Logistic Regression for Detecting Fraudulent Financial Statement of Listed Companies in China

Abstract

This paper examines published data to develop a model of Logistic Regression for detecting factors associated with Fraudulent Financial Statement (FFS). After an exhaustive exploitation of prior work used financial ratios, 21 ratios are selected as potential predictors of FFS and a series of experiments have been conducted to determine the optimal parameters for Logistic model. Then, we propose an appropriate model for detecting FFS of listed companies in China and compare its predictive ability with other detecting models using a data set of 174 listed companies in China including 87 with FFS and 87 with non-FFS during the period 1993-2007.The results demonstrate that the predictive ability of the model proposed in this paper is higher than other models at about 10% by using the optimal parameters determined and indicate the importance of financial ratios, which could benefit both internal and external auditors, taxation and other state authorities.

A Review on Fraud Research : A Study of Vote Count

“…Overseas research on fraud detection methods is also more diverse such as research by (Adrian, 2015) which detects fraud using data analysis, 2000 2001 2003 2004 2005 2005 2005 2008 2008 2009 2009 2010 2011 2011 2011 2012 2012 2012 2012 2013 2013 2014 2014 2014 2014 2015 2015 2015 2016 2016 2016 2016 2017 2017 2017 2018 2018 2018 Theme Distribution per Year research by (Dalnial, Kamaluddin, Sanusi, &Khairuddin, 2014) that detects fraud by using financial statement analysis and research by (Meenatkshi&Sivaranjani, 2016) who examined fraud detection using data mining methods. There are still many studies on fraud detection with various methods such as (Yue et al, 2009) detection using logistic regression, and (Zainudin&Hashim, 2016) which examine the use of ratios to detect fraud.…”

“…Likewise, research relating to fraud prevention was also carried out by several researchers in Indonesia such as (Widaningsih, 2015) who examined the Effect of Internal Auditor Professionalism on Fraud Prevention and Detection. Likewise, research on fraud detection methods using the beneath m score method examined by (Efitasari, 2013) and from abroad (Perols, 2011); (Yue, Wu, & Shen, 2009) and (Kaminski, Sterling Wetzel, & Guan, 2004).…”

Intelligent Fraud Detection in Financial Statements Using Machine Learning and Data Mining: A Systematic Literature Review

“…The studies in this SLR utilized Logistic Regression (LR) method frequently in their researches. In particular, 16 studies employed logistic regression [53,86,62,58,11,51,34,82,57,83,74,42,12,37,35,76]. Logistic regression models are applicable in multi-class and binary classifications.…”

Statistical Medical Fraud Assessment: Exposition to an Emerging Field

“…For instance, more than 80% of the papers reviewed in Phua et al () have skewed data with less than 30% fraud. The sparsity of the fraud data can be addressed with methods such as non‐negative matrix factorisation, singular value decomposition and principal component analysis (Zhu et al , ). Secondly, both legitimate and fraudulent claims have dynamic patterns due to heavy competition in health care industry and updates in the policy and legal frameworks.…”