Major Companies Not Making Full Use of Big Data to Spot Fraud

Most companies are spurning the chance to improve their anti-fraud and anti-bribery efforts by not taking full advantage of big data analysis, according to research from business consulting firm EY.

Most companies are spurning the chance to improve their anti-fraud and anti-bribery efforts by not taking full advantage of big data analysis, according to research from business consulting firm EY.

[ 17 Steps to Implement a Public Sector Big Data Project ]

EY found that 63 percent of senior executives surveyed at leading companies around the world agreed that they need to do more to improve their anti-fraud and anti-bribery procedures, including the use of forensic data analytics (FDA).

The survey polled more than 450 executives in 11 countries, including finance professionals, heads of internal auditing and executives in compliance and legal areas. They were asked about their use of FDA in anti-fraud and anti-bribery compliance programs.

Bribery and corruption was reported as the top perceived risk at 65 percent, with other significant fraud risk areas cited including asset misappropriation and financial mis-statement.

Paul Walker, head of EY's forensic technology and discovery services in the UK, said: "Our findings suggest that while companies may be doing some forms of FDA, many could be missing important opportunities to improve their anti-fraud and anti-bribery efforts.

"By combining multiple data sources and leveraging advanced FDA tools, companies are now able to gain new and important insights from their business data."

The research suggests that the vast majority of companies are not working with sufficient data volumes given the size of their corporate revenues. Only 18 percent of internal audit professionals polled are working with data volumes in excess of one million records.

Among financial services respondents, only 21 percent report working with data volumes nearing and over one million records, which is still low for such a data-intensive industry, said EY.

Overall, 71 percent of companies with over $1 billion (APS600 million) in revenues are working with data sets under one million records. "The use of smaller than expected data volumes, relative to corporate revenues, raises the question that many companies may be missing important fraud prevention and detection opportunities by not mining larger datasets," said EY.

Advanced FDA tools like statistical analysis and data-mining technologies are used by only 11 percent of respondents, according to the survey. Interestingly, cost does not appear to be a major obstacle, with just 10 percent of respondents indicating that FDA is "prohibitively expensive".

EY said traditional spreadsheet and database applications can struggle with the increasing volumes, velocities and varieties of data generated by global companies.

Advanced FDA technologies including statistical tools that incorporate predictive modelling, anomaly detection and risk-scoring algorithms, can mine such big data to detect potentially fraudulent transactions in real- or near-real-time, said EY.

Also, said the consulting firm, the effective use of natural language processing, or text-mining, combined with data visualisation, can handle a wide variety of sources, including both structured and unstructured data, to improve overall detection, reduce risk and increase return on the investment in FDA.

Respondents who are using FDA technologies beyond spreadsheet and database tools have observed earlier detection of misconduct - 15 percent more than others in the survey - and improved results and recoveries - 11 percent more than others.

This story, "Major Companies Not Making Full Use of Big Data to Spot Fraud" was originally published by Computerworld UK.

Insider Resume Makeover: How (and When) to Break the Rules
Join the discussion
Be the first to comment on this article. Our Commenting Policies