Applying Benford’s law to detect accounting data manipulation: an empirical study of non-performing loans (NPL) scenario in banking industry of Bangladesh

Abstract

Non-performing loans (NPLs) in Bangladesh’s banking sector nearly tripled during the 2023–2025 period, signaling a potential financial crisis. Rising NPLs reduce banks’ interest income and deplete their capital base, making financial institutions more vulnerable and threatening the country’s economic stability, which heavily depends on the banking sector. It has been alleged that manipulation of accounting data further exacerbates the problem by understating liabilities and overstating assets, thereby presenting an unrealistically favorable financial position. In this context, the study applies Benford’s Law to detect possible manipulation of banks’ loan data, particularly across different categories of NPLs, including Special Mention Accounts (SMA), Substandard (SS), Doubtful (DF), and Bad Loans (BL). Benford’s Law, which describes the expected frequency distribution of leading digits in naturally occurring datasets, was used to analyze NPL data from major commercial banks. Statistical tests, including the chi-squared test and the Kolmogorov–Smirnov (K–S) test, were conducted to assess the conformity of the leading-digit distribution with Benford’s Law. The results revealed significant deviations from the expected distribution, indicating potential data manipulation or fraudulent reporting in the NPL records.