you, including algorithmic trading and Expert Advisors. No pre-approval. No source code to submit.

FTMO permits Expert Advisors on challenge, verification and funded accounts with no approval step. What it limits instead is activity: 200 orders open at one time, 2,000 positions per day, and a hyperactivity threshold at more than 2,000 server requests per day. Separately, total capital across your accounts is capped at $400,000 per trader or per strategy, and identical strategies detected across accounts can be suspended.

Two of those catch automated traders regularly. Neither is the one people worry about.

What counts as a server request?

An instruction your terminal sends to FTMO's trading server. FTMO's forbidden trading practices page defines the threshold as an account becoming hyperactive through "more than 2,000 server requests per day", and the same clause spells out what generates them: trades and pending orders being opened, modified, or closed.

Modified is the word doing the work, and it's the reason this limit matters more than the position count.

Opening and closing a trade is two requests. Every adjustment in between is another one. Moving a stop loss is a request. Moving a take profit is a request. Updating a pending order's price is a request. FTMO's rules page names stop and target updates explicitly, so there's no room to read the limit as covering entries only.

Now do the arithmetic that nobody publishes.

A strategy that opens a position, adjusts its stop twenty times as the trade develops, and closes it, has sent twenty-two requests for one trade. Ninety trades like that in a day and the account is at the threshold. Run three strategies on the same account and you're dividing that budget three ways.

For most people this is a non-issue. A swing strategy taking four trades a week with a fixed stop sends single digits per day, and the cap is irrelevant. It becomes real in exactly one pattern: anything that moves a stop repeatedly as price moves. That mechanism is common, it's often sold as a safety feature, and the request cost of it is disclosed by nobody.

If you're running something that adjusts stops continuously, count what it actually sends before you scale up position frequency. The number is knowable in advance and it's much easier to check than to explain afterwards.

What doesn't count

This is where most automated setups are quieter than people assume.

Watching your account costs nothing at the broker. Balance, equity, open positions and current prices all sit in your terminal's own memory, kept current by the broker's price feed. Anything reading those values is reading locally. It never leaves your machine, and it doesn't matter how often it happens.

Price streaming isn't a request either. The broker sends prices for the symbols you have open, continuously, as a subscription. That's identical whether an automated system is watching the chart or you are.

So the broker only hears from you when a trade actually happens. An idle strategy waiting for a signal sends nothing. A strategy that fires twice a day sends a handful of requests. The cap is about activity, not about being connected.

One caveat worth stating rather than glossing: a system watching for candle closes has to check the clock against its data, and while that's normally answered from what's already on the machine, whether a terminal occasionally refreshes that from the broker isn't something I've measured. I'd rather say that than publish a number I invented. It doesn't change the practical picture, because the volume that reaches 2,000 comes from order modifications, not from watching.

The other two caps

The 200 concurrent limit is the one grid and ladder strategies should look at. A grid placing pending orders at intervals accumulates them, and a wide grid on several symbols reaches three figures faster than it feels like it should. If the tool you're using lets you cap open positions, set that cap below 200 with room to spare rather than discovering the ceiling during a trending session.

FTMO's stated response to hyperactivity on its FAQ is that it may alert you and ask you to adjust your EA's logic or parameters. The forbidden practices page frames the same threshold more severely. Those two pages don't read identically, which is worth knowing: don't treat the softer wording as the operative one, and don't design a strategy that sits near the line.

The $400,000 rule, and why third-party EAs trip it

This one has nothing to do with server load and it's the one that costs people funded accounts.

FTMO caps total capital at $400,000 per trader or per strategy, across all your accounts at once, before any scaling, applying to 1-Step and 2-Step together. Equivalent figures exist in other currencies. If identically traded strategies are detected across multiple accounts and the combined capital passes the limit, FTMO states it reserves the right to suspend the accounts involved.

Read the "or per strategy" carefully, because it's not about you.

The clause works across traders, not just across your own accounts. FTMO's own guidance on third-party EAs makes the point directly: if you're running an EA that other traders are also running, you're all trading the same strategy, and the combined allocation is what's measured. You can be well inside your personal limit and still be part of an aggregate that isn't.

That's the actual risk in buying an EA off a marketplace and putting it on a funded account. Not that FTMO bans it. That a few hundred other buyers did the same thing, and none of you can see the total.

One timing detail that gets missed: the allocation cap bites at the funded stage. FTMO places no limit on how many challenge and verification accounts you hold. So the strategy that carried you through three challenges can be the reason the fourth funded account doesn't get issued.

Does building your own logic change anything?

For this rule, yes, and it's the practical takeaway.

What FTMO measures is whether the strategy being traded is identical across accounts. Not what software produced it. A strategy you've designed — your own conditions, your own sizing, your own rules — isn't running on anyone else's account, so there's no aggregate to fall foul of.

The same logic applies in reverse to anything pre-built. A template used exactly as supplied, by a number of people who all found it in the same place, is in the same category as a marketplace EA for the purposes of this clause, regardless of what platform it came from. If you start from a template, change it into something that's yours.

Nothing here means FTMO has approved, classified or cleared any particular tool. Firms don't do that, and any page suggesting otherwise is selling you something.

What's not restricted, and is worth knowing

Two things people assume are prohibited and aren't.

Copying your own strategy across your own FTMO accounts is permitted, subject to the allocation cap. What isn't permitted is holding accounts through separate registrations to get around the cap, or having someone else manage your account. The line is between your accounts and other people's.

VPS use is generally fine. Automation has to run somewhere, and FTMO's rules accommodate that.

The restrictions that do exist are behavioural: exploiting price feed errors or delays, holding opposing positions across accounts to distribute profit artificially, and trading that resembles gambling rather than a strategy. Those are about what the strategy does. None of them is about it being automated.

The honest limitation

You can't verify the $400,000 aggregate before it affects you.

If you're running your own logic, this is theoretical. If you're running anything pre-built, there's no way to find out how many other accounts are trading it, and no disclosure that would tell you. You find out when an account gets suspended or a funded account isn't issued, and by then the challenge fee is spent.

There's no clean fix. What reduces the exposure: build or meaningfully modify what you run, don't put the same untouched strategy on several accounts, and treat any EA marketed as widely used as carrying a cost that isn't in its price.

What to do before your next challenge

Count what your setup actually sends. Opens, closes, and every stop or target adjustment. If you're near four figures a day, you're closer to the threshold than the marketing suggested.

Cap your concurrent orders below 200 deliberately, especially for anything that ladders into a position.

Check what you're running is yours. If it came pre-built and you haven't changed it, assume other people are running it too.

Read FTMO's own pages rather than a summary. Their FAQ and forbidden practices pages are public, they're the operative documents, and the rules on them change. This article was checked against both on 2 September 2026.

Vectorhelm lets you build your own strategy logic and connect an MT5 account without writing code, with limits you set at strategy and account level. It isn't open yet. Early access is here if you want it when it is.

Vectorhelm is execution software. It runs the rules you define on accounts you own and control. Nothing here is trading advice, and no configuration of risk limits prevents losses. Prop firm rules change; verify against the firm's own published terms before relying on anything here.