Data-Quality Matters: Lenders Today Face Growing Pressure to Ensure That the Quality of Their Loan Data is Good. Here are Some Things to Think About to Ensure Better Data Quality (Technology‪)‬

For as long as mortgage banking has existed, there have been data-quality and integrity issues. As one executive with a megalender once said, "There are only two types of data in the mortgage business: dirty data and very dirty data." The list of where the data problems come from is lengthy, and includes automated and manual processes. * Here in the 2009 mortgage market, data quality is ever more important and under intense scrutiny. As a result of the market downturn, the secondary market has had to ask originators to buy back bad loans based on, among other things, data-quality issues--thus creating a direct business impact on the industry. * This article is based on work by the Mortgage Bankers Association (MBA) to address data quality head-on from a lender and secondary market perspective. Yet it is important to point out that while data-quality problems exist throughout the mortgage universe, we are only choosing to focus on a specific area in need. * Over the years, the mortgage process has become increasingly automated. Manual, paper intensive processes have given way to dynamic and demand-driven workflow automation. Introducing new forms of technology and automation does not always improve data quality and security; in fact, complexities of new systems often do the opposite. * At the same time, regulation and compliance requirements have been rapidly increasing, which has introduced more complexity and additional technology systems to typical operations. An example is the ever-expanding methods by which lenders obtain leads/applications. Twenty years ago, people predominantly went into a bricks-and-mortar establishment and filled out a loan application with a loan counselor (and many still do). Today, you have Internet-direct consumer Web sites, aggregators, call centers, broker networks and many more methods of obtaining the initial loan application data--all of which could introduce data-quality problems into the process. We have moved forward technologically, but backward in terms of increased complexity. The potential for data-quality issues without increased investment in tools and processes to obviate them is tremendous.