Most prohibited strategies at least require the market to cooperate. Cross-account abuse doesn't. The scheme is pure arithmetic: buy ten challenges, put half the accounts long and half short on the same instrument, and let variance do the rest. The losers cost a challenge fee each; a winner becomes a funded account extracted for payouts. The trader never had an edge — they had volume and a spreadsheet.
It's the silent killer because every individual account looks clean. Reasonable position sizes, sensible stops, a plausible equity curve. The fraud only exists between accounts — in the correlation of entries, exits and directions across your book — which is precisely where a human reviewer, staring at one account at a time, will never look.
The Three Flavours of Cross-Account Abuse
- Inverse hedging across accounts. The classic. Opposite positions on the same symbol across multiple evaluation accounts — often under different names, funded by the same operator. Guaranteed pass rate on a percentage of accounts, with the fees on losers treated as cost of goods.
- Mirror copying into evaluations. A signal — from a paid group, a funded account elsewhere, or a bot — is copied identically into dozens of purchased accounts. If the signal has a hot streak, every follower passes simultaneously with identical histories, and your firm funds one strategy pretending to be fifty traders.
- Pass-for-hire services. A third party trades many customers' evaluations at once, inevitably producing near-identical trade sequences across unrelated accounts. The customer never demonstrates skill; the account's history belongs to someone else entirely.
All three leave the same forensic residue: trade histories that agree with each other far too much — same symbols, same timestamps within seconds, same or precisely inverted directions, proportional sizes.
Why This Fraud Is Uniquely Expensive
A martingale or grid account fails on its own eventually; the strategy self-destructs. Cross-account schemes never self-destruct — they are designed so that some account always wins. The expected cost scales with your own success: the more challenges you sell, the more raw material the scheme has. And the damage lands at the worst point in the funnel, the funded stage, where payouts are real money and clawbacks are reputationally expensive. Firms that discover the pattern only at payout review face an ugly choice — pay a fraudulent account or fight a public dispute. The economics of failure here are covered more broadly in why prop firms fail; cross-account abuse is one of the recurring causes.
The uncomfortable math: with inverse hedging across N accounts, roughly half survive any directional move. The abuser's cost is fixed challenge fees; the upside is your funded capital. Only detection — not drawdown rules, not profit targets — breaks the equation.
Detection: Reading Patterns Across the Whole Book
Because the scheme is invisible within any single account, detection has to run at the level of the firm's entire trade history. The Singuard Prop Firm CRM ships cross-account and copy-trading pattern detection as part of its prohibited-strategies library: the engine ingests every account's positions and closed trades — synced from the platform every 500 milliseconds — and flags accounts whose histories correlate in ways honest, independent traders' never would:
- Mirrored entries — the same instrument traded in the same direction at nearly the same moment across accounts, repeatedly.
- Inverse pairs — systematically opposite positions on the same symbols at overlapping times, the signature of hedged group-passing.
- Repetition over time — one coincidence means nothing; the engine keys on sustained agreement across many trades, which is statistically damning.
Flagged clusters surface for review rather than auto-failing by default — correlation cases deserve a human decision — and you choose the consequence per rule, from flag to suspend. Every automated flag and every staff decision is written to the permanent audit log, so when you deny a payout on a mirrored cluster, the evidence trail is already assembled: which accounts, which trades, which timestamps.
Closing the Loop with Identity and Payout Controls
Pattern detection is the core, but the surrounding controls make the scheme uneconomic end to end:
- KYC before money moves. Automated identity verification via Sumsub, Onfido, Veriff or manual review makes multi-account operation under fake identities materially harder — see identity fraud in prop firms. No verified identity, no payout, ever.
- Payout gates. Minimum trading days, minimum trades and payout cycles checked before a withdrawal request is even allowed give the pattern detector time to accumulate evidence before real money is at stake.
- The account analyzer. When a cluster is flagged, staff open each account's equity curve, per-rule pass/fail status and trade breakdown side by side in the account analyzer — turning an investigation from hours of CSV archaeology into minutes of review.
- Legitimate copytrading, ring-fenced. Note the distinction: eTrader offers internal copytrading as a broker product — approved providers, disclosed following, configured rules. Abuse detection targets covert copying into evaluations, where the entire point is that the firm doesn't know.
What to Publish in Your Rules
State plainly in your challenge terms that cross-account correlation, group passing, account sharing and third-party trading are prohibited, and that trade histories are analysed across accounts. This deters casual abusers before purchase, and — because Singuard's engine documents exactly what it detected — gives you contractual and evidentiary footing for the determined ones. Honest traders lose nothing; the only people inconvenienced are the ones running twenty accounts.
"Copy-trading abuse is invisible account by account and obvious across the book. That's why detection has to live at the platform level."
— Alex Onta, Executive Director, eTrader & Prop Firm CRM
Key Takeaways
- Cross-account abuse is arithmetic, not trading: hedged or mirrored accounts guarantee some passes regardless of skill.
- Each account looks clean alone — detection must correlate trade histories across the whole book, which only software can do continuously.
- Singuard flags mirrored and inverse patterns automatically from 500ms-synced history, with consequences you choose and a full audit trail.
- KYC gates, payout rules and the account analyzer close the loop, making the scheme uneconomic before real money moves.
Frequently Asked Questions
How Can a Firm Tell Copying from Coincidence?
Repetition. Two traders taking the same EURUSD breakout once is a market; the same accounts agreeing on symbol, direction, timing and proportional size across many trades is statistically implausible. The engine flags sustained correlation, and staff make the final call with the evidence in front of them.
Does This Ban Legitimate Copytrading?
No. Broker-side copytrading — like eTrader's internal copytrading, with approved signal providers and configured allocations — is a disclosed product. What's detected here is covert copying into evaluation accounts designed to defeat the assessment itself.
What Should Happen When a Cluster Is Flagged?
Most firms route flags to human review: open the accounts in the analyzer, compare histories, then apply the published consequence. Every step — the automated flag and the staff decision — is recorded in the audit log, which is exactly the trail you want in a dispute or chargeback. See it live at demo-prop.sghk.org.