The Research On The Fraud Identification Based On XBRL Financial Report

As our country’s capital market is not mature, the financial fraud of listed companiesis very serious, false financial information will not only affect the healthy development ofmarket economy, they will also bring bad influence to society. The related theoreticalresearch about financial fraud problem has not yet formed a relatively mature system ofresearch at home and abroad, especially the study of XBRL financial reporting fraudidentification are much rarer, and its empirical research also was still in the stage ofexploration. Strengthen the recognition and control of XBRL network financial report fraud,not only is the inevitable choice of perfecting the capital market credit body, but also is theinevitable choice of maintaining the normal order of capital market and rebuildinginvestors’ confidence. This paper mainly study the effect of XBRL technology foraccounting fraud and the XBRL environment how to identify financial fraud effectively.Through theoretical analysis and empirical research, this paper received threeconclusions: first, the unique advantages of XBRL technology has a great role for financialreport reported, it is imperative to promotion and the development of XBRL financialreport. Second, XBRL has various effects on the accounting information quality, auditing,internal control and supervision and so on, this effects are quietly changing theenvironment of financial reporting fraud, so studying the characteristics of fraud companythat under XBRL environment is particularly important in this article, through independentsample T test, we get15significant indexes, as shown in the table5-3. Finally, in this paper,we introduce the financial indicators and corporate governance index combining with thecharacteristics of XBRL technology on the basis of previous studies, we also introduce therelevant deviate from the horizontal and vertical deviation index indicators, then weestablish the fraud identification model based on XBRL financial reporting using Logisticregression, it has high recognition rate, and let the fraud and paired samples whose datawas shown by XBRL format back to generation model, by comparing this model and thepredecessors’ study model, we found that the overall accuracy higher than that of thepredecessors’ research model, so we can confirm the necessity of the XBRL environmentto modeling.