Model Deteksi Kecurangan Berbasis Fraud Triangle

The case of issuers violation in stock market is one of the most frequent case that should be solved by the stock market’s Regulator Board. In Indonesia, the authority to do the surveillance in stock exchanges is handled by the Capital Market Supervisory Agency and Financial Institution (or called ‘Bapepam-LK’ in Indonesian). There is a gap between society’s expectation toward Bapepam-LK as the regulator and its performance in solving several companies violation cases which demands some alternative solutions. Based on the above description, this study will empirically investigate the development of fraud detection model using fraud triangle based on the cases of violation committed by public companies in Indonesia. In detail, the problems in this research are: (1) Is there any difference related to the triggering factors of Fraud Triangle between the company that commits fraud and the company that does not; (2) Is there any difference in terms of pressure between the company that commits fraud and the company that does not; (3) Is there any difference in terms of opportunity between the company that commits fraud and the company that does not; (4) Is there any difference in terms of rationalization between the company that commits fraud and the company that does not. This research was carried out to the public companies who are registered in Indonesia Stock Exchanges (BEJ). Generally, there are two sample group in this research. The first sample group consiststed of the companies who had committed fraud and the second group as comparison consisted of the non-fraud companies. In this research there are 98 companies as the research samples which consisted of 23 companies who had committed fraud and 75 companies who did not commit fraud. The analysis tool used for this research was logistic regression because the measurements of dependent variable used the categorical that is dummy variable, code (0) was used for the non-fraud companies and code (1) was used to indicate the companies who committed fraud. The research result shows that from four hypothesis proposed in this research, only one variable which fits in to the model (variable in equation) because posesses the significance score above 0.05. The interpretation is that the higher the audit report (rationalization), will make the company’s probability to commit fraud is also higher. From the above explanation, it can be concluded that the fourth hypothesis (H4) is accepted because the audit report (rationalization) is proven to have the ability in forming the model to predict fraud in a company.