Explainable algorithmic trading with embedded audit and tax trails: a technology-enabled behavioral oversight model

Purpose This study aims to introduce a governance-oriented accounting information system (AIS) framework designed to detect and regulate addictive behavioral patterns within algorithmic trading environments. It explores how accounting logic, auditability and tax trail mechanisms can be embedded into automated trading systems to enhance transparency, compliance and investor well-being. Design/methodology/approach Using five years of S&P 500 open, high, low, close and volume data, the framework integrates XGBoost and long short-term memory models within an adaptive ensemble enhanced by meta-learning for volatility-sensitive thresholding. The AIS layer records algorithmic transactions through tax audit trails, enforces wash-sale disallowances, and ensures continuous financial reporting compliance. Explainable artificial intelligence (XAI; SHapley Additive exPlanations) provides interpretability, while behavioral archetype modeling identifies overconfident and addiction-prone trading patterns across market regimes. Findings The integrated model enhances risk-adjusted performance and interpretability while ensuring full auditability and accounting traceability. Behavioral simulations highlight that overconfident and momentum-driven profiles exhibit higher addictive tendencies and tax disallowance risks, emphasizing the need for accounting-based oversight in automated decision systems. Research limitations/implications The framework operationalizes responsible algorithmic trading by merging behavioral finance insights with accounting information infrastructure. It supports continuous auditing, regulatory alignment with frameworks such as FINRA, MiFID II and the EU AI Act, and offers a replicable model for integrating governance analytics and tax trails in financial automation. Originality/value To the best of the authors’ knowledge, this study is among the first to link addictive trading behavior with accounting-based audit systems. By embedding AIS, tax audit trails and XAI into algorithmic trading, it redefines the accounting function as an active safeguard for ethical, transparent and psychologically sustainable financial automation.