The Big Banks have suddenly decided to move traditionally underserved borrowers into loans guaranteed by the Federal Housing Adminsitration (FHA).

For years, FHA lending made up only a small fraction of mortgage loans. One source quotes a Wells Fargo representative who says that FHA lending made up only three percent just a few years ago.

The main difference between FHA loans and conventional products is that there is a lower down payment requirement. FHA asks for a downpayment of at least 3.5 percent. Conventional loan terms fluctuate, but in the past year most borrowers have had to put up at least 20 percent in order to buy a home. For investors the number has been much higher.

FHA counters the risked posed by higher LTVs by requiring borrowers to purchase mortgage insurance at the outset of the loan, and to pay a premium for new insurance for as long at the LTV remains at a certain level. The result is that FHA loans, while bearing the same interest rate, will usually have a much higher APR.

There are other appeals for an FHA loan. Third parties (home builders, relatives, employers, non-profits) can contribute to the down payment. On the other side, sellers are allowed to contribute to closing costs.

Things have changed dramatically since the credit crisis. One of the interesting things is that there has been an improvement in the kinds of borrowers that use FHA. According to HUD, the average credit score for an FHA loan  went up from 621 in early 2008 to 692 by the middle of 2009. Do the math: that is a gain of more than 70 points!

Some lenders appear to have made a strategic decision to make loans to underserved groups through FHA. The table that follows lists twenty banks from the S&P 900, along with the share of loans that they originated through FHA. There are four segments: minority borrowers, low-income borrowers, low-income neighborhoods, and rural loans.  

The Big Banks are Making their Loans through FHA

The Big Banks are leading the charge. These top ten are most of the largest banks in the country. Citigroup is an absence, but they have pulled back in mortgage lending.

This analysis includes both FHA and conventional loans on site-built homes. Only loans by owner-occupants are included. There are no home equity loans here – only purchase and refinance loans.

Other observations:

Some banks rely on FHA in rural areas. Fifth Third scored high on this list because they are so dependent upon FHA outside of the cities. It is the same story with M&T Bank. Astoria Financial has a high score, but it can be discounted due to sampling size.  Almost all of AF’s loans are made in the New York metro area.

The big bank that isn’t relying on FHA? BB&T. The Best Bank in Town makes plenty of loans, but they’re not going after the FHA borrowers. For a long time, BB&T has insisted that they are happy to walk away from loans. Maybe their thinking is that they don’t want to let a guarantee push them into making a loan that they otherwise wouldn’t originate.  

A few banks are making almost all of the FHA loans. When the numbers show that JP Morgan, Bank of America, and Wells Fargo are active in FHA than are other banks, then it follows that their share of the market is very large.  These three banks accounted for 79 percent of FHA loans to minority borrowers.  Throw in SunTrust and Fifth Third, and five banks make 90.3 percent of FHA loans to minorities. For low-income borrowers and loans made in low-income neighborhoods, the top five’s share is nearly as dominant – 85.7 percent and 83.5 percent.

This leads into another issue. Rural lending is not the province of small town banks. There are about 60 lenders in this analysis. The big five had 73 percent of all FHA in rural areas, and 55.6 percent of all conventional loans.

In honor of Veteran’s Day, Bank Talk is going to take a look at the market for loans guaranteed by the Veteran’s Administration.

One of the benefits set aside for our Veteran’s is the opportunity to utilize VA financing to buy a home. There are some significant advantages. For one, VA loans can still be had without a down payment. There’s also the psychological benefit. While many people enter a bank branch unsure if they will be well-received by a lender, veterans can feel secure in knowing that there is a designated program to help them.

I’ve pulled data for all 2009 VA loan applications reported by institutions that are regulated by the Office of the Comptroller of the Currency. The OCC regulates national banks. Any lender where the acronym “N.A.” is appended to their name is an OCC bank.

The VA market is no different from the rest of the field. Most VA mortgages come from just a few lenders. (more…)

Supporters of the new “neighborhood schools” plan for Wake County, North Carolina might assume that families will be able to “vote with their feet” and move to a school of their preference.

The data hints at inequity in how mortgage loans are made, but that is not really the point that I am trying to make. Nor is this a conclusion that should be seen as limited to Wake County. In most districts, school funding is at least loosely correlated with property values.  That means that it is a norm for well-off neighborhoods to have better schools.  Thing is, it doesn’t have to be that way. The reason that I am looking at Wake is only because they’re doing their best to develop a problem.  Since the most recent school board election, their district has been dismantling its old assignment plan. That system sought to put a ceiling on the concentration of poverty in the student populations of its schools.  Now, they are redistributing under a new plan that is loosely described as “neighborhood schooling.”

Critics argue that schools in low-income areas will gradually succumb to the challenges of the plan.  Data show that low-income students perform at better levels when they have a socioeconomically diverse set of classmates. For years, that insight has guided the reassignment policy. It meant that Wake set a ceiling on the share of low-income students that could be assigned to any school. The results were evident: Wake was recognized as a leader in urban school policy.

Unfortunately, the truth is different. CRA-NC has studied mortgage lending in Wake County for three years, from 2006 to 2008.  Our results show that minority families are obstructed from moving to upper and middle class neighborhoods.  They have a harder time getting a loan than white families do, at all points of the economic strata.  Household income does not make much of a difference.  In fact, the greatest discrepancy in access to mortgage loans is between upper income white and upper income minority borrowers.

Here are a few data points that I am working on right now:

• When low income minorities try to buy a home in a middle income neighborhood in Wake County, they are denied for a conventional home purchase loan at a rate that is 2.25 times greater than the denial rate for white borrowers with similar income profiles.
• When low-income minorities (income less than 80 percent of the average for Wake County) try to buy a home in an upper income neighborhood (greater than 120 percent AMI), they are denied at a rate that is 1.88 times higher than white borrowers with a similar profile.

Even more interesting are the numbers that show how minority and African-Americans are challenged to buy into well-0ff neighborhoods, regardless of income.  I switched my focus from low-income African-Americans to upper income African Americans.  Now we are focusing on only African-Americans with incomes greater than 120 percent of the average.  This is well-off group.  To qualify for this group, a household would have to have an income of at least $84,000.

The picture is even more troubling when minority borrowers are filtered to include only African-American borrowers.  Then, the same comparisons become even greater: 3.28 times more difficult for upper income black borrowers to buy a home in a middle income neighborhood, and 3.63 times more difficult to buy a home in an upper income neighborhood.  Translated simply, this says that even white black households make a lot of money, they have hurdles to go with banks.

This data is from 2008*.

The discrepancies were even greater in 2007: very low income African-America borrowers that applied for a mortgage to buy a home in an upper-income neighborhood were turned down at a rate that was almost five times greater than that of white borrowers with a similar profile.

Research shows that even as incomes between white and African-American households are more equivalent now than there were in the 1960s, there is still quite a lag in terms of assets. The Pew Foundation reported that white households had between 12 and 13 times more in assets, including in home equity, than did African-American households. The ratio was slightly less for Hispanic households.

All of this should underscore that residents will not be able to “vote with their feet” in response to the new neighborhood schools policy. If people try to move to a good school, it may be not be so easy.

We don’t want to consign anyone to being trapped in a bad school.  It is contrary to having an opportunity society, and it makes no sense for the ongoing economic development of our community.

*single-family homes, home purchase only, conventional loans for owner-occupied properties.

Perhaps the difficulty that African-American borrowers experience when they apply for a loan is not merely about race, as many have suggested, but also about the economic differences between men and women.

Advocates often compare how capital is allocated within communities.  Who is getting loans, and who is not? The income status of a borrower is very relevant, as is the income status of the surrounding neighborhood.  HMDA data is set up to allow for those kinds of questions.  We also look at borrower race.  HMDA data can provide answers for many of these questions, too.

I decided to use one of the lesser-utilized data points in HMDA to do some initial descriptive research about how race and sex interact in the mortgage marketplace. HMDA allows applicants to indicate their sex and the sex of their co-applicant. Most borrowers fill out the data. I looked at a set of originated loans made in Charlotte, North Carolina in 2008.

This is just an initial take and the findings are only descriptive. I can’t draw any inferences for policy with these numbers because there are so many other variables that aren’t captured here.

In loans originated by African-American borrowers, women are almost as likely as men to be listed as the lead applicant. Women applied for 3,306 of the 6,713 mortgage loans taken out by African-Americans in Charlotte in 2008.  That is about half of all new loans. In instances where white borrowers made loans, women made only 30.4 percent of the applications. In most of the instances when an African-American woman took out a loan, there was no male co-borrower.  Only 605 of those 3,306 loans listed a male co-borrower (18 percent), and in 2,612 instances there was no other applicant at all.  The rest are instances when the other borrower was also female, or when the application was taken over the phone and the sex status of the co-borrower was not known.  White woman are more likely to have the support (29.1 percent) of a male co-borrower, although the numbers show that it is still an exception rather than the rule for both.

African-American women are more likely to apply as single heads of households, and minority neighborhoods have a greater share of single head of household families.

It is even true that the share of African-American women seeking a loan that are single grows higher in higher-minority neighborhoods.  The discrepancy grows in minority neighborhoods.  In communities where more than half of all residents are African-American, women 52.8 percent of all approved applications.  The prevalence of women as lead household financial decision-makers is more prominent in minority neighborhoods, regardless of borrower race.  White women were more likely to file for a mortgage without a co-applicant when they lived in a majority-minority neighborhood as well.

Women-only borrowers are taking out modest loans.  The average loan amount was much less – just $154,000, compared to almost $190,000 – in African-American households.

I don’t have the numbers, but I believe that there is a lot of data out there to suggest that women are paid less than men, after controlling for job type.

The data includes first-lien loans on owner-occupant single-family homes purchased with a conventional prime loan in 2008.

Piggy-back loans (second lien) were one of the emblems of the go-go days of subprime lending.  Lenders took the risk, perhaps because buyers of mortgage-backed securities had the appetite, to make loans that were vulnerable to default. The results were ruinous.  No one planned for 11 percent unemployment.  No one planned for a time when home values would drop by more than 40 percent in one year.

“Piggy-backs” (also called “80-10-10s”) played a big role in the end of a number of lenders: Washington Mutual, ownit.com,and E-Trade are among some of the institutions felled by these loans.

It should have been a warning to regulators that people were choosing to substitute a payment to a private mortgage insurer with new debt on a second mortgage. Here is a link to a “mortgage advisor” who suggest that (more…)