Using business technologies to detect fraud

An accounts payable clerk makes fictitious entries to steal money from the company to pay off his debts; a rogue trader creates false entries and counterparty accounts to manipulate his trading position so as to receive a bigger bonus; or an accounting employee pays expense claims for employees who are not on the employee master file.

These are just three among countless potential fraud situations that could devastate a company, not only from a financial perspective, but also in terms of its loss of reputation.

So how do you manage fraud risk proactively ¡V before suspicions are raised? How do you use widely available business technologies to create better anti-fraud programmes, or for early identification of suspicious transactions? 

Some companies invest heavily in various information-management technologies, such as business intelligence, data warehousing, data mining, customer relationship management, analytics, integration and content management news, and so on. These technologies help generate different types of data that are used mainly in operational and management processes. They also make report generation more efficient and provide analytical views from existing data. Companies can generate weekly management-information-system data, bi-weekly sales forecasts, or calculate monthly profit margins easily and automatically. New reports or queries can also be generated simply, if data are available.

A number of these information-management technologies can be employed in different ways, one of which is to create a fraud-detection process to highlight suspicious transactions. You can use these technologies to perform in-depth analyses on large volumes of data embedded in your company¡¦s key sub-ledgers, including accounts receivable, accounts payable, revenue, inventory, payroll and fixed assets.

To get started, here are few simple steps:

ƒÜƒnKnow what you have. Make an inventory of what systems you have in your environment (reporting tools, business intelligence, extract, transform and load, enterprise resource planning, customer relationship management, identity management and so on).

ƒÜƒnUnderstand your data. What types of data are generated in your environment? Is the data correct? Is it consistent across systems? If not, what needs to be done to improve the data quality? Is manual reconciliation required? How long will the process take?

ƒÜƒnCreate exception reports for various fraud scenarios. To perform an in-depth analysis, use your information-management technologies to match existing data against fraud scenarios (for example, to identify an employee whose address is similar to a vendor¡¦s, or identify payments made to two invoices with the same reference numbers). List the exception reports that each management division needs to monitor. These will highlight abnormal transactions that require investigation. Systems can even be configured to generate alert messages or automatically send a notification to a supervisor or upper management, to enable prompt investigation of suspicious transactions.

Existing technologies can be used to detect fraud. The key to success is the availability of good-quality data. If such data is not available, it is time to discuss why data that is important to fraud monitoring does not exist in your environment. This will provide good feedback on which to base performance improvement in your operations.

If you take the suggested steps immediately, they will improve your ability to create a robust fraud-monitoring mechanism in your organisation.

Vilaiporn Taweelappontong is a partner of the advisory services of PricewaterhouseCoopers Thailand. She can be contacted at vilaiporn.taweelappontong@th.pwc.com