Investigating ESG misconduct through forensic auditing: lessons from the Indonesian palm oil sector

Abstract

This study develops a novel forensic auditing framework to detect ESG fraud in Indonesian palm oil firms, addressing the absence of empirical models in agribusiness sustainability reporting. A sequential explanatory mixed-methods design was applied to 75 firm-years (2020–2024) from 15 IDX-listed companies. Quantitative analysis using the Beneish M-Score revealed a mean of -2.05 (SD = 0.32), with 18.7% of cases exceeding the manipulation threshold (-1.78). Total Accrual to Total Assets (TATA) was the dominant fraud signal (OR = 107.8, p < 0.001), linked to biological asset overcapitalization. Logistic regression confirmed that higher ESG disclosure scores significantly predict earnings manipulation (β = 0.92, p < 0.001). Qualitative triangulation via satellite imagery (Global Forest Watch) and semi-structured interviews (n = 25) identified fraud in 16% of cases through geospatial discrepancies. Post-forensic audit intervention reduced M-Scores by 0.58 (Cohen’s d = 1.12, p < 0.01) and improved ESG land accuracy by 6.2%. This research is the first to integrate Beneish M-Score, blockchain traceability, and satellite cross-verification in palm oil ESG assurance. Findings expose systemic greenwashing under POJK No. 51/2017 and validate forensic auditing’s role in restoring credibility. Policy recommendations include mandatory third-party geospatial verification and a national early warning dashboard integrating M-Score and satellite data. The framework offers a replicable model for fraud-prone agribusiness sectors worldwide.