自早期美國安隆(Enron)、世界通訊(WorldCom)等爆發的財務醜聞後,近年來證券市場的上市櫃公司財務報表舞弊案件一直不曾斷過,這股歪風不只國內有,甚至蔓延到全世界的各行各業,隨著監控力的加強,越來越多的財務報表舞弊事件浮出水面,顯示了企業在內部舞弊的偵測與外部立法防治的不足,為了健全企業的公司治理與降低不當成本的支出,有必要對金融舞弊及其它不法的行為採取防治措施。 為了將因財報舞弊所損失的金額降到最低,企業必須透過積極的措施去發覺舞弊危險因子,因此,本研究提出了一個財務報表舞弊的查核機制,運用社會網絡密度分析財務報表內的投資關係,並結合財務分析指標,對28個產業內共1308間企業的財報進行班佛分析找出可能異常的節點進行分析,並求證其真實性與合理性。根據實驗結果顯示,該產業中其中不乏有些高異常的企業曾經發生財務造假的案件,而其他可能異常筆數較多的企業將列為嫌疑觀察對象,需再對其財報科目進行深入探討找出是否有可疑的現象,然而並沒有一種財務報表的鑑定是絕對能偵測出舞弊現象的,有時只能發掘些微的異常,使稽核人員能夠及早發覺財報異常,提高查核的效率與降低人力成本。
Since the American Enron, WorldCom occur financial fraud incidents. In recent years, the company of stock market continue to occur fraudulent financial statement, not only domestic but also spread to the whole world. With the strengthen of Monitoring, more and more financial fraud incidents surfaced, display the company are inadequate for internal fraud detection and external legislation. In order to improve corporate governance and reduce the cost of inappropriate expenditure, must take prevention measures to reduce financial fraud and other illegal behavior. In order to reduce amount of loss due to fraudulent financial statement, company must actively to detect fraud risk factors. Therefore, this study proposed audit mechanism to detect fraudulent financial statement, coupling social network density and financial index analyze investor relations from financial statements, and using benford’s law to detect abnormal nodes to verify authenticity and rationality from 28 industry total 1308 company’s financial statement. According to the experimental results show that some high abnormal company have occurred financial fraud case in the industry, other high abnormal record company will be treated as a suspect of observation object, must study financial statement in deep. However, there is no absolute mechanism to detect fraud, sometimes only discovered anomalies so that auditors can find abnormal financial statements quickly, improve efficiency and reduce personnel costs.