A Bayesian Regression Analysis of Corporate Governance and Financial Reporting Quality in South Africa

This study investigates the relationship between corporate governance mechanisms and financial reporting quality in South African listed firms using a Bayesian regression framework. Drawing on a sample of firms listed on the Johannesburg Stock Exchange (JSE) over the period 2015-2024, the study examines how board and audit committee characteristics, ownership structure and CEO-Chair duality relate to earnings quality, proxied by abnormal accruals estimated using the modified Jones model. The Bayesian approach permits the formal incorporation of theoretically motivated prior information regarding governance effects, the direct quantification of parameter uncertainty through full posterior distributions, and rigorous hypothesis testing via posterior probabilities. Posterior estimates are obtained via Markov Chain Monte Carlo (MCMC) sampling using four chains of 4,000 iterations each, with convergence assessed using the Gelman-Rubin diagnostic and effective sample size. The results indicate substantial negative associations between board independence, audit committee financial expertise and earnings quality, with credible intervals that exclude zero. Ownership concentration and the presence of a non-executive chair also exhibit statistically meaningful posterior associations with reduced abnormal accruals. Conversely, board size and CEO duality display posterior mass concentrated near zero. The findings provide robust inferential support for the monitoring role of independent directors and the disciplining effect of expert audit committees and underscore the methodological contribution of Bayesian inference for governance research in emerging markets. The study contributes to corporate governance scholarship by demonstrating how Bayesian regression formalises and extends classical frequentist analyses, yielding interpretable probabilistic statements that are directly informative for regulators, standard-setters and audit practitioners.