Accrual Earnings Management and Firm Valuation: Distributional Evidence From JSE-Listed Firms

Abstract

This study evaluates whether investors on the Johannesburg Stock Exchange (JSE) incorporate accrual earnings management (AEM) into firm valuation decisions. Using a panel dataset of 171 non-financial firms and 1,881 firm-year observations over the period 2015-2025, discretionary accruals are estimated using the performance-matched Kothari (2005) model, while firm value is proxied by the natural logarithm of Tobin’s Q. The analysis employs both ordinary least squares (OLS) and quantile regression to capture average and distributional effects. The OLS results indicate a weak positive association between AEM and firm value. However, quantile regression reveals that this relationship is not uniform across the distribution of firm value. AEM is positively and significantly associated with firm value at the median (Q50) quantile, while remaining insignificant at the lower (Q25) and upper (Q75) quantiles. Wald tests provide evidence of differences between the lower and median quantiles, indicating conditional value relevance. Robustness analysis using the Modified Jones (Dechow) model yields no significant relationship, suggesting that AEM pricing is sensitive to measurement and driven by performance-adjusted accruals. The study contributes to the earnings management literature by demonstrating that the value relevance of AEM in emerging markets is distribution-dependent and sensitive to model specification. Methodologically, it highlights the importance of quantile regression in uncovering heterogeneity masked by mean-based estimators. From a policy perspective, the findings suggest that regulators and standard setters should strengthen monitoring and disclosure frameworks, as investors appear to respond selectively to higher-quality, performance-adjusted earnings signals.