大數據分析審計 —大數據分析軟體個案研究

Abstract

審計理論應用於審計方法執行程序，係查核人員於瞭解及測試受查核公司內部控制制度，包括: 1. 控制環境； 2. 風險評估機制； 3. 資訊環境系統； 4. 控制活動； 5. 監督機制 查核人員於瞭解及測試受查核公司內部控制制度以後，包括對受查核者： 1. 管理階層舞弊風險評估及 2. 管理階層對內部控制制度的遵行與否 並且對受查核者內部控制執行測試後決定綜合風險評估(Combined Risk Assessment)，藉以決定證實測試的時間、性質及範圍，然而前述證實測試的執行，在資產負債表方面，主要係採抽樣方式以年底或期中期間的餘額驗證為主，茲舉應收帳款為例，我們多著重於年底或期中期間的應收帳款存在及發生，然而卻無法明確定義產生銷貨的應收帳款歸屬期間(年度)。在損益表方面，主要係採抽樣方式驗證全年或特定期間個別組成金額，卻無法由損益表項目逐月波動性驗證出損益表各組成科目之合理性。 茲舉本論文中的東芝公司為例，由於管理階層為達成營業目標，在公司遭受重大危難變化後，採取異於正常的會計處理方法，包括提前認列收入，亦同時提前認列應收帳款，在傳統以抽樣方式驗證年底或期中期間餘額的審計方式，的確難以偵知出相關的舞弊或錯誤。 惟有採取大數據分析審計方式，將受查核客戶全年度或期中期間會計資訊依特定格式設定，透過大數據分析審計軟體分析而產生逐月、逐年或逐期的比較資訊，審計人員依照專業判斷，定義出錯誤或舞弊之狀況，依照大數據審計軟體所呈現之結果，針對波動性異常之會計項目進行逐筆交易的檢視，以確認是否的確存在錯誤或舞弊之情事，而非僅依抽查方式進行，如此不但使查核係針對高風險項目執行查核，亦可使查核更具效率。

In traditional auditing approach, auditors will understand and test an audit client’s internal control including 1. Control environment, 2. Risk assessment, 3. Information technology system, 4. Control activities, and 5. monitor functions The evaluation includes to understand the fraud risks of the management and the entity level controls of the management. Auditors will test: 1. The effectiveness of the management preventing from the fraud and 2. Monitoring the internal control of the company Based on the results of test of controls, auditors will determine the timing, nature and scoping of their substantive testing. In balance sheets, auditors normally sampling audit the year end or period end account balances. Take accounts receivable as an example, auditors will put emphasis on auditing the existence of accounts receivable, however, the correctness of cut-off is often neglected. In statements of income, auditors sampling test the items in the income statements, however, the fluctuations of trends by periods are often neglected. Take Toshiba Corporation (“Toshiba”) as an example, Toshiba intends to meet its earnings projection after a big nature disaster. They adopt improper revenue cut-off to recognize the following year’s revenue in the current year, which resulted in both over accrued revenue and accounts receivable in the current year. It proved that traditional audit approach may not detect errors or frauds. Under big data analysis auditing, auditors use big data analysis software to run the audit client’s accounting information. There will generate various diagrams of month over month, year over year, or period over period. Through the fluctuations or trend analysis, auditors can decide their audit procedures based on their professional judgements. The audit approach will be detail testing of the high risk transactions (accounts) but not sampling testing. In this approach, auditors can focus on the auditing of high risk transactions (accounts) and perform the audits efficiently.