Data quality: the cost of dirty data in the secondary market.(Tower on Tech)

The problems of inconsistent data names, definitions, structure, cleanliness, usability, transfer and governance are well known in the mortgage industry. We subsist on a steady diet of missing, incomplete and inaccurate data elements for loan origination, loan servicing, portfolio management and securitization processes.

Dirty data persist despite the proliferation of automated computer systems for core lending processes and subprocesses. Similar to a "water-bucket brigade," where water is lost as water buckets pass from hand to hand, duplicative legacy systems, the lack of one database of record and less-than-seamless integration between lending subsystems result in lost, inaccurate and manually re-entered data. This results in slow, expensive loan processing; weak underwriting decisions; inaccurate loan pricing; excessive quality-control costs; incorrect portfolio management; loan buybacks; and other costs to lenders and mortgage investors.

In his May 2007 Mortgage Banking column entitled "Data Quality: Crucial for Every Organization," Gabe Minton discussed the benefits of good data quality--first by defining it, then by describing progress made defining data field names, definitions and structure using MISMO[R] guidelines. He then addressed the …