کارایی مدلهای آماری والگوهای یادگیری ماشین در پیش بینی گزارشگری مالی متقلبانه The efficiency of statistical and machine learning models in fraud financial statement
محورهای موضوعی : اقتصاد مالیحسن ملکی کاکلر 1 , جمال بحری ثالث 2 , سعید جبارزاده کنگرلویی 3 , علی آشتاب 4
1 - گروه حسابداری، دانشگاه آزاد اسلامی، واحد ارومیه، ، ارومیه، ایران
2 - گروه حسابداری، دانشگاه آزاد اسلامی واحد ارومیه، ارومیه، ایران
3 - گروه حسابداری، واحد ارومیه، دانشگاه آزاد اسلامی،آذربایجان غربی ،ایران
4 - گروه حسابداری، دانشگاه ارومیه، ارومیه، ایران،
کلید واژه: G32, M42, واژههای کلیدی:گزارشگری مالی متقلبانه, مدلهایآماری, مدلهای یادگیری ماشین. طبقه بندی JEL: M41,
چکیده مقاله :
وجود تقلب و تداوم آن در صورتهای مالی،آثارگسترده ای بر سلامت مالی شرکت ها و توسعه پایدار بازار سرمایه دارد. روش های متداول حسابرسی در پیشگیری و کشف صورت های مالی متقلبانه، نتوانستهاندباتقلبهایحسابدارینوظهور به دلیل فقداندانشموردنیازداده کاوی،پیچیدگی تقلب های جدید و عدم تجربهکافیحسابرسان کناربیایند. در این پژوهش، انواع مدل های آماری و یادگیریماشین در دست یابی به الگویی با کارایی بالا در پیش بینی گزارشگری مالی متقلبانهاستفاده شد. از 20 متغیر در قالب الگوی پنج ضلعی تقلب با تاکید بر ساختار کنترل های داخلی در 166 شرکت هایفعال در بورس اوراق بهادار تهران طی سالهای 1388 الی 1397 و مقایسه بین مدل های مورد بررسی،باکمکآزمـونمقایسـة نسبت ها،نشان میدهدکهبه لحاظ آماریمدل هاییادگیریماشـیندرپیشبینیگزارشگری مالی متقلبانه نسـبتبـه مدل هایآماری،کارایی و دقتبیشتری دارند. ترکیب الگوریتم درخت تصمیم گیری CHAID، C5 و C&R بالاترین دقت در پیش بینی گزارشگری مالی متقلبانه را با دقت بالای 61/92 درصد در پیش بینی تقلب نشان می دهد. روش های داده کاوی بر پایه مدل های یادگیری ماشین و بویژه ترکیب آنها بطور موفقیت آمیزی در پیش بینی و کشف تقلب در صورتهای مالی می تواند مورد استفاده قرار گیرد.The efficiency of statistical and machine learning models in fraud financial statement Hassan Maleki KaklarJamal Bahri SalethSaeed Jabbarzadeh KangarloeeAli AshtabThe existence and persistence of fraud in financial statements can have adverse impact on the sustainable development of the capital markets as well as the financial health of companies. Using conventional audit procedures which is applied to prevent and detect fraudulent financial statements, auditors fail to cope with emerging accounting frauds. This can be due to many reasons, such as the lack of the required data mining knowledge, the complexity and infrequency of financial frauds, and the auditors without much experience. Accordingly, due to importance of identifying fraud in capital market, different types of statistical and machine learning based models were examined to establish a rigorous and effective model to detect financial statements fraud in this study. For this purpose, 20 variables in the form of the pentagonal fraud with emphasis on the structure of internal controls (pressure, opportunity, justification, capability, arrogance and internal control structure) were used from 166 manufacturing companies listed on Tehran stock exchange over the period 2009-2018. Based on the statistical indices obtained, machine learning based models exhibited higher predictive ability and accuracy than statistical based models in predicting financial statement fraud. The results also showed that C5, CHAID and C&R decision tree models were highly accurate in prediction of fraudulent datapresented in fnancial statement. Accordingly, the efficacy of combination of CHAID, C5 and C&R decision tree algorithms which had the highest accuracy in prediction of fraudulent financial reporting was examined. The high accuracy of 92.61% of the combination of these algorithms in fraud prediction shows that data mining methods based on machine learning models and especially their combination can be used successfully in fnancial statement fraud prediction.
The existence and persistence of fraud in financial statements can have adverse impact on the sustainable development of the capital markets as well as the financial health of companies. Using conventional audit procedures which is applied to prevent and detect fraudulent financial statements, auditors fail to cope with emerging accounting frauds. This can be due to many reasons, such as the lack of the required data mining knowledge, the complexity and infrequency of financial frauds, and the auditors without much experience. Accordingly, due to importance of identifying fraud in capital market, different types of statistical and machine learning based models were examined to establish a rigorous and effective model to detect financial statements fraud in this study. For this purpose, 20 variables in the form of the pentagonal fraud with emphasis on the structure of internal controls (pressure, opportunity, justification, capability, arrogance and internal control structure) were used from 166 manufacturing companies listed on Tehran stock exchange over the period 2009-2018. Based on the statistical indices obtained, machine learning based models exhibited higher predictive ability and accuracy than statistical based models in predicting financial statement fraud. The results also showed that C5, CHAID and C&R decision tree models were highly accurate in prediction of fraudulent data presented in fnancial statement. Accordingly, the efficacy of combination of CHAID, C5 and C&R decision tree algorithms which had the highest accuracy in prediction of fraudulent financial reporting was examined. The high accuracy of 92.61% of the combination of these algorithms in fraud prediction shows that data mining methods based on machine learning models and especially their combination can be used successfully in fnancial statement fraud prediction.
Annisya, M., Lindrianasari.,& Asmaranti, Y. (2016). Pendeteksian Kecurangan Laporan Keuangan Menggunakan Fraudو Jurnal Bisnis dan Ekonomi, 23(1), 72-89.
Research in Governmental and NonProfit Accounting, 11, 213–228.
yang Terdaftar di Bursa Efek Indonesia. Jurnal Akuntansi dan Auditing Indonesia, 19(2), 112-125.
یادداشتها
_||_