Can A.I. Fix Income Calcs For Mortgages? Funny You Should Ask
Fannie Mae’s Lender Sentiment Survey
has foreshadowed the growing attention to automating the loan approval process.
That survey underlined the extent to
which lenders are viewing increased efficiency as key to reversing their
declining profitability expectations. Freddie Mac has talked about the growing
need for lenders to embed artificial intelligence (AI) into their internal
operations and now CoreLogic has jumped in with a specific area in which it sees
a need for streamlined software, borrower income verification.
CoreLogic researcher Brandon Brahms says
any underwriter knows that the processes to approve a borrower is inefficient
and time-consumer. While there have been technological improvements, the
process still requires some manual data entry, use of spreadsheets, and pencil
and paper exercises. Nowhere, he says,
is this truer or more painful for both underwriter and borrower than income calculations.
Underwriters have to consider many
potential variables – passive and portfolio income streams, salaries, raises,
bonuses, gig income, with multiple touch points and review cycles. Brahms says one has to consider the human
factor as well; different underwriters, given the same inputs, will come up with
different results as they interpret underwriting guidelines and calculation
processes in different ways.
He points to the solutions
that may be out there, already utilized in other applications, that may be
appropriate for lenders who still rely on old solutions. Do-it yourself tax
prep software, investment tools and other technology-based solutions are
guiding users to make informed decisions and could provide a potential
blueprint for the mortgage industry to follow.
Brahms says, “By automating the
collection of a borrower’s income data and standardizing a workflow for income
analysis and calculation, underwriters would benefit from accelerated
processing and more accurate and consistent results.” Smart software could suggest new sources of potential
income and identify missing documents, leading to smarter underwriting, saving
time and thus money.
Anything that smooths and speeds the
process makes sense. That said, it’s fair to wonder if Brahms’ blog article is also a bit of foreshadowing; that
CoreLogic will soon announce that they have developed just such a product that
will, in Brahms words, make “the clunky and inefficient processes underwriters
use to calculate a borrower’s income… distant memories – like landlines or
balancing a checkbook.”