How Prop Firms Detect Trade Copying (& How to Stay Compliant)

Can prop firms detect trade copying? Absolutely. "My funded account was terminated for group trading but I trade alone." Posts like this appear weekly in r/PropFirmTrading. The trader usually used a cloud trade copier to manage multiple accounts and tripped the firm's detection systems.

Understanding how prop firms detect trade copying is essential for any trader managing multiple accounts. This is not a guide to cheat — it is a guide to stay compliant while legitimately automating your own strategy.

Prop Firm Detection Method 1: Trade Correlation Analysis

This is the primary detection method. Firms analyze every trade across their entire user base and flag accounts with high correlation.

What They Measure

• Entry time correlation: Do your accounts consistently enter trades within 100ms of each other?
• Symbol correlation: Do your accounts trade the same pairs in the same sequence?
• Direction correlation: Are the directions identical (forward copying) or perfectly opposite (reverse copying)?
• Lot size ratios: Are lot sizes between accounts always in the same ratio (e.g., always 2:1)?
• Exit timing: Do positions close within seconds of each other?

The Math Behind Detection

Firms calculate Pearson correlation coefficients across your trade data. A correlation above 0.85 (positive or negative) triggers manual review. Above 0.95 is an automatic flag. For reference:

Detection Method 2: IP and Infrastructure Analysis

Prop firms log the IP address of every connection to their MT5 servers. They flag:

• Same IP, multiple accounts: 5 accounts all connecting from the same VPS IP
• Same IP subnet: All accounts on the same cloud provider's /24 subnet
• Known copier IPs: Some firms maintain lists of cloud copier service IP ranges
• VPN/proxy detection: Unusual latency patterns that suggest routing through intermediaries

Detection Method 3: Behavioral Pattern Analysis

Human traders have natural variation in their execution. Automated copy trading produces unnaturally consistent patterns:

• Identical order comments/magic numbers across accounts
• Same EA names appearing on multiple accounts
• Execution microsecond patterns — trades that arrive with the exact same millisecond offset
• Login-to-trade timing — accounts that connect at the same time and trade at the same time, every day

Detection Method 4: Third-Party Services

Major prop firms use specialized detection services:

• MetaFynd: Cross-firm trade correlation database
• TradersLab: Behavioral analysis and copy trading detection
• Custom analytics: Large firms like FTMO have proprietary in-house systems

These services compare your trades not just against other accounts at your firm, but across multiple firms. If your FTMO account and your FundedNext account show high correlation, both firms will know.

How to Stay Compliant While Automating

The key insight: prop firms ban copy trading (copying someone else's signals), not algorithmic trading (automating your own strategy). Here is how to automate legitimately:

1. Use a Local EA, Not a Cloud Copier

A local EA runs on each MT5 terminal independently. There is no shared cloud infrastructure, no correlated IP addresses from copier servers, and execution timestamps vary naturally because each terminal processes the webhook independently.

2. Distribute Accounts Across VPS Instances

Don't run 10 accounts on the same VPS. Cluster 3-4 accounts per VPS, using different providers if possible. This creates IP diversity that matches normal trader behavior.

3. Use Slight Lot Size Variations

Instead of always using exactly 0.5 lots, consider using risk-based lot sizing where each account calculates lot size independently based on its own equity. This produces natural lot size variation between accounts.

4. Avoid Identical EA Names and Comments

Set unique EA names and trade comments per account. Instead of "TradingView_Signal" on all 5 accounts, use unique identifiers.

5. Understand Your Firm's Specific Rules

Each firm has different policies:

Cloud Copier vs Local EA: Detection Risk