Why most EAs fail prop firm rules — and how to avoid it

An Expert Advisor is well-suited to the discipline a prop evaluation demands — fixed rules, no emotional sizing, hard stops on every order. And yet most EAs fail prop accounts. The reasons are predictable, almost mechanical, and most of them are fixable on the EA side once you know what to look for.

Here are the ten most common ways an Expert Advisor breaches a prop account, ranked by how often we see them. Read this list before you license your EA onto a funded account.

1. Retail sizing on a prop account

The single most common breach. A trader sets the EA's risk-per-trade to 1% — fine on a personal account, dangerous on prop. Three losing trades in a row puts the account within range of the daily-loss limit. Five losses runs into the overall floor. A 0.5% risk-per-trade setting halves the breach probability; 0.25% gets you to the survivable zone.

2. Holding through high-impact news

Most prop firms forbid open positions within ±2 to ±5 minutes of high-impact releases (NFP, FOMC, CPI, central bank decisions). The breach is automatic — even if the trade wins, the position-was-open evidence is in the trade log.

An EA without an event filter will, on average, blunder into a restricted window every 4–6 weeks. The fix is a news calendar feed plugged into the EA — see news-event filters for prop firm accounts.

3. Weekend / overnight holds

Many firms forbid holding any position past Friday close (server time) or through scheduled futures rollovers. An EA that opens a Friday afternoon scalp and holds into the close has just breached the rule — irrespective of P/L.

The fix is a Friday-close cutoff in the EA: stop opening new positions after a configured local time, and close existing positions at the rule deadline. Most reputable EAs ship this as a built-in toggle.

4. Daily-loss limit spike from a single trade

A position runs against you, the stop is in the trade plan, the stop is wide. By the time it triggers, the realised + unrealised loss has already breached the daily-loss limit. Most prop firms compute the daily limit on equity — meaning the unrealised P/L of an open trade can breach the limit before your stop ever fires.

5. Trailing-drawdown breach on a normal pullback

You pass the evaluation, the funded account is live, the EA prints two strong weeks. A normal post-run consolidation gives back 4% from peak — and on a trailing-DD account that is the breach point. The strategy never lost money relative to starting balance; it just gave back relative to peak.

A trailing-DD-aware EA does two things differently: it sizes positions off the trailing buffer rather than off starting equity, and it pulls position size down as the peak-to-current gap grows. The result is a wider safety margin in the moments when the floor is closest.

6. Martingale / averaging-down logic

Strategies that add to losers produce equity profiles that are fundamentally incompatible with prop drawdown rules. The peak-equity moment comes when the recovery closes — meaning the trailing floor is at its tightest exactly when the next averaging cycle begins. The variance crushes the rule.

If your EA uses any form of averaging, either disable the averaging on a prop account or do not use that EA on prop. There is no setting that makes martingale and trailing-DD coexist safely. Our own MartingaleHedge EA ships with hard kill-switches and is explicitly not recommended for trailing-DD prop accounts.

7. End-of-day swap / rollover bookings

Several firms compute the daily limit at 5pm New York time. An open position at that moment gets swap charged, which appears as a small balance debit on the next trading day. If you sized aggressively and ended the previous day at the edge of the limit, that rollover debit can be the breach you cannot see coming.

Fix: close all open positions 30–60 seconds before the firm's daily-reset time, or hold positions only on symbols with zero/positive swap. The EA needs to know the firm's reset clock — different from your local time.

8. Lot-size limits per position

Some firms cap the maximum lot size per position (e.g. 5 lots on a $100k account). An EA that sizes purely off risk-per-trade can land above this cap on a tight-stop opportunity, and the order gets rejected — or worse, partially filled — in ways that confuse downstream stop logic.

Set a hard maximum-lot input on the EA. Floor it to the firm's rule. The trade-off (sub-target sizing on tight stops) is small compared to the rejection / partial-fill chaos otherwise.

9. EA running on multiple accounts

A subtler rule. Some firms forbid running the same EA simultaneously on multiple funded accounts — they detect identical trade timestamps across accounts and treat it as "EA copy trading", which violates the funded-account terms.

If you must run the EA on multiple firms' accounts, stagger the timeframes (different M5/M15 setups) and the risk inputs so trade sequences are visibly distinct. Or simply do not — the rule exists because the firm needs to attribute risk per trader.

10. Strategy bans (EA-specific)

A small group of firms forbid specific strategy classes: HFT, latency arbitrage, grid systems, news straddles, tick-scalping. Some forbid EAs entirely (rare, but they exist — typically the most discretionary-focused firms). Read the prohibited strategies clause in the firm's contract before paying.

The fix that catches most of them

A prop-aware EA layer — sometimes shipped as a separate utility — sits in front of the strategy and enforces:

• Daily-loss tripwire on equity (not balance), at 80% of the firm's limit
• Trailing-DD watcher that closes positions before the floor is touched
• News-window blocker tied to a calendar feed
• End-of-day flatten at the firm's rollover clock
• Weekend flatten on Friday before close
• Hard maximum lot per position
• Maximum simultaneous open positions

These checks live above the strategy itself — they do not affect the EA's entry / exit logic, they just refuse to let it breach a rule. AlphaLab-AI ships this as Prop Guard, runnable alongside any other EA.

The bottom line

Most EA-driven prop breaches are not strategy failures — they are housekeeping failures. The maths of the strategy is fine; the EA just was not built with prop rules in mind. Run the EA in a prop-aware configuration, layer a rule-enforcement utility on top, and 80% of the common breaches go away before they happen.