PREPRINT (Version 1) available at Research Square.
ISSN/ISBN: Not available at this time. DOI: 10.21203/rs.3.rs-2270360/v1
Abstract: Data growth is exploding in the era of big data. To maximise the value of data, data quality management has become crucial. Establishing a strategy for identifying the quality of scientific data is crucial and essential. Benford’s law has evolved into an effective tool for assessing data quality and spotting anomalous data across multiple sectors. Benford’s law is a digital analytic method that determines the probabilistic distribution of digits for numerous common phenomena. The fraud detection method uses deviations from the expected Benford’s Law distributions as strong signs of fraudulent behaviour. The Wire card fraud, which resulted in losses of several billion euros, is regarded as one of the most notable financial scandals of the decade. This paper examines the digit structure of Wire card’s financial figures from 2005 to 2019, financial figures of the Bank of England, financial figures of the SEC’s Accounting and Auditing Enforcement Releases (AAERs), financial figures of publicly traded U.S. firms, sample financial data, and test data by analysing their conformity with the expected frequency distributions based on Benford’s law. The results indicate that small accounting fraud in large datasets cannot be identified using the Benford analysis of the first digits alone. This work still demonstrates how Benford’s law can be applied to small datasets with a high fraud count in accounting and auditing, hence introducing data analytic approaches for fraud detection. The future evolution of Benford’s Law necessitates that academics from many disciplines conduct additional research on its foundation, strengthen its integration with other data processing technologies, and then broaden its application.
Bibtex:
@misc{,
author = {Rakshit Kaushik},
title = {Using benford’s law and RMSE to predict financial fraud using firm-reported data},
year = {2022},
doi = {10.21203/rs.3.rs-2270360/v1},
}
Reference Type: Preprint
Subject Area(s): Accounting