Detecting Earnings Manipulation Practice by the M-Score Model: Evidence from the Listed Power Companies of Bangladesh

Abstract

Purpose: The study aims to detect earnings manipulation practices, if any, and to investigate the relationship between company-specific characteristics and earnings management in the power sector of Bangladesh. Methodology: In this study, a panel dataset comprising data from annual reports of 22 power sector companies listed in the DSE was utilized. The dataset spanned seven consecutive years, from 2014 to 2020. To identify earnings manipulation practices, the researchers employed the Beneish M-Score model. Findings: The results of this study indicate that approximately 30% of the Fuel & Power sector companies listed in Bangladesh engage in information manipulation. Moreover, it was found that 64% of these manipulators received an unqualified opinion from auditors. Among the 22 selected power companies, 18 exhibited significantly higher M-Scores for at least one year during the period of 2014-202. Regression analysis shows that accrual quality has a significant positive association with earnings management, while the firm size and audit quality are negatively related to earnings manipulation. However, firm age and audit opinion did not demonstrate any significant influence on earnings management. Originality/Value: This study marks the pioneering use of the Beneish MScore model in the Fuel & Power sector of Bangladesh to detect earnings management practices. The findings suggest that having more non-cash items in the income statement allows management to manipulate, and large firms with strong corporate governance are less likely to manipulate information. These findings are valuable for decision-makers and stakeholders such as investors, policymakers, and the government. Limitations: Only one sector has been chosen for investigation in this study. Selecting more samples from each industry could give a broader picture of earnings manipulation practices by the companies in Bangladesh.