Abstract
Purpose: This paper aims to facilitate the development of more effective models to detect fraud. Financial statement fraud continues to be a costly problem to society and detection of it is difficult. Better aids, such as fraud detection models, are needed to improve detection. The selection of explanatory variables is crucial to developing such models, and yet, surprisingly, the selection process in prior financial statement fraud detection studies is not standardised. This paper proposes a new framework for this purpose.
Design/methodology/approach: The new framework adapts the well-known Fraud Triangle, which was originally designed to explain drivers of fraudulent behaviour and needs to be adapted for use in fraud detection. The adaptation is done by the inclusion of a new Suspicious Information factor. This new framework also incorporates modifications suggested in prior research as extensions to the existing factors.
Findings: Publicly available variables to operationalise each factor of the new Fraud Detection Triangle (FDT) framework already exist and there is preliminary empirical support for each factor in the FDT framework. Research into additional variables that measure R and S factors would be beneficial as less focus has been placed on them in prior research.
Originality/Value/Practical Significance: The FDT is more suited to fraud detection than the previous fraud triangle. As a guiding theory for variable selection (previously done primarily on an ad hoc basis), it can facilitate the development of more accurate detection models. The variables analysed to operationalise the framework are also more comprehensive compared with prior research.
Design/methodology/approach: The new framework adapts the well-known Fraud Triangle, which was originally designed to explain drivers of fraudulent behaviour and needs to be adapted for use in fraud detection. The adaptation is done by the inclusion of a new Suspicious Information factor. This new framework also incorporates modifications suggested in prior research as extensions to the existing factors.
Findings: Publicly available variables to operationalise each factor of the new Fraud Detection Triangle (FDT) framework already exist and there is preliminary empirical support for each factor in the FDT framework. Research into additional variables that measure R and S factors would be beneficial as less focus has been placed on them in prior research.
Originality/Value/Practical Significance: The FDT is more suited to fraud detection than the previous fraud triangle. As a guiding theory for variable selection (previously done primarily on an ad hoc basis), it can facilitate the development of more accurate detection models. The variables analysed to operationalise the framework are also more comprehensive compared with prior research.
Original language | English |
---|---|
Publication status | Published - 2016 |
Event | 8th Asia-Pacific Interdisciplinary Research in Accounting Conference (APIRA) - RMIT University, Melbourne, Australia Duration: 13 Jul 2016 → 15 Jul 2016 Conference number: 8 https://www.rmit.edu.au/events/all-events/conferences/2016/july/apira-2016 |
Conference
Conference | 8th Asia-Pacific Interdisciplinary Research in Accounting Conference (APIRA) |
---|---|
Abbreviated title | APIRA |
Country/Territory | Australia |
City | Melbourne |
Period | 13/07/16 → 15/07/16 |
Internet address |
Fingerprint
Dive into the research topics of 'Addressing the problem of financial statement fraud: Better detection through improved models'. Together they form a unique fingerprint.Student theses
-
Financial statement fraud detection using supervised learning methods
Author: Gepp, A., 10 Oct 2015Supervisor: Kumar, K. (Supervisor) & Bhattacharya, S. (External person) (Supervisor)
Student thesis: Doctoral Thesis
File