An Empirical Study On Fraudulent Financial Reporting Of Listed Companies In China

With the gradual development of Chinese capital market, accounting informationdisclosed by public companies, especially the information on accounting earnings andearnings quality, has become an important source of information to investors whenthey're making their investment decisions. Meanwhile, accounting earnings, andmetrics derived from it, are widely used in regulations and contracts, such ascompensation arrangements and debt agreements. However, accompanying thedeveloping role of accounting information, we also observe a frequent break out ofaccounting frauds in Chinese capital market. Fraudulent accounting information notonly impairs the benefits of a large number of investors, but also interrupts the healthydevelopment of capital market. For this reason accounting frauds have attractedsignificant attention from regulators, investors, creditors and other stakeholders. Amajor concern shared by all these parties is how to identify accounting fraud in itsearly stage. Motivated by this question, this paper provides empirical evidence on motivationsfor accounting fraud, the association between corporate governance structure andaccounting fraud, and financial characteristic of fraud firms. It also uses logisticregression to establish a model in identifying fraud firms. We hypothesize that accounting frauds are more likely to happen when publicfirms have strong external financing needs and when weak governance structureoffers the necessary opportunities. Moreover, fraud firms are likely to experiencesome abnormal changes in their financial ratios. These abnormal changes can eitherbe the causes of the fraud or be the consequences of the fraud. Specifically, we test four hypotheses in the paper: 1) Accounting frauds are morelikely to happen in firms with strong external financing needs. 2) Accounting fraudsare more likely to happen when top management has larger equity holdings of thefirm. 3) Accounting frauds are more likely to happen in firms with weak governancestructure.4) Accounting frauds are more likely to happen in firms with abnormalfinancial numbers or ratios. Our sample is selected from public firms subject to accounting enforcementactions by the China Securities Regulatory Commission(or Shenzhen Stock VIExchanges, or Shanghai Stock Exchanges) for alleged earnings overstatements. Thereare 30 fraud firms left after imposing certain sample selection criteria. Our finalsample consists of 55 fraud firm-years and 55 non-fraud firm years, spanning theperiod from 1995 to 2002. We use various statistical methods to test our hypotheses.These statistical methods include parametric estimation, nonparametric estimation andfactor analysis. We also establish a logistic regression model to help distinguish fraudfirms from non-fraud firms using pubic available information. While we fail to find a statistically significant relation between externalfinancing needs and accounting fraud, we find that the likelihood of accounting fraudis positively related to the number/proportion of company shares held by topmanagement , especially members in the board of directors. Thus, an importantmotivation for accounting fraud is top management incentives to benefit form inflatedstock prices. In terms of corporate governance structure, we find that the likelihood ofaccounting fraud is higher in firms with dispersed stock controls, where effectivemonitoring of top management's behavior is more difficult. Additionally, firms withhigher proportion of equity holdings by other companies are less likely to engage inaccounting fraud, possibly because they can use other "legal"means such as thirdparty transactions or debt restructurings to manage earnings. Finally, our results suggest that although fraud firms manages to show similarprofitability as control firms through earnings manipulation, their asset qualities, assetturnovers and cash flows are significantly worse when compared w...