Home loans in just 30 minutes. Closing in just five days.
It’s a great deal for some, made possible by an increasingly popular computer program that mixes your income with your credit history, and presto, a number pops out. Get a good number, get the loan–and speed up the closing process.
Technology marches onward, but some mortgage lenders are beginning to wonder if it also is trampling right over perfectly creditworthy people who, for various reasons that can’t be reduced to computer code, wind up with lousy numbers and a “no” from the loan officer.
The growing reliance upon computer-aided “credit scoring” brought three times the usual number of participants to a recent Colorado Mortgage Lenders forum.
Many were scratching their heads over a program that can reward a loan applicant with a wallet full of credit cards and a history of late payments, while penalizing those who shun plastic and pay their bills in cash, on time.
And some wondered if the new technology, by virtue of the assumptions built into the computer program, is giving minorities lower credit scores than whites–eroding gains lenders have made in recent years toward balancing a lending system that historically has favored whites.
“We have a lot of questions about credit scoring and how it affects the marginal borrower,” said Lori Richardson, assistant vice president of national retail production at Knutson Mortgage Co. in Denver.
Getting to the bottom of the matter, however, may be difficult because the company that wrote the computer program, Fair, Isaac and Co. (FICO) of San Rafael, Calif., says its formula is a trade secret.
Company officials say their intention is to make it easy for mortgage lenders to determine the risk of making loans. By doing the heavy number-crunching necessary, the computer program generates a score: the higher the score, the lower the risk.
Here’s how it works: Fair, Isaac used ten factors to analyze different segments of the population based on income, geographic area and credit history to determine who might default on a loan. Those trends are used to determine which loan applicants are likely to default. Among the factors included in the score cards are:
– The number of lines of credit open and whether they are paid off.
– How long the applicant has had credit.
– Whether any loans are from high-rate finance companies.
– How many credit cards the applicant has and how new they are. The software is designed to find applicants who take out new credit cards to pay the balances on their old cards.
– Any bankruptcies or foreclosures.
– How many inquiries have been made by the applicant to lenders. The software is designed to catch applicants who are applying for multiple credit lines to pay off debts.
Credit scores fall between 400 and 900. It is rare to see a score below 520, and a score of 660 or above is considered good.
Lenders give the program high marks for convenience.
“It’s a great way for us to summarize quickly who gets a mortgage without having to go over each late payment or credit card balance,” said Linnea Clayton, regional sales manager for Factual Data, a mortgage credit report company. “It rewards the good borrower, as it should, by removing some of the hassle when applying for a mortgage.”
“People who have good FICO scores can close on their homes in as little as five days,” she said. “Before FICO, it could have taken as many as 90 days.”
But speed wasn’t the issue among lenders at the Denver forum. Fairness was.
They brought to the table the example of a homeowner who wanted to move up to a bigger house after a promotion. His income was more than enough to cover the mortgage, but he had a history of late credit payments.
In the computer program, income trumped late payments and the applicant’s score of 706 qualified him for red-carpet treatment and a streamlined loan that closed in five days.
Then there was the case of a recently promoted secretary who shopped numerous lenders for her first mortgage. Her income, while adequate, provided less cushion. She paid her bills on time. But her credit cards were too new to suit the computer program, which also marked her down for checking with so many other lenders. Score: 584. Loan denied.
“The thing that bothers me,” Richardson said, “is that scores are adversely affected by things like how new your credit cards are or if you pay cash, or how many inquiries you have on your record.”
One Colorado Springs lender said credit scoring often becomes an exercise in trying to fit square pegs into round holes. Only 40 percent of her applicants have scores high enough to get mortgages with the best rates, she said.
“We are in an automated world,” said Caroline Marshall, executive vice president of Classic Mortgage. “If you fit into the window created by credit scoring–and from what I’ve seen so far, only 40 to 45 percent of us do–you have it made.”
A comparatively minor irritant is the fact that the scores become part of the borrower’s credit history, and can’t be changed. That means would-be borrowers whose low scores are derived from a faulty credit report must submit mountains of documentation to have the low score set aside.
More worrisome to lenders is the pressure they feel to rely on credit scores.
Because most lenders make their money by selling loans to investors such as the Federal Home Loan Mortgage Corp. (Freddie Mac), lenders must live by the rules the investors set. And investors such as Freddie Mac are reluctant to touch loans with credit scores below 620, unless the lender can provide costly documentation. Making loans to low-scoring borrowers can even trigger an audit by Freddie Mac.
More troubling still is a 1994 Freddie Mac report that black loan applicants were three times as likely as whites to have scores low enough to disqualify them for a mortgage. Hispanics were twice as likely as the general population to have low scores.
The study determined that borrowers with lower incomes, non-traditional sources of income, multiple jobs and loans from family members are more likely to be turned down for a loan. Although applicants with that kind of financial profile are more likely to be minorities, Freddie Mac concluded race was not a factor in the higher denial rate.
Fair, Isacc and Co. has spent two years trying to tackle these issues, said product manager David Shellenberger.
He conceded that women and minorities may fall through the cracks of a credit-scoring computer program written to accommodate the broadest definition of the market possible, even though the computer program contains no information about a borrower’s race or gender.
The company is continually tinkering with the program, Shellenberger said, to eradicate inequities. For example, the program now gives less weight to credit checks made by finance companies, which can bring a credit score down.
Shellenberger said that nationwide the program yields favorable scores about 85 percent of the time, a much higher rate than the 45 percent reported by the one Colorado Springs lender.
Above all, he said, a credit score should be only one factor in a person’s credit report, not the determining factor–especially in the case of borrowers who may not match up to the idealized profile but still are credit-worthy.
“The score is meant to be used as a guideline, not as a meat cleaver,” Shellenberger said.
None of which reassures State Sen. Joan Johnson, a Denver Democrat.
“I find (credit scoring) absolutely appalling,” said Johnson, sponsor of a new law that increases penalties on credit bureaus that maintain faulty information on consumers. She said she would have liked to address concerns about credit scoring.
“But I didn’t know it even existed when the bill was written,” she said. “After I found out about it and how many inaccuracies got into the credit scores, I wanted to include a provision addressing it but it was just too big an issue.”
Despite its critics, credit scoring appears to be here to stay.
“Quick approval is already here in the mortgage industry. You can’t do that without some sort of credit scoring and automated loan underwriting system,” said Steve Stingley, president of Colorado Springs-based Peoples Mortgage Corp. “If we can’t offer that, the consumer will find it somewhere else.”
