An iOS App for Accelerating the Detection of Fraudulent XBRL Instance Documents

Although fraudulent financial documents are detected manually, smartphone apps may be created to accelerate the detection of fraudulent financial documents. These apps are programmed to study the conformity of a document to the Benford's law. If unacceptable conformity is concluded, it is more possible that this document is fraudulent. As a preliminary test of such an idea, this study creates an iOS app to evaluate the conformity of an XBRL (eXtensible Business Resource Language) instance document to the Benford's law. This conformity evaluation is implemented using leading and first-two digital probabilities. The conclusion of evaluation results is determined by visual comparison of actual and theoretical digital probabilities, mean deviation, Chi-square, Kuiper, and Kolmogorov-Smirnov test statistics. The latter three types of test statistics are concluded with respect to 10 %, 5 %, 1 % and 0.1 % significance levels. The current iOS app is run without needing any XBRL taxonomy. It demonstrates that a smartphone can be a handy tool for accelerating the detection of fraudulent XBRL instance documents.