Market Integrity by Design: Governance Intelligence in Malaysia

This study examines whether governance quality, ownership concentration, ESG disclosure and trading-information anomalies are associated with market integrity risk among non-financial publicly listed companies in Malaysia. Using a balanced panel of 40 Bursa Malaysia companies from 2016 to 2025 (400 firm-year observations), the paper separates explanatory association from predictive early warning through firm and year fixed effects with firm-clustered standard errors, dynamic sensitivity analysis, upper-tail quantile regression, machine-learning classification and illustrative conditional 2026-2030 scenario projections. Board independence and ESG disclosure are consistently associated with lower market integrity risk, whereas ownership concentration, volatility and digital trading intensity are associated with higher risk. Upper-tail evidence indicates that concentrated ownership is especially consequential when board independence is weak. In the fitted random-forest model, ownership concentration produces the largest permutation-importance signal for suspicious market-integrity classification. The study interprets these layers according to distinct evidential roles: regression estimates describe conditional associations, machine learning provides early-warning discrimination, and scenario paths illustrate sensitivity to alternative governance and market conditions. The paper contributes a four-hypothesis governance-intelligence framework linking governance, disclosure and market signals to risk-based surveillance, board oversight and stakeholder value without treating statistical anomalies as proof of misconduct.