Anomalous Detection in Gst Collections: A Forensic Analysis using Benford's Law

Abstract

In the modern financial landscape, ensuring the accuracy and integrity of financial data is crucial. Forensic auditing provides a robust mechanism to detect and investigate irregularities, frauds, and errors in financial records. The aim of the present study is to detect irregularities and potential frauds in GST collections from Karnataka using Benford's Law and to assess the conformity of GST data (CGST, SGST, IGST, and CESS) to expected patterns and distributions. The dataset comprises of monthly GST collections of Karnataka State from July 2017 to May 2024, segmented into CGST, SGST, IGST, and CESS. The methodology for this forensic audit study involves two primary techniques: Benford's Law analysis and Anomaly Detection. The analysis indicates that CGST and IGST show acceptable conformity to Benford's Law with moderate distortion, suggesting relatively reliable data. In contrast, SGST and CESS show marginal conformity with higher distortion, indicating potential anomalies or irregularities.