Complexity and lack of standardization limit CREF technology

Register now

Standardized documentation and single-family lending are a natural pairing. Or at least a much more natural pairing than exists with commercial and multifamily, where the idea of standardization bumps up against a complex set of variables.

"The appraisal for a residential mortgage is probably four pages long but the appraisal for a commercial property on average is about 200 pages long," Paul Orlando, executive vice president and chief information officer at Flagstar Bank, said in an interview. "The residential mortgage has a very standardized format and the agencies — Fannie Mae and Freddie Mac — mandated that paper file and it's accompanied with data extraction."

Which is not to say that the industry wouldn’t benefit from some standardization. The question is, how to get there? Might it be possible to pull some of the processes in single-family over to the commercial side of the business?

Orlando thinks so. For example, while in residential lending, tax returns are used to model past and future income calculations for borrowers, a similar system could be applied to extract data and automate cash flow analysis generated by businesses that support commercial real estate.

A possible next step, he said, would be to train artificial intelligence engines to look at the various documents of commercial lending and automatically recognize specific data points. With content imaging and data extraction, the time it takes to file away and reference the forms would get dissected and repurposed for analysis.

Looking for more digital mortgage innovation?

"If they take all the contracts in place on a strip mall, for example, each individual business is going to have its own contract," Orlando said. "You need to read through all those contracts to see how much the rent is, how much is getting paid upfront for build-outs. The AI data extraction has come a long way so you can set them up to look at those documents and find specific things."

Of course, all of this would require a commitment and capital investments that the industry has so far seemed unwilling to make.

Orlando, who will be speaking at the MBA Technology Conference in Dallas on Monday, says none of that will change unless key segments of the industry press for it. CREF lenders need to market the value proposition within their sector of the mortgage industry so vendors become aware of the existing opportunity and focus needed. Because even AI needs to be told it’s supposed to be helping optimize pieces of the commercial lending process.

For reprint and licensing requests for this article, click here.
Mortgage technology Artificial intelligence Digital mortgages Commercial mortgages Commercial real estate lending Multifamily Flagstar Bancorp Fannie Mae Freddie Mac