I was a bit under the weather last week, and turned to my favorite service: Netflix. I had the “Inside Job” DVD lying around, and thought it might put me to sleep. Quite the contrary…it’s a very good and many times infuriating movie about the financial meltdown.
Then I found this news clipping from Huffington Post on the US Justice Department using Deutsche Bank AG’s subsidiary, MortgageIT for at least $1 billion for “defrauding taxpayers by repeatedly lying to a federal agency when securing taxpayer-backed insurance thousands of shoddy mortgages.” (Deutsche Bank contends that most of the questioned activity happened before their acquisition of MortgageIT. That’s a whole separate risk topic.)
Essentially, the Justice Department claims that MortgageIT did not follow the Federal Housing Administration’s (FHA) requirements for validating loan information before it funneled the mortgages to the Department of Housing and Urban Development (HUD). Once funneled to HUD, the US government is responsible if the mortgages go bad.
My data antennae went right up…sounds like data integration and checking for consistencies across fields to me…let’s check.
Here’s a memo from a fraud division that went to HUD that lays out some of the rules: An Underwriting Review of 15 FHA Lenders Demonstrated That HUD Missed Critical Opoprtunities To Recover Losses to the FHA Insurance Fund. The memo outlines these areas:
- Income/employment history
- Qualifying ratios
- Gift funds
- Credit history
- Rent verification
- Borrower investment
- Skipped mortgage payments
Let’s focus on Qualifying ratios and Credit History. Before we dive in, there are of course baseline data integration, data quality, and information governance requirements for these two categories. Proper information must be identified, found, and integrated. Data Stewards must be in place to interpret guidelines into implementable information policies.
(These definitions come from the FHA/HUD memo listed above.)
Qualifying Ratios: “Effective April 13, 2005, the mortgage payment-to-income and total fixed payment-to-income ratios were increased from 29 and 41 percent to 31 and 43 percent, respectively. If either or both ratios are exceeded on a manually underwritten mortgage, the lender is required to describe the compensating factors used to justify the mortgage approval.”
EIM impact: A cross-column validation solution seems like it would help here. In the first example, Estimated Payment must not be more than 41% of Income. The Data Steward needs to interpret this policy and document it. Then the policy needs to be executed when a mortgage is proposed. With these policies in place, compliance could definitely be automated. Automated also gets you auditability and clear visibility via reporting/analytics. These features would have been in invaluable in this scenario—not only to ensure proper compliance, but to defend yourself against claims of non-compliance. Here, SAP tools like Data Services, Information Steward, and Business Intelligence can help.
Credit History: “HUD requires the lender to consider collection accounts in analyzing a borrower’s creditworthiness. The lender must explain all collections in writing.”
EIM impact: Again, a cross-column validation solution seems like the ticket. First verify that a credit check was done. Then interrogate the credit check to determine if any accounts have gone into collection. Finally, verify that there is a Collection Reason filed filled out (with more than n/a or tbd…perhaps on minimum length?) for each Collection Account identified. As in Quality Ratios, these checks could be automated, with all of the auditability and visibility features, too. No surprise that SAP tools like Data Services, Information Steward, Business Process Management, and Business Intelligence can help again.
Finally, I bet Deutsche Bank wishes it could have run some of this analysis before acquisition of MortgageIT, too. Instead, they are stuck fighting an expense lawsuit with the Justice Department. Ouch.