Informatica Perspectives

If banks and financial institutions had invested in more data integration and business intelligence tools to spot issues arising within their portfolios, could they have avoided the recent credit mess?

Perhaps, to a degree. But it is human beings that are ultimately making the risk judgments, and oftentimes, bad decisions may have looked good at the time they were made.

Still, technology has improved to the point where troubles could have been more effectively flagged. Perhaps the current troubles in the financial sector will spur the acquisition and development of a new generation of intelligence and analytic tools that will prevent future debacles based on questionable subprime lending practices.

A new survey of 316 financial services executives conducted by The Economist Intelligence Unit found that many blame lack of access to relevant, timely and consistent data as an obstacle to better risk management. “Timeliness and quality of information is highlighted by 44% of executives as being one of the three main challenges in implementing enterprise-wide risk management. As organizations become more complex, the need for sophisticated data infrastructure to gather and process risk information is becoming more pressing than ever before. Just over one-half of the respondents (56%) agree that it is essential to have an enterprise risk data infrastructure in place.”

Many observers feel today’s data integration and business intelligence tools may have been enough to do the job. Ephraim Schwartz, for one, notes that various analytic tools may have provided banks “an early warning” of the risks to their loans and securities. Writing in a recent issue of ComputerWorld, he spoke to several experts, who pointed out that business intelligence metrics could have highlighted debt quality versus the cash reserves financial institutions have on hand, and deliver alerts when things started to get out of hand. However, gathering and consolidating such data from various parts of the enterprise is a non-trivial task.

Complex event processing tied to pervasive BI systems would have made the ability to deliver such alerts more automatic. I heard it pointed out that a typical bank may have 10 million events going on each day, so being able to identify those few business-critical events is something the business needs to carefully weigh. Putting analytics, CEP, and data integration into a common middleware layer that would enable both businesses and regulators to access and share this critical data would have provided greater transparency.

In his article, Schwartz observes that “technology in the financial services industry has been focused on capacity — whether an application can handle high volume and volatility — rather than on process. There is no process flow map that tells organizations who owns what pieces of what risk.” He advocates adopting a process flow approach similar to that already in place within the retail sector.

Once again, humans ultimately give approval to risky decisions, and a degree of risk is essential to drive businesses forward. But technology can play a role in gaining better insights on data coming in from a myriad of sources – be it mortgage brokers, property appraisers, credit bureaus on the outside or asset and risk management systems on the inside – and offer a way to get a better handle on risky events before they erupt into the next crisis.