Learn to Rank Risky Investors: A Case Study of Predicting Retail Traders' Behaviour and Profitability

2025-09-20

AI Summary

To address the challenge of real-time identification of high-return–high-risk retail traders by financial market makers, this paper proposes a profit-aware learning-to-rank (LTR) framework. Methodologically, we design a Profit-and-Loss–Augmented Binary Cross-Entropy (PA-BCE) loss function to explicitly model P&L signals and introduce a self-cross-trader attention mechanism that jointly encodes individual trading behavior and inter-trader relational features, all implemented within a Transformer-based LTR architecture. Our key contribution is the first integration of P&L sensitivity into the ranking objective—overcoming the limitation of conventional risk-control models that neglect return-oriented metrics. Experiments demonstrate that our approach outperforms state-of-the-art baselines (e.g., Rankformer) by +8.4% in F1-score and achieves 10–17% higher average profitability, significantly improving both the precision of high-risk trader identification and the accuracy of ex-ante profit prediction.

Technology Category

• Machine Learning: Learning Preferences or Rankings
• Search and Optimization: Learning to Search
• Multiagent Systems: Adversarial Agents

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking
• Economics, Online Markets and Human Computation: Trust and reliance of crowd workers and data experts on GenAI

Abstract

Identifying risky traders with high profits in financial markets is crucial for market makers, such as trading exchanges, to ensure effective risk management through real-time decisions on regulation compliance and hedging. However, capturing the complex and dynamic behaviours of individual traders poses significant challenges. Traditional classification and anomaly detection methods often establish a fixed risk boundary, failing to account for this complexity and dynamism. To tackle this issue, we propose a profit-aware risk ranker (PA-RiskRanker) that reframes the problem of identifying risky traders as a ranking task using Learning-to-Rank (LETOR) algorithms. Our approach features a Profit-Aware binary cross entropy (PA-BCE) loss function and a transformer-based ranker enhanced with a self-cross-trader attention pipeline. These components effectively integrate profit and loss (P&L) considerations into the training process while capturing intra- and inter-trader relationships. Our research critically examines the limitations of existing deep learning-based LETOR algorithms in trading risk management, which often overlook the importance of P&L in financial scenarios. By prioritising P&L, our method improves risky trader identification, achieving an 8.4% increase in F1 score compared to state-of-the-art (SOTA) ranking models like Rankformer. Additionally, it demonstrates a 10%-17% increase in average profit compared to all benchmark models.

Problem

• Identifying risky traders with high profits for financial risk management
• Capturing complex dynamic behaviors of individual traders effectively
• Overcoming limitations of traditional fixed-boundary risk detection methods

Innovation

• Learning-to-Rank algorithms for trader risk ranking
• Profit-Aware binary cross entropy loss function
• Transformer-based ranker with self-cross-trader attention