[Infowarrior] - Credit Score Is the Tyrant in Lending

[Richard Forno]

July 23, 2010
Credit Score Is the Tyrant in Lending

By JOE NOCERA

http://www.nytimes.com/2010/07/24/business/24nocera.html?_r=3&ref=business&pagewanted=print

The other day, a mortgage broker named Deb Killian called me, more or less out of the blue. Ms. Killian has been in the business since 1994. She and her husband run Charter Oak Lending Group, a small firm based in Danbury, Conn., that they founded in 1996. She is a member of the board of the National Association of Mortgage Brokers. By her estimate, she has closed more than 3,500 loans during her career.

Ms. Killian was calling because she was upset about one element of the mortgage underwriting process that, in her opinion, had gotten completely out of hand. That element was the borrower’s FICO score — in other words, his or her credit score.

Essentially, she says, a person’s credit score has become the only thing that matters anymore to the banks and other institutions that underwrite mortgages. Yes, the banks all mouth pieties about how credit scores are just one of many factors that go into their underwriting decisions. But every day she receives notices from banks and other lenders showing just the opposite: as they fiddle with the terms for one or another of their loan products (something they do constantly, by the way), invariably, the credit score is the dominant — and sometimes the only — criterion mentioned. Most of the time, needless to say, the minimum credit score needed to get the mortgages has been increased.

To make matters worse, she says, clients are having a difficult time just maintaining their current credit scores — even when they have done nothing to merit a downgrade.

“This is the example that drove me over the edge,” she said.

A woman seeking a mortgage had a credit card with a $3,000 limit. She had $1,500 worth of debt on the card, which meant, in industry jargon, that she had a 50 percent debt utilization. The client had moved the balance to a different bank, and that bank immediately lowered her credit limit to the amount she had borrowed: $1,500. Without taking on an additional penny of debt, the woman’s debt utilization had suddenly jumped to 100 percent.

Which, as Ms. Killian knew only too well, the FICO algorithm frowns on. Once that information made its way to the credit bureaus, the woman’s credit score dropped. And because it had, the interest rate on the mortgage she hoped to get increased. Ms. Killian had a hundred stories like that.

Yes, she told me, she knew that underwriting standards had been way too lax during the bubble. But to her mind, the mortgage lenders had swung too far in the other direction, depriving perfectly creditworthy borrowers of the chance to get a mortgage at a reasonable rate. “If the loan criteria says you need to have a 700 credit score, and you have a 699, you don’t get the loan,” she said. “It makes me nuts.”

Was I interested, she asked me finally, in writing a column about this problem?

I most certainly was.

•

“Mortgage brokers,” said Craig Watts, the head of communications at FICO, with a tone of bemused exasperation. “Some of them are kind of cranky these days.”

The mortgage brokers, he went on to say, were seeing things only from their own narrow perspective — the perspective of someone who wouldn’t get a commission if their clients couldn’t get a mortgage. Thankfully, from his spot high on the mountaintop in FICO-land, Mr. Watts could give me a broader, more sophisticated take on the topic. Thank goodness for that.

A FICO score, he patiently explained, is merely a tool that lenders use to help manage their risk; criticizing it is akin to criticizing “a saw because the construction job turned out badly.” With big banks making thousands of credit decisions every day, they couldn’t possible do it without some standardized benchmark; a credit score provided that benchmark. Over the years, he added, the algorithm had gotten very good at predicting the odds of a borrower defaulting.

In fact, FICO scores are not the best predictor. The amount of equity a person has in his home, his debt-to-income ratio, his job stability and his cash reserves are all better predictors than credit scores, according to Dave Zitting, the chief executive of Primary Residential Mortgage, a leading mortgage lender. And yet, he said, “The credit score  has become the line in the sand for the banks.”

It is easy enough to understand why, I suppose. During the bubble, Wall Street used credit scores to decide which subprime loans it would buy and securitize. The lower the credit score, the better, because Wall Street needed risky loans to generate yields that would entice investors. They pushed lenders to make loans to people with low credit scores, which became shorthand for risky loans.

In the aftermath of the bubble, credit scores have remained shorthand for a borrower’s creditworthiness — except that now borrowers need to have high credit scores instead of low ones. And yet, credit scores are no more accurate than any other risk model. There are people with low credit scores who are quite creditworthy. There are people with high scores who aren’t. Treating credit scores as if they were infallible — which is what the banking industry is now doing — is beyond foolish. It is hurting the recovery.

The two most important credit score hard-liners are Fannie Mae and Freddie Mac, which of course are currently wholly owned subsidiaries of the federal government. Because Fannie and Freddie are practically the only entities willing to buy and securitize mortgages, they have enormous clout; most lenders simply won’t make a loan if Fannie or Freddie won’t buy it. Their bottom line number is 620 — the company will buy mortgages only if the borrower has a credit score of 620 or above. Which means, given the current state of the mortgage market, that anyone with a score below 620 can’t get a mortgage. Even if that score is 619.

But the difference between a 620 score and a 619 is utterly meaningless. The credit scoring industry likes to make it sound as if their results are scientific; they’re anything but. It is not FICO that comes up with a borrower’s score — it just sells the algorithms. The companies that do are the big three credit bureaus, TransUnion, Equifax and Experian. They gather input about the prospective borrower’s lending history from various lenders like credit card companies and auto dealers, plug them into a formula and derive a credit score.

You would think, given the critical importance of an accurate score, that there would be rules about the information that is submitted to them. There aren’t. Lenders can submit information about your credit history to one of the bureaus, all of them or none of them. Some of them turn over information right away; some take months; some don’t do it at all. Some are sticklers for accuracy; others are sloppy. The point is that the credit score is derived after an information-gathering process that is anything but rigorous.

Or, rather, I should say, the three different numbers that are derived. Almost always there is a difference — sometimes a big difference — in the credit scores generated by the three bureaus. (Which, when you think about it, is another indicator of how haphazard the process is.) What happens then? I found a Web site called lendingart.com, which listed every big mortgage lender’s credit score requirements. The lenders all said that if two credit scores differed, they took the lower score; if there were three credit scores, they took the one on the middle. CitiMorgage, in its description, said that if a borrower had one score of 691 and another of zero, zero was the operative score.

Let’s go back again to that borrower trying to qualify for a loan that conforms to Fannie Mae’s criteria. Suppose one credit bureau has given him a score of 625 — which means he qualifies — and another gives him a score of 618, meaning he doesn’t. Then he doesn’t get the loan. Can someone explain how that constitutes sound underwriting?

And that doesn’t even get into the question of whether the prospective borrower is someone who once had credit problems and has now cleaned up his act — and his score is improving — or someone whose credit is in decline. The credit bureaus are incapable of tricking out that kind of nuance. Actually, they don’t really care. Nor do they take into account examples like the one Ms. Killian mentioned, where people’s credit scores are being hurt by credit card companies that are cutting back their credit limits.

And finally, they don’t take into account the many, many mistakes that are found in credit reports. My own credit reports, which I looked up for this column, are a case in point. Although my score was O.K. — the low 700s — the reports themselves were full of unpleasant surprises. They listed credit card accounts I didn’t have, and failed to list at least one big one that I did have. Two of them noted that five years ago, I was late on a car payment. (I was?) My daughter’s old Brooklyn address was listed as my former address. According to Experian, I was still writing for Fortune magazine. It said I no longer lived in a house that I just bought two months ago. TransUnion, meanwhile, listed The New York Times as my former employer. Currently, TransUnion said, I am an employee of Rite Aid.

Rite Aid? I know, I know — it is supposed to be up to me to catch their mistakes (which is also why they don’t have to care about the mistakes.) But what I find incredible is that we have imbued credit scores with these magical predictive powers — and yet the companies coming up with the scores can’t even get the borrower’s address and employer right. It would be funny if it didn’t matter so much.

This was the week, of course, that President Obama signed the financial reform bill into law, which calls for the establishment of a new consumer financial protection agency. The credit scoring business would certainly seem to be a worthy area for the new agency to dive into.

Wouldn’t you agree, Professor Warren?