As the world expands, with more and more data being made available, the opportunity to abuse this vast pool of information for the purposes of fraud - or to manipulate it for personal gain - is enhanced. Deloitte's Fariel Hoosen writes on the matter.
Approximately 72 years ago the term “information explosion” was coined. Between then and now the amount of data which has become available has increased exponentially.
We find ourselves inundated with data, and in order to transform this data into information and knowledge we need to apply analytical intelligence.
In today’s highly data drenched environment the application of data analytics specific to fraud detection and prevention has become crucial across all sectors. Forensic Data Analytics is now a driving force in identifying fraudulent trends within vast amounts of data.
The sources of data we would analyse are:
- Payroll and Human Resources
- Financial statements
- Fixed Assets
- Expenses and Disbursements
- Asset Management
- Any form of structured data can have analytics applied to the data in order to pick out characteristics indicative of fraud.
Data analytics and fraud detection
From our experience we have noted a general sense of reactive behaviour to data related fraud events. An event is when a business’s data system is misused by performing actions such as:
- Unauthorised payment transactions to a vendor;
- Passing unauthorised manual journals;
- Duplicate payment on an invoice with two different bank account numbers;
- Unauthorised payroll run; and
- Amendments to banking details on either the payroll or vendor data in order to channel funds into a different bank account.
In order to analyse and interrogate the data we would use appropriate software tools in order to develop and apply investigation specific analytics to the data in order to detect anomalies which could potentially be fraudulent or displays fraudulent characteristics.
The Forensic Data Analytics applied can be highly focussed in support of an investigation or, if the fraud investigation is still in the early stages, analytics can be harnessed to streamline the investigation team with directing the focus on particular areas.
Data analytics and fraud prevention
Using historic data and being able to pin point data events which led to fraud, waste or abuse can be very powerful. Developing statistical models based on this data in order to predict potential fraudulent behaviour will allow business to make sense of a seemingly random act. Data triggers can be developed to highlight when similar data events occur in order to highlight anomalies.
These mechanisms would be put in place on a continuous basis and red-flags will be raised, and these anomalies will then be investigated. These mechanisms will be used by internal stakeholders to monitor the activity around areas where fraud has occurred, and mitigate risk by assessing the information provided from these mechanisms.