The Mortgage Credit Squeeze on Low/Moderate-Income Families and Consumers of Color

(Source: Habitat for Humanity)

As we continue to think about the aftermath of the 2008 housing crisis (particularly as we approach its 10-year anniversary) it's important to understand how the housing credit market has changed and who it's currently serving. The Home Mortgage Disclosure Act (HMDA), enacted in 1975, requires that mortgage lenders report to the federal government information about the loans they make and the characteristics of the borrowers who receive credit. The latest annual data collected through this law tells an interesting story about homeownership in America and what populations are being served (and which are not) by mortgage credit markets.

Recently the Center for Responsible Lending (CRL) released an analysis of the 2015 HMDA data and shared some interesting observations. While credit is certainly flowing, it's not flowing in all directions:

Black and Hispanic families, as well as families with low- to moderate-incomes are struggling to obtain mortgage loans. CRL notes in its commentary that although large banks and financial institutions continue to benefit from cheap and easy access to U.S. Treasury funds, the vast majority of the already meager amount of credit flowing to these underserved groups has mostly come in the form of government-backed loans. Conventional loans (i.e., those not backed by the FHA or some other government program) seem to be mostly going toward servicing the needs of white borrowers. Further, consumers from communities of color and low- to moderate-income families are paying more on the whole for mortgage credit. And interesting, half of all new mortgage loans (both purchase money and to refinance) have been originated by non-depository intuitions (i.e., insurance companies, mutual funds, investment trusts etc. - rather than traditional community or even national banks) Cribbing from the report, here are some high points:

Highlights from the 2015 HMDA data include: 

• The share of loans made to African-American and Hispanic borrowers in 2015 rose modestly, but remained well below the population share they compose. In 2015, African-Americans received 5.5 percent of loans up from 5.2 percent in 2014; Hispanic borrowers received 8.3 percent up from 7.9 percent in 2014. These percentages fall far short of the share of the U.S populations that these groups represent. African-Americans compose 13.3% and Hispanics 17.6% of the total national population.
• The share of loans made to low and moderate-income borrowers rose slightly in 2015 to 28 percent from 27.1 percent. Although modestly higher than the share in 2014 the 2015 share was lower than it was from 2009 to 2013.
• Most home purchase loans made to African-American and Hispanic borrowers continued to be through government insured programs (including FHA, VA and others) and reliance on these programs continued to increase. In 2015, 70.2 percent of loans to African-American borrowers and 62.6 percent of loans to Hispanic borrowers were government backed. These shares compare to just 36.0 percent of loans made to non-Hispanic white borrowers.
• In keeping with recent trends, a very small share of conventional home-purchase loans were made to African-American and Hispanic borrowers. In 2015, just 2.7 percent of conventional home-purchase loans were made to African-American borrowers and just 5.1 percent of these loans were made to Hispanic borrowers. These percentages are virtually unchanged from the levels in 2014.
• The share of African-American and Hispanic borrowers that received “higher-cost” loans fell dramatically from 25.6 and 28.4 percent respectively in 2014 to 16.2 and 18.5 percent respectively in 2015. This is a result of changes in the cost structure of government-insured loans (including FHA). The shares, however, remained well above the share of non-Hispanic white borrowers that received higher-costs loans, just 6.2 percent in 2015
• The share of both home-purchase and refinance loans made by non-depository lenders continued to increase. In 2015, 50 percent of all first-lien owner-occupied home-purchase loans were made by non-depository mortgage lenders. This share has been increasing in recent years and is the highest level since 1995.

In looking at this data, one question that comes to my mind is whether the new underwriting standards imposed by the Dodd-Frank Act might have anything to do with the credit squeeze on low- to moderate-income families. And, in turn, might that have anything to do with the disparity in lending to communities of color. For the uninitiated, the Dodd-Frank Act imposed upon all originators of mortgage credit (whether banks, brokers, financial institutions...whoever) an obligation to ascertain a borrower's "Ability-to-Pay" prior to making a loan. The law (and accompanying regulations which are part of Regulation Z of the Truth in Lending Act) provide a number of factors that an originator must use to make that determination (current or expected income and assets, current employment status, projected monthly mortgage payments, other property related monthly expenses, debts and competing financial obligations, credit history and monthly debt-to-income ratio). As long as the originator makes a "good faith" determination that the borrower can afford the loan, then its obligation to assess the "Ability-to-Repay" has been met.

But, true to form, lenders do not like risk. The chance of being found to have failed to make a good faith, ability-to-repay determination looms large (especially since such a finding can have severe consequences). Because of this, most all lenders are opting instead to make loans that meet the Dodd-Frank Act's safe-harbor. These are loans that, because they must meet such strict underwriting/low-risk guidelines, are deemed to meet the good faith ability-to-repay requirement (the safe-harbor comes in non-rebuttable and rebuttable flavors--depending on the risk tolerance of the bank, although both are pretty low-risk overall). For more on the requirements for these loans, you can check out my article The Unfinished Business of Dodd-Frank: Reforming the Mortgage Contract forthcoming in the SMU Law Review here. These loans are called "qualified mortgages" and, unsurprisingly, mortgage lenders have embraced them in a big way. Data from the National Association of Realtors quarterly survey of mortgage originators reveals that over 90 percent of all mortgage loans in the second quarter of 2016 were qualified mortgages. The chart below indicates that, with some small variance, this trend has been consistent since the first quarter of 2014. Compare this to the share of non-qualified mortgage loans (i.e., those that do not meet the safe harbor, but rather merely meet the good faith ability-to-repay requirement): less than one percent and declining.

So, with that bit of (or too much) background, how might these heightened underwriting guidelines (and mortgage originators' love for the qualified mortgage) help explain the 2015 HMDA data above? Well, one of the big requirements for a loan to be a "qualified mortgage" is that it cannot result in the borrower's debt-to-income ratio going above 43 percent (i.e., all monthly debt obligations, including the mortgage payment itself, cannot eat up more than 43 percent of the borrower's income in a year). On the one hand, that means that the debt will likely be more manageable and a default less likely. On the other hand, it doesn't allow for a person to borrow very much money to buy a home. When you think about student loans, auto loans for a two-car family, and various other expenses that form part of the slings and arrows of everyday life for low- to moderate-income American families, it's pretty easy to get to 43 percent (to say nothing of the lack of affordable housing in this country). So, in order to make the math work the borrower obviously has to have more income to drive the ratio down. That's a problem for low to moderate-wealth families. That's where race and ethnicity come into play. The typical black household has a mere 6 percent of the average wealth of a white household, and Hispanic families do just a little better at 8 percent. To put some hard numbers on that, take a look at the Census Bureau chart below, showing household incomes by race and ethnicity over the period from 1967 to 2014.

So if there is any correlation (or even causation) between the 2015 HMDA data and the way mortgage originators are dealing with Dodd Frank's mandatory underwriting requirements, the next questions becomes what to do about it. In essence this is a natural tension - one between access to credit on the one hand and safety and soundness on the other. Government regulators are concerned that lending institutions need to be more robust in their underwriting to make sure individuals are not taking out loans for which they can never hope to repay. On the other hand, homeownership is still an obsession in the United States and policy makers generously reward those who obtain it - therefore making it very important for individuals to be able to acquire a home (in other words, have access to mortgage credit). Reconciling this tension and striking the right balance is something that housing finance reformers have yet to truly tackle (see, e.g., the perpetual receivership of Fannie Mae and Freddie Mac). And there's also the real and ever-present role that lending discrimination plays in the housing finance market (a role that has been well-documented over the years).

In any case, the first step in addressing some or all of these issues is to see the problem and begin to discuss what might be causing or contributing to it. Unfortunately, neither of the candidates have made housing policy reform (prior DNC and RNC commentary provided) much of a tangible issue in this presidential campaign. We'll see if that changes when one of them gets to the White House.