Self-referral is the most common fraud type in prop trading affiliate programs. The mechanic is simple: an affiliate signs up as a partner, generates a tracking link or coupon code, then uses it to purchase challenges for themselves. At a $50 CPA on a $200 challenge, the affiliate effectively gets a 25% personal discount funded by the prop firm commission budget. Some affiliates extend this to friends, family, and trading communities where they coordinate group purchases.
Same payment method used for affiliate payout and challenge purchase
Payment fingerprint matching
Very high
Affiliate IP address matches customer IP on purchase
IP correlation analysis
Medium (VPNs create false negatives)
Single affiliate drives 3+ purchases from same device fingerprint
Device fingerprinting on checkout
High
All referred customers purchase within minutes of each other
Timestamp clustering analysis
Medium-high
Affiliate has zero content or social presence but generates steady conversions
Content audit vs. conversion ratio
High
Fake Account Clusters
A more sophisticated version of self-referral involves creating multiple customer accounts to generate fake challenge purchases. The affiliate creates 10 accounts with different email addresses, uses a VPN to vary IP addresses, and purchases the cheapest available challenge tier on each account. With a $30 CPA per purchase and a $100 challenge price, the affiliate spends $1,000 on challenges and earns $300 in commissions -- a net loss. But if the affiliate also disputes the charges with their bank, they recover the $1,000 while keeping the $300 commission.
Fake account clusters combined with chargeback fraud create a double loss: the prop firm pays the CPA commission and loses the challenge revenue to chargebacks. This combination is the highest-damage fraud pattern in prop trading affiliate programs.
Detection Through Purchase Behavior
Legitimate prop trading customers exhibit predictable behavior patterns. They research challenge rules, compare account sizes, and typically purchase a challenge that matches their trading experience. Self-referral and fake account purchases often deviate from these patterns.
Legitimate customers browse the site for 5-15 minutes before purchasing -- fake accounts go directly to checkout
Real traders often select mid-tier or high-tier challenges ($50K-$200K) -- fraud clusters concentrate on the cheapest tier
Genuine customers create accounts days or weeks before purchasing -- manufactured accounts register and purchase in the same session
Organic customers use diverse payment methods -- fake clusters often share BIN ranges or payment processors
Real traders attempt the challenge and generate trading activity -- fake purchases show zero platform logins after purchase
Prevention Controls
The most effective control against self-referral is a clear policy combined with technical enforcement. Prohibit self-referral in the affiliate agreement, then back it up with automated detection. Configure your affiliate platform to flag purchases where the affiliate and customer share identifiers -- email domain, payment method, IP address, or device fingerprint. Set up automatic hold on commissions when flags are triggered, pending manual review.
Add a self-referral prohibition clause to the partner agreement with specific consequences (commission clawback, account termination)
Implement cross-matching between affiliate registration data and customer purchase data
Require a minimum number of unique referred customers (e.g., 5 unique buyers) before first payout
Apply a 14-day hold period on commissions for new affiliates to allow fraud patterns to surface
Monitor post-purchase behavior -- flag affiliates whose referred customers never log into the trading platform
Key Takeaways
Self-referral is the most common prop trading affiliate fraud -- affiliates use their own codes to get effective discounts on challenges
Fake account clusters combined with chargebacks create double losses for prop firms
Detect self-referral by cross-matching affiliate and customer identifiers: email domain, payment method, IP, device fingerprint
Legitimate customers browse before buying, choose varied challenge tiers, and log into the trading platform -- fake accounts skip all three
Enforce a minimum unique-buyer threshold and hold period before first payout to surface patterns early