According to a recent FT.com article, US lawmakers are seeking to expand the reach of anti-money laundering regulations after a Senate investigation found that hundreds of millions of dollars in suspect foreign funds have been able to land in the US. As we can see, money laundering and terroristic financing continues to rise despite the huge investments banks have made in packaged Anti-Money Laundering (AML) software, custom built solutions through global system integrators, re-engineered business process, people, and policies. How can this be? Were these bad investments to begin with or is there something else causes these issues?
One potential cause is the way banks approach data integration and data quality to support their anti-money laundering requirements. If you are from my generation, we often correlate the term “data integration” as ETL for data warehousing or “real-time data integration” as message queues or EAI technology. Also, managing data quality required a PhD in Computer Science or a black belt in statistics or data analysis. How times have changed.
Today’s data integration technology is definitely more than ETL. It’s about support all forms of data movement from batch to real-time, accessing data regardless of format, structure, and location for analytical and operational systems. It’s about accessing and leveraging your “data as a service” through SOA based architectures to avoid being subject to stringent and fixed data management processes that can delay the delivery of time sensitive data to those who need it most. It’s about enabling common business users to understand, inspect, and manage the quality of their data in collaboration with IT developers vs. throwing requests over the wall and not knowing what was done with it.
Anti-money laundering software, custom built or off-the-shelf is only as effective as the quality, consistent, completeness, and timeliness of the data. These are complex systems that perform specific functions however are not designed to access, cleanse, and deliver data from the various sources of data. For example, one bank I spoke with recently, purchased a leading AML applications several years ago which required 36 separate and unique extracts from 16 unique source systems ranging from deposit banking systems, online banking, credit card processors, mortgage origination and servicing systems, consumer lending applications, trading systems, etc. in which in large institutions, each category can include up to 5 different systems for each category due to mergers and acquisitions over time.
Many of these sources were legacy systems designed to transact but not designed to make its data accessible to downstream systems. They were dealing with non-conforming data formats, increasing data volumes and mountains of unstructured files including .PDF, spreadsheets, and text files that contained the data their AML solutions need to be effective. Unfortunately, they chose to hand code their data integration and data quality processes, extracting flat files from mainframes via ODBC, scanning documents through OCR readers, and writing Java or C++ scripts to transform the data into a usable format, and attempting to fix data quality issues in a spreadsheet. The amount of time and cost of performing these tasks were significant and not surprisingly ineffective thus raising the question from executives, “Is the AML solution delivering the value they were promised 3 years ago?”
We live in uncertain and dangerous times and the applications built to identify these suspicious activities require timely, trustworthy, and consistent data to do its job. The only effective way of meeting what the business needs is to leverage proven data integration and data quality solution that can scale and perform to handle any data integration and data quality need whether for anti—money laundering or any other business need where timely, trustworthy, and relevant data is required. To learn more about how Informatica 9 for your data integration and data quality needs.
�
