Investigating the Impact of Earnings Management and Internal Control on the Financial Performance of Banks Using a Spatial Artificial Intelligence Approach

This study examines the relationship between earnings management and financial performance of banks listed on the stock exchanges in Iran and Iraq, focusing on the role of internal controls. This study utilizes the statistical population of ۴۴ banks from Iraq and ۲۲ banks from Iran during ۲۰۱۰-۲۰۲۳ using the combined methods of deep learning, machine learning and spatial metrics. Deep neural networks identified different patterns in the two countries and deep learning algorithms determined the relative importance of the variables. Earnings management showed a significant positive relationship with financial performance (spatial coefficient: ۰.۵۹۵۳, Z-statistic: ۶۱.۸۷). Income per employee (coefficient: ۰.۳۲۱۸, Z-statistic: ۳۳.۵۷) and sustainable investment return (coefficient: ۰.۲۸۷۱, Z-statistic: ۲۹.۹۱) also had a positive impact on financial performance. Operating expense margin, ratio of non-performing loans to total loans, and general and administrative expenses negatively affected financial performance. Internal controls played a mediating role, strengthening the relationship between Earnings management and performance (coefficient: ۰.۴۱۲۷, Z-statistic: ۴۳.۰۸). Earnings Management has a significant role in the financial performance of banks and internal controls strengthen this relationship. It is suggested that the monetary authorities of both countries establish stricter rules for financial reporting, set minimum standards for manpower productivity, and require banks to implement advanced control systems. These findings suggest that bank regulators in both countries should adopt stricter accounting rules, set minimum standards for employee productivity, and mandate the implementation of advanced control systems. For bank managers, the results highlight the importance of focusing on earnings management strategies, improving staff productivity and enhancing internal control mechanisms in order to improve financial performance.

کلیدواژه ها:

Earnings Management ، financial performance ، internal control ، Deep Learning ، Spatial Econometrics