A couple weeks ago, the Association of Community Organizations for Reform Now (ACORN) published The Great Divide reporting on Home Purchase Mortgage Lending Nationally and in 115 Metropolitan Areas.

The report "looks in detail at the denial rates and ratios for borrowers of different races and incomes when they apply for ... home purchase loans" for 2002. One of our local papers has already published an article coverint the report and another interviewed us on Monday in preparation for their article.

The report is very interesting reading and is available at:

http://www.acorn.org/fileadmin/Community_Reinvestment/Reports/Great_Divide_2003_Main.pdf

Happy reading!

The problem with the ACORN report on housing-secured lending is that, like the FRB reports before it, the desired and intended premise was that some borrowers are denied credit merely because of race, ethnicity, and other prohibited factors. However, the report does not do a true comparison of borrowers' FICO scores, debt-to-income (DTI) ratios, loan-to-value (LTV) ratios, and other presumed factors which underlie the report premise. A white borrower with a FICO score of 775 and manageable DTI and LTV ratios would likely be granted an "A"-rate loan, while a borrower (of any color or ethnicity) with a FICO score of 582, and with high DTI and LTV ratios, would likely be denied credit. This is common sense. The industry is driven by investors, and the investors set the criteria for selection and purchase of loans. The real way to resolve the lending dilemma is through financial literacy training. Consumers need to better understand debt management and its affect on credit status and getting the best loan rate. The ACORNs of the world overlook this important issue. What banker today who is competent and who wants to promote optimal earnings would deny a borrower with strong credit, income and debt ratios simply because of the borrower's race, gender or ethnicity? It doesn't happen. Let's see a report of comparable FICO scores and ratios, and you'll see that ALL borrowers -- regardless of race or ethnicity -- are treated similarly.

The Acorn article says that Blacks with the same income as Whites were turned down more often.....but fails to mention why - our experience is that credit history and D/I played a big part in our denials to Blacks. We did not turn them down because of any particular race. Sorry - I feel a little defensive. We go over each denied file twice to ensure that the loan could not have been made.

And therein lies the discrepancies that arise from HMDA LAR analysis w/o the benefit of a file review. Fortunately with the changes in 2004 there will be more data on the LAR to be interpreted... or mis-interpreted as the case may be.

While I am sure they try to do good work, it isn't a full picture.

Excellent point, Andy!

This reinforces my comments in other posts on how important the knowledgeable use of demographic and CRA/HMDA data is.

Combined, this is extreamly useful information that allows compliance and marketing people to intelligently use the data that relates to their specific situation. It stops an institution from being painted with the broad brush of a national survey, and creates tons of business opportunities with existing customers.

Since the release of the study, 2 of our 3 major newspapers have done articles on the denial rates. The San Jose Mercury News did not name names, however, the Fremont Argus named us and the other large local lenders. The San Francisco Chronicle has been quiet (so far).

How about the rest of the country? Any articles you can link me to?

I'm sorry for you Al, that stinks.

Al, I feel for you buddy. It's been a while since I read the analysis but I felt sorry for some communities that showed 600+% increases in some very undesireable categories.

Many years ago an examiner team told our board of directors that we had a serious problem as we had a 100% declination rate for hispanics (Gee is that a race or ethnicity, sorry I digressed). This caused quite a furor among senior execs and board members demanding of me how I could be tolerate such unfair lending behavior. We had had 2 such applicants in the exam period!

The following year we did much better. We had 1 applicant and he was approved.

I still remember the somber tone of the examiner when he said our HMDA data had an undesirable "cluster" of loans in a middle income census tract. He thought it could be a real Fair Lending issue.

Yeah - we had FIVE loans to a mobile home park operator who purchased the units for their park managers. The mobile home parks were all in the same census tract.

SHEESH!

Misinterpretation of data is one of the fallicies of reports and statistics - like most consultant's opinion, they're geared to the desires of the person footing the tab.

Several years ago, we were involved in a DOJ inquiry because of faulty data analysis from Washington. After looking at 1000+ loans, DOJ quietly left. The charge - we turned down twice as many minority applicants as we did the prior year. The reason - we took three and a half times as many apps. (They didn't see that number.)