Covering Disruptive Technology Powering Business in The Digital Age

Home > DTA news > News > Using data to deal with ever-evolving nature of fraud
Using data to deal with ever-evolving nature of fraud
January 23, 2017 News big data Fraud Analytics


PRIME MINSTER Prayut Chan-o-cha, aware of the damage corruption causes, made anti-fraud measures one of the national priorities for both the public and private sectors. Thai private company leaders can fast-track anti-fraud programmes, reducing risk to their bottom line, which addresses a government priority.

Recent studies in Thailand suggest fraud is the third largest challenge to doing business in Thailand, estimating that more than 50 per cent of companies could be victims of fraud and that a significant percentage of affected companies, even after remediation efforts, remained uncertain as to the size of the fraud they experienced and thus lacked confidence in the scope of anti-fraud efforts.

Rapidly evolving fraud landscape

While traditional fraud detection and remediation mechanisms remain in place, fraud is also rapidly evolving. Fraudsters are becoming more sophisticated and frequently use emerging technologies, causing an increased fraud risk from rapidly occurring larger frauds and compelling risk-aware companies to employ more advanced fraud management strategies to effectively combat fraud. Best-practice fraud management standards and industry specific guidance from government regulators have been updated to address the changing landscape. These frameworks include using fraud analytics and big data as key elements of fraud detection programmes.

Companies must use the power of their own data to detect and combat fraud. Advanced analytics technologies and other detection mechanisms, like whistleblower programmes, can and do detect a large percentage of previously unknown company fraud.

Traditional rules-based versus advanced technologies in the fight

Thai companies need to embrace new approaches and use fraud analytics. Depending on the size and complexity of a company’s operations, deploying rules-based expert-driven fraud analytics may be sufficient or more advanced analytics techniques may be required to identify potential fraud.

Rules-based analytics are generally effective but advanced techniques, not relying on expert rules, may detect unknown suspicious patterns or new fraud types. As both rules-based and advanced analytics may be appropriate, some companies use a hybrid analytics approach to ensure nothing is missed in potential fraud areas.

The size of big data in some companies saw the growth of impressive and effective advanced analytics approaches to detecting fraud, such as unsupervised clustering not relying on expert rules. These technologies allow for patterns or anomalies to be detected that may have gone undetected by rule sets, such as an unknown fraud type or a new twist on an old fraud.

Fraud analytics case study – Thailand

Does fraud analytics detection work for Thai companies? The chief executive of a manufacturing company recently engaged my team to conduct an internal controls review. Being unsure whether fraud was occurring due to a weak control environment, the team used rules-based fraud analytics to investigate payments in purchasing and payroll.

The financial data was analysed for a two-year period. Quickly, suspicious transactions were discovered in both functions including large purchases from certain suppliers of materials the company did not use. The company confirmed that these were fraudulent. The use of analytics assisted the company in understanding the true weakness in its controls and the exposure to fraud it faced and it moved quickly to strengthen anti-fraud control.

Using analytics techniques is highly effective, but it is also essential companies review their business, build better fraud management frameworks and effectively implement the key pillars on anti-fraud, namely proper planning, prevention, detection and response strategies.

Global studies by the Association of Certified Fraud Examiners regularly report best-practice fraud management programmes detect fraud earlier and reduce losses by up to 50 per cent, particularly through analytics and whistleblower systems.

Richard Batten is a forensic director with Deloitte Southeast Asia and is based in Thailand. He is also part of Deloitte’s Financial Crime Strategy and Response Network. He has extensive fraud management experience including rules-based systems design, strategic leveraging of big data fraud analytics and emerging analytics methodologies used for preventing and detecting fraud and corruption.

This article was originally published and can be viewed in full