With interest rates rising and refinances tanking, a lot of lenders are searching for a bigger share of the home purchase market. The time-honored strategy is to call on local real estate agents and be the lender of choice for their customers looking to buy homes.
But SMR Research Corp., a Hackettstown, N.J.-based research and consulting firm, says it has developed a predictive model to help lenders bypass real estate agents and target directly the very people looking to buy homes and need a mortgage loan in the near future.
SMR’s new predictive statistical model rank-orders homeowners for their likelihood to buy a new house and need a purchase mortgage over any upcoming six-month time period.
“We think our scores will help,” says Stu Feldstein, SMR’s president and co-founder. “Using them, lenders can see which of their existing customers may be moving, and they also can prospect for brand-new customers.”
The company is currently scoring about 39 million households with the new model.
Feldstein says its tests show that the highest-scored households, totaling about one million, are more than three times more likely than the national average to buy a new home with mortgage financing over the next six months. The scores will be updated monthly.
SMR’s scores range from near-zero to as high as 10,000, Feldstein said, with 1,000 the national average likelihood to move.
SMR is working with a strategic data partner, which Feldstein described as a “large publicly traded company” but was prohibited from mentioning by name due to contractual restrictions.
How does SMR know who might be in the market for a new home?
Actually, SMR started by “working backwards” and identifying those people who are least likely to move and then backing them out, Feldstein said. Those include people who recently moved or refinanced their mortgages as well as people with reverse mortgages. Those people can be readily found in public courthouse records. SMR also used several other proprietary predictors to winnow out other people to further narrow the field.
It was then able to take the remaining population and rank-order them by other variables to predict those most likely to move.
For mortgage lenders, the company can further narrow the field to include only those homebuyers most likely to need a loan, for example, by excluding homeowners who already own their existing homes free and clear. If clients want to know who’s most likely to move for reasons other than providing mortgages, SMR also can score the debt-free homeowners.
George Yacik has been covering the residential mortgage business for more than 20 years and writes frequently for industry publications. He was vice president of SMR from 1992 to 2008. He can be reached at gyacik@yahoo.com.
