Automating detection of insider trading, money laundering and fraud

One of the most interesting uses of enterprise decision management or EDM is in automating detection of various kinds of “bad” activity. Whether the concern du jour is insider trading, money laundering or just plain old fraud. There are a number of reasons why EDM is such a powerful approach.

• Automation is fundamentally the best approach both because transaction volumes are so high and because manual approaches require large numbers of staff thus increasing the risk of a crook subverting the process.
• Insider trading, money laundering and fraud can occur in any channel, in any system and typically use a variety of processes. Thus centralized management of the decisions as to whether a given transaction is fraudulent or problematic is much more effective than embedding those decisions in specific systems.
• Although this sounds terribly cynical, much of the time an organization is not so much concerned with catching those committing, say, money laundering so much as being able to show they were compliant with the rules defined by the government. Business rules, because they make it so much easier to demonstrate compliance, are thus a great implementation approach.
• Much of the know-how for catching crooks and money-launderers comes from law enforcement specialists, experience of other organizations and prior problems and business rules are also a great way to capture these nuggets of expertise and do so in a way that can be managed, updated and shared.
• Analyzing data to find patterns of fraud and to identify transactions outside the usual patterns of a specific trader or type of traders are classic uses of predictive analytics. Predicting how likely a transaction is to be problematic in one way or another, and doing so as the transaction executes using a predictive analytic model, feeds powerful information to the rules that decide how to act
• The problem is constantly changing making agility critical, reinforcing the value of business rules, and making adaptive control techniques useful in constantly and systematically challenging approaches to see what the best approach should be going forward

Enterprise decision management, EDM, with its focus on the independent automation and management of decisions and its use of business rules, predictive analytics and adaptive control, is ideal for this kind of automation. This post was prompted by the Societe Generale case and a post on the analytical engine. There are some posts on my various blogs you might also find interesting:

• AML and business rules
• EDM and Enterprise Fraud management
• Using EDM to stay a step ahead in fraud
• Identity theft and decisioning

Plus there is a white paper I wrote about healthcare fraud here.