How to Comply With Prop Firm Rules for Automated Trading

How to Comply With Prop Firm Rules for Automated Trading: 2026 Guide

TLDR

A firm allowing Expert Advisors is not a blank check to run any code you want — it is permission to use the technology, subject to a separate list of banned strategies: high-frequency trading, latency arbitrage, tick scalping under a minimum holding time, grid and Martingale position sizing, and copy trading between accounts in the same challenge batch. Firms detect violations through a mix of automated triggers (request volume caps, trade-duration analysis, entry-timing patterns around news) and manual account review, so compliance has to be built into the EA's code itself — a news-window filter, a minimum trade duration, a request-rate limiter — not left to the trader remembering the rules while the bot runs unattended.

What "Automated Trading Compliance" Means for an EA Specifically

Compliance for automated trading has two separate layers that are easy to conflate. The first is account-level risk compliance — daily loss limits, maximum drawdown — which applies identically whether a human or a bot is placing the trades. The second, specific to EAs, is strategy-level compliance: a list of trading techniques and behaviors that are banned outright, regardless of whether the account is within its risk limits.

This second layer is the one that catches automated traders off guard, because an EA can be perfectly profitable and perfectly within its daily loss and drawdown limits while still running a banned strategy — and get the account closed for that reason alone.

For risk-limit compliance specifically, see Viprasol's guide to ensuring prop firm compliance during live trading. This guide covers the strategy-level rules that apply specifically to bots.

Why EA Compliance Is Different From Manual Trading Compliance

A bot can violate a rule thousands of times before anyone notices. A manual trader who accidentally scalps a few trades under the minimum holding time might do it a handful of times a week. An EA with a bug in its exit logic can repeat the same violation on every single trade, all day, unattended — turning a small compliance gap into a clear, unmissable pattern the moment a firm reviews the account.

"EAs allowed" is permission for the technology, not the strategy. Nearly every major firm now permits automated trading in some form. That approval covers running code on the platform — it says nothing about whether the specific strategy that code executes is compliant. Grid trading, Martingale sizing, and latency-based strategies remain banned at most firms regardless of the platform-level EA policy.

Detection is partly automated and looks specifically for bot-shaped patterns. Consistent sub-second to sub-minute trade durations, unusually high win rates on very short holds, and abnormal server request volumes are patterns that are far more characteristic of automated trading than manual trading — which means EAs are, in practice, watched more closely for exactly these signatures.

Strategies Prop Firms Actually Ban

Strategy	Why It's Banned	Typical Detection Signal
Latency arbitrage	Exploits price-feed delay, not trading skill	High win rate on very short holds, entries timed just before favorable moves
HFT / tick scalping	Below the firm's minimum holding time	Average trade duration under a firm-specific threshold, often under a minute
Grid trading	Masks risk until one adverse move blows the account	Manual review, flagged rather than auto-detected
Martingale sizing	Position size grows after losses, hides true risk	Drawdown pattern review, escalating lot sizes after losses
Copy trading across same-batch accounts	Circumvents per-account risk and capacity limits	Correlated trade timing and lot sizes across accounts

Per FXNX's breakdown of prop firm EA rules, latency arbitrage is specifically banned because it exploits delays between an EA's faster price feed and the prop firm's slower feed to predict price moves milliseconds ahead — not because the technique is difficult to detect, but because firms classify it as exploiting infrastructure rather than demonstrating trading skill.

How Firms Detect Non-Compliant EAs

Detection happens on two tracks running in parallel: automated platform triggers, and manual account review.

Automated triggers. Per an analysis of prop firm EA detection methods, accounts that trigger over roughly 2,000 server requests in a day face automatic termination through platform-level detection — no human review involved. This alone is a reason to keep an EA's polling frequency and order-check loops reasonable rather than hammering the server on every tick.

Trade-duration and entry-timing analysis. The same analysis notes that a consistent pattern of sub-minute trade durations combined with an unusually high win rate is a specific signature reviewers look for, since it suggests the EA's edge comes from execution speed rather than a genuine market read. Reviewers also examine entry timing relative to price movement — entries that consistently land just before a favorable move look like latency exploitation even if the EA's logic is not explicitly designed that way.

News-window violations. FTMO's rule, as one of the most widely referenced examples, blocks new trades or trade closes within 2 minutes before or after a scheduled high-impact news release. An EA that does not have this window hard-coded will violate it mechanically, every single time a news event falls inside its trading hours — this is exactly the kind of rule a bot breaks far more consistently than a human would.

Statistical Detection: Beyond Simple Threshold Triggers

Server request caps and minimum-hold floors are hard thresholds, but the more sophisticated risk teams have moved to statistical profiling on top of them. Rather than flagging a single trade that crosses a fixed line, these systems build a behavioral profile of each account and compare it against a population model built from thousands of other accounts — a strategy can be flagged for looking statistically unusual even if no single trade violates an explicit rule.

Temporal correlation analysis is described as the strongest signal specifically for latency arbitrage: detection systems check for a statistically significant correlation between the moment a fast external price feed diverges from the broker's own quote and the moment an order arrives, with genuine arbitrage activity clustering orders at very small time deltas after that divergence — a pattern normal discretionary or even normal algorithmic trading does not produce. Cross-account clustering is the other major layer: firms compare trade timing, entries, and exits across many accounts simultaneously, which means a rule-compliant strategy can still get flagged if its timing pattern statistically resembles other accounts already identified as violating the rules, purely by pattern-matching rather than any rule the account itself broke.

The practical takeaway for EA design is that "technically compliant on every individual trade" is necessary but not sufficient. An EA whose aggregate timing behavior statistically resembles a banned pattern — even for legitimate reasons, such as a mean-reversion strategy that happens to enter shortly after volatility spikes — needs documented reasoning ready, because population-level statistical review will not always distinguish intent from a technically similar automated fingerprint.

News Trading and Weekend Holding Rules

These rules vary meaningfully by firm and by account tier, so treat any specific numbers as a starting point to verify against your firm's current terms, not a universal standard.

Rule	Typical Restriction	What an EA Needs to Enforce
News trading window	No open/close within ~2 minutes of high-impact news	Economic calendar feed, hard entry/exit block during the window
Weekend holding	Prohibited at some firms, allowed at others	Forced flat-close before Friday market close if the firm requires it
Minimum holding time	Often around 1 minute, firm-specific	A minimum-hold floor in the exit logic itself, not just intended behavior

Rules also genuinely differ between firms rather than following one universal standard — treating "prop firm rules" as a single fixed list is itself a common source of accidental violations. FundedElite's 2026 rules, for example, explicitly permit news trading while still banning HFT — the opposite emphasis from firms like FTMO that restrict the news window tightly but are comparatively more permissive on trade duration for non-HFT strategies. An EA built against one firm's rule set is not automatically compliant on another, even when both firms broadly "allow EAs."

Practical Example: How a Review Actually Unfolds

A concrete scenario makes the detection mechanics less abstract. Consider an EA trading a mean-reversion scalp on a 1-minute chart, with an average holding time of 45 seconds and a 68% win rate:

Signal Observed	What It Suggests to a Reviewer	What Would Clear It
45-second average hold, under the firm's 1-minute threshold	Possible HFT / scalping violation	Documented mean-reversion logic showing the exit is signal-driven, not time-driven
68% win rate on short holds	Possible latency exploitation	Entry timestamps showing entries do not consistently precede favorable feed updates
No trades during the 2-minute news window, every time	Compliant behavior — a positive signal	Nothing further needed; this is what a hard-coded filter looks like from the reviewer's side

The account in this scenario is not automatically closed — the pattern is flagged, and what determines the outcome is whether the trader can produce a documented explanation quickly. A trader with entry/exit logic written down in plain language, and clean news-window compliance already visible in the trade log, clears this kind of review in minutes. A trader with no documentation and no visible news-window discipline gives the reviewer nothing to distinguish the account from an actual violation.

Building Compliance Directly Into the EA

Manually remembering these rules does not work for an unattended system — the rules have to be enforced in code, the same way risk limits should be. A compliance-aware EA needs at minimum:

An economic calendar integration that blocks new entries and exits inside the firm's news window, not a manually maintained list of dates that goes stale.
A minimum trade duration floor enforced in the exit logic, so the strategy structurally cannot close a trade faster than the firm's threshold — not just a strategy that "shouldn't" trade that fast.
Request-rate throttling on any polling loop, quote checks, or order-status queries, kept well under the volume that triggers automated termination.
A weekend flat-close routine that runs automatically before Friday close if the firm prohibits weekend holding, rather than depending on the trader remembering to close manually.
Position-sizing logic that structurally cannot Martingale — size calculated fresh from account risk parameters on every trade, never derived from the previous trade's loss.

If you want this engineered properly rather than bolted onto a retail EA template, hire an MQL5 developer who builds specifically for prop firm rule sets, or see Viprasol's ranked comparison of prop firm compliant EA developers.

EA Verification and Source Code Ownership

Beyond strategy rules, most firms require you to demonstrate the EA is genuinely yours, or that you understand its logic well enough to explain it — a compiled binary alone is generally not accepted as proof of ownership during a review. Keeping entry and exit logic documented in plain language, alongside the code itself, is what lets you respond quickly if a firm's review team asks how the strategy works. This documentation is also what separates a defensible review conversation from an account closure when a firm flags a pattern that has a legitimate explanation.

Step-by-Step Compliance Checklist Before Going Live

Confirm the firm explicitly permits Expert Advisors for your specific account type — not all tiers at a given firm allow the same automation.
Get the firm's current banned-strategy list in writing and check it against your EA's actual logic, not just its intended behavior.
Hard-code the news window and minimum holding time rather than relying on the strategy's natural behavior to stay compliant.
Audit the EA's request volume against a full trading day before going live, to confirm it stays well under any automated termination threshold.
Document the entry and exit logic in plain language, kept alongside the code, ready to produce if a review is triggered.
Forward-test on a demo account first, watching specifically for accidental rule violations — a minimum-hold breach or a news-window trade — not just profitability.
Review the strategy's aggregate timing profile, not just individual trades, against the kind of statistical clustering patterns firms flag — an occasional fast trade is different from a consistent timing signature that resembles a banned pattern.
Confirm which regulatory layer actually applies — a prop firm's own rules for offshore forex/CFD trading, SEBI's Algo-ID framework for any separate Indian-exchange automated strategy, or both if running different strategies on different platforms.

India-Specific Considerations: SEBI's Algo Framework vs Prop Firm Rules

These are two separate regulatory layers that Indian traders sometimes conflate, and it is worth being precise about which one actually applies to a prop firm EA. Prop firm rules (news windows, minimum holding time, banned strategies) are the firm's own contractual terms, enforced by the firm regardless of the trader's location. SEBI's algorithmic trading framework is an Indian securities-market regulation that applies to orders placed through a SEBI-registered broker on Indian exchanges — it does not govern a forex EA running on an offshore prop firm's MT4/MT5 servers, since that activity is not routed through an Indian broker or exchange at all.

Where SEBI's framework does become directly relevant is if the same trader also runs algorithmic strategies on Indian equities or derivatives through a domestic broker's API. From April 1, 2026, every order placed by a registered algorithm must carry a unique exchange-issued Algo-ID, API access is restricted to a static IP address registered with the broker, and strategies that place 10 or more orders per second (measured per exchange, per second) cross a threshold requiring formal registration through the broker before the strategy can keep running. Brokers carry full responsibility for every algorithm operating through their platform under this framework, which is a meaningfully different accountability model from a prop firm simply banning a strategy in its terms of service.

The practical rule of thumb: a prop firm EA trading forex/CFDs through an offshore broker is governed entirely by that firm's rules, not SEBI's Algo-ID regime. A trader who also runs automated strategies on NSE/BSE instruments through an Indian broker needs to separately confirm that activity against SEBI's April 2026 requirements — the two compliance obligations do not overlap, but both can apply to the same person running different strategies on different platforms.

Common Mistakes When Building Compliant Automated Trading

Assuming "EAs allowed" means the strategy is allowed. Platform-level permission and strategy-level compliance are two separate approvals. Confirm both.

Treating the news filter as optional because the strategy "rarely" trades near news. Rarely is not never, and a bot running unattended will eventually hit the exception if it is not blocked in code.

No minimum-hold enforcement in the exit logic. A strategy that is not intentionally scalping can still occasionally close trades under the minimum holding time purely from normal volatility — enforce the floor structurally, not by trusting the strategy's typical behavior.

Excessive polling frequency. Checking quotes or order status far more often than the strategy actually needs increases request volume for no benefit and moves the account closer to an automated termination threshold.

No documentation trail. An EA that works fine but has no plain-language explanation of its logic puts the trader in a weak position if a firm's review team flags an unusual pattern that has an innocent explanation.

Assuming individual-trade compliance is sufficient. Statistical and cross-account review can flag an account whose aggregate timing pattern resembles known violations even when no single trade breaks an explicit rule — design and document the strategy's logic with that population-level review in mind, not just per-trade rule checks.

Conflating SEBI's algo framework with prop firm rules. A forex EA on an offshore prop firm platform is not governed by SEBI's Algo-ID regime; assuming either that SEBI rules apply to it, or that prop firm rules satisfy SEBI requirements for a separate Indian-exchange algo strategy, are both compliance gaps worth closing explicitly.

Build vs Buy: When to Get a Developer

Use a reputable off-the-shelf EA with a documented compliance track record if you are testing one firm's rules with a standard strategy type and want to move quickly.

Get a custom-built or audited EA if you are running a strategy with genuinely short holding periods that could resemble banned patterns even when legitimate, running across multiple firms with different rule sets, or need documented, review-ready logic for funded account audits.

Related Glossary Terms

For more definitions, visit the AI and software glossary.

Latency Arbitrage: A banned strategy that exploits price-feed delay between a faster and slower broker feed to predict near-term price moves, rather than trading on genuine market signal.

Grid Trading: Placing a series of orders at fixed price intervals in both directions; flagged by most prop firms because it can mask true risk until a single adverse move triggers a large loss.

Martingale: A position-sizing method that increases size after a loss to recover it on the next win; banned or flagged at most firms because it hides risk exponentially.

Minimum Holding Time: The shortest duration a trade must remain open to avoid being classified as prohibited scalping or HFT.

News Trading Window: A firm-defined period, commonly around 2 minutes, before and after high-impact economic releases during which new trades or closes are prohibited.

Algo-ID: A unique exchange-issued identifier that, under SEBI's April 2026 framework, must be attached to every order placed by an algorithm through a registered Indian broker, enabling trade-level audit and tracking.

Temporal Correlation Analysis: A statistical detection method that checks for correlation between the moment a fast price feed diverges from a broker's quote and the moment an order arrives, used to identify latency arbitrage patterns.

FAQ

Does "EAs allowed" mean any strategy is allowed?

No. Platform-level permission to run automated trading software is separate from strategy-level compliance. Banned strategies like latency arbitrage, HFT, grid trading, and Martingale sizing remain prohibited at most firms regardless of whether EAs are technically permitted.

How do prop firms detect banned EA strategies?

Through a mix of automated triggers — such as server request volume caps and trade-duration analysis — and manual account review that looks at entry timing, win rate on very short holds, and drawdown patterns consistent with grid or Martingale sizing.

What is the typical news trading restriction for EAs?

A commonly referenced example is FTMO's rule blocking new trades or trade closes within roughly 2 minutes before and after a scheduled high-impact news release. Exact windows vary by firm and should be confirmed against current terms.

Can weekend holding get my automated account closed?

At firms that prohibit weekend holding, yes — an EA needs a forced flat-close routine before Friday market close if the firm requires it, since this is not a rule that enforces itself.

Do I need to prove I own my EA's source code?

Most firms want evidence the strategy is genuinely yours, or that you can explain its logic in detail — a compiled binary alone is generally not sufficient proof during a review. Keeping documented entry and exit logic alongside the code is the practical way to satisfy this.

Does SEBI's algo trading framework apply to my prop firm EA?

Generally no, if the EA trades forex or CFDs through an offshore prop firm's MT4/MT5 servers — that activity is not routed through a SEBI-registered broker or an Indian exchange. SEBI's April 2026 framework, including the Algo-ID requirement and the 10-orders-per-second registration threshold, applies to algorithmic orders placed through Indian brokers on Indian exchanges, which is a separate compliance obligation from prop firm rules.

Can a technically rule-compliant EA still get flagged?

Yes. Statistical and cross-account detection compares an account's timing and entry patterns against a population model built from many accounts, which means a strategy that never explicitly breaks a stated rule can still be flagged if its aggregate behavior statistically resembles patterns already associated with violations. Documented, plain-language reasoning for the strategy's logic is what resolves this kind of review.

Should I hard-code compliance rules per firm, or build one general-purpose rule set?

Per-firm, not general-purpose. Rules genuinely differ — some firms restrict the news window tightly while being comparatively permissive on trade duration, others are the reverse — so a single hard-coded rule set copied across firms risks being either too restrictive (missing legitimate trading windows) or non-compliant (missing a rule the previous firm did not have). Treat each firm's current terms as the source of truth and configure the EA's compliance module per deployment rather than assuming portability.

Need an EA built with compliance enforced in code from day one, not bolted on after a rule violation? Book a free 30-minute consultation to review your target firm's rule set.

How to Comply With Prop Firm Rules for Automated Trading: 2026 Guide