Integrating Non-financial Data into a Creative Accounting Detection Model: A Study in the Saudi Arabian Context

Global financial scandals have demonstrated the harmful impact of creative accounting, a practice in which managers creatively manipulate financial reports to conceal a company’s performance and influence stakeholders’ decision-making. Studies showed that Saudi-listed companies engage in creative accounting when preparing financial statements. However, big data analytics has found practical applications in auditing, and recently, the use of Deep Learning in financial statement fraud detection has yielded remarkably accurate results. Therefore, our research aims to train a hybrid learning Creative Accounting Detecting Model (CADM) proposed by [18]. This study seeks validation for non-financial data to be used in CADM training. Among the chosen factors that represent non-financial data, the reputation of the external auditor is the most influential factor in the credibility of information extracted from financial statements. The analysis also revealed that the accounting subject showing the most variability in interpretation within the Saudi business environment is the disclosure of management compensations. Additionally, many innovative accounting systems are still awaiting adoption in Saudi Arabia despite the government’s implementation of advanced interconnected systems. Lastly, a consensus was reached on most of the recommended data sources, particularly those obtained from Saudi authorities. Despite providing the foundation for the non-financial data integration phase, the results will provide insights into the reliability and transparency of financial statements of Saudi-listed companies. It can enhance multiple stakeholders’ decisions and inform Saudi regulators about areas requiring their attention in financial reporting. However, this study is limited by the sample size and the methods employed in analysing the results.