The housing crisis cost the nation about $13 trillion when tallying all losses from lawsuits, mortgage-backed securities litigation and taxpayer bailouts, a representative with fraud analytics firm Interthinx said.

Ann Fulmer, vice president of industry relations for Interthinx, made that revelation while speaking at the Mortgage Bankers Association's Risk Management and Quality Assurance Forum in Dallas. To put this $13 trillion into perspective, U.S. gross domestic product — the nation's entire output of goods and services — hit $15 trillion in 2011.

"We did not pay attention to data integrity on the way up, and so we have wiped out almost an entire year of gross domestic product in the United States," Fulmer said.

When looking at just REOs, short sales, RMBS suits and crisis-related litigation, the meltdown cost the mortgage/housing industry roughly $2 trillion, Fulmer explained. In addition, U.S. taxpayers and the government lost $3 trillion, while losses tied to homeowner equity reached approximately $8 trillion.

So how should data analysts search for potential fraud in the future to prevent overwhelming losses?

Fulmer advised attendees to look for areas of obvious risk when conducting forensic loan file reviews. In the future, parties may fudge on loan applications or other documents as they try to comply with new mortgage regulations from the Consumer Financial Protection Bureau and other agencies, Fulmer suggested.

For example, if a new loan is required to have a 20% down payment under the qualified mortgage rule in the future, scammers who cannot afford 20% down may try to manipulate that particular number. 

"If that becomes the standard, then that is going to be the problem area. People will be making it (the down payment) up or manipulating the cash value," Fulmer said. "And if you don't get that right then your loan-to-value ratio is going to be off."

The goal, she said, is to "follow the money" when dealing with compliance on the loan origination side. Areas to survey closely include down payment funding and sources, income and employment, assets and liabilities as well as collateral values.

"If you don't get it right up front then you have a defective loan," Fulmer told the crowd.

To get fraud detection and loan scoring right, the MBA panel overwhelmingly agreed that financial firms should be using automated compliance tools at the point of loan origination.