DATA-DRIVEN COMPLIANCE FRAMEWORKS FOR ANTI-MONEY LAUNDERING (AML) AND TAX RISK MANAGEMENT IN FINANCIAL INSTITUTIONS

Abstract

This study addresses the problem that financial institutions often deploy fragmented governance, analytics, and automation controls across AML and tax functions, which can weaken evidence traceability and reduce compliance effectiveness. The purpose was to test whether a data-driven compliance capability framework predicts AML compliance effectiveness and tax risk management effectiveness within an enterprise financial institution case, using a quantitative, cross-sectional, case-based design. Purposive sampling targeted role-relevant staff across compliance, risk, audit, tax/finance control, operations, and IT or data governance support; 210 questionnaires were distributed, and 184 valid responses were retained (response rate 87.6%). Key variables included four independent dimensions, Data Governance and Data Quality (DGQ), Analytics Capability (AC), Automation and Integration Maturity (AIM), and Organizational Readiness (OR), plus two dependent outcomes, AML effectiveness (AML-EFF) and tax risk management effectiveness (TRM-EFF). The analysis plan applied descriptive statistics, reliability testing (Cronbach’s alpha), Pearson correlations, and multiple regression. Findings show moderately high overall maturity (DDCC index M = 3.74, SD = 0.61), with AML-EFF (M = 3.72, SD = 0.66) slightly higher than TRM-EFF (M = 3.63, SD = 0.69). Reliability was strong (α = 0.82–0.90). Correlations were positive and significant (e.g., AC with AML-EFF r = 0.64; DGQ with TRM-EFF r = 0.58; OR with TRM-EFF r = 0.60; all p < .001). Regression explained substantial variance (AML model R² = 0.53; tax model variance explained = 49%), with AC the strongest AML predictor (β = 0.33, p < .001) and OR the strongest tax predictor (β = 0.34, p < .001); AIM was significant for AML (β = 0.12, p = .048) but not for tax at 0.05 (β = 0.10, p = .061). Implications indicate institutions should prioritize governed data and readiness as foundations across both domains, intensify analytics capability for AML performance, and treat automation as value-adding only when paired with strong governance and execution discipline.