Amid a national focus on redlining and other forms of racial inequality, the Consumer Financial Protection Bureau has identified persistent gaps in home-loan denial rates by race and ethnicity, but is stopping short of pointing to a pattern of discrimination.

The CFPB’s recent analysis of 2019 Home Mortgage Disclosure Act said denial rates are improving slightly across demographic groups, but historical racial disparities have not abated. Black homebuyers were denied loans last year at nearly three times the rate of non-Hispanic white homebuyers, roughly the same gap as 2018. The CFPB also found gaps by race and ethnicity in interest rates and refinance volumes.

But the CFPB, which released the preliminary results in June followed by further analysis in August, said the disparities require further study and left out any suggestion of discrimination. Against the backdrop of national protests over racial equity issues, some observers say the agency is being too timid. The bureau is now empowered to look at more HMDA data than it has before.

Critics say the agency’s cautious stance generally reflects how the Trump administration has undercut fair-lending efforts. Some suggested the CFPB sought to avoid a deeper analysis by claiming it was releasing the information as quickly as possible.

“It’s regrettable because lots of people have been waiting to see what this data shows about how much of the persistent racial disparities is explainable by facially neutral creditworthiness standards and how much is not, and also which … standards contribute most to disparities in pricing and denial rates,” said Diane Thompson, the founder of the Consumer Rights Regulatory Engagement and Advocacy Project and a former acting assistant director in the CFPB’s Office of Regulations.

The CFPB found that Black and Hispanic borrowers had notably higher denial rates last year than non-Hispanic white and Asian borrowers. Denial rates for conventional home purchase loans in 2019 were 16% for Blacks, 10.8% for Hispanics, 8.6% for Asians and 6.1% for whites, according to the 69-page report the CFPB released in June.

No conclusions, few referrals

HMDA data is regularly used by bank examiners in supervisory exams and fair-lending investigations. The 2019 data contained information from 5,500 institutions that originated about 8.1 million loans, a 26% jump from 2018, driven primarily by refinancings spurred by low interest rates. The HMDA represents roughly 88% of closed-end originations.

For the first time, the bureau was able to analyze non-public credit score data — long considered a key to unlocking the puzzle of disparities in lending — but the CFPB said more analysis and data is needed in order to reach any conclusions.

“The bureau is identifying disparities and cautioning that, in and of themselves, and even with the analysis by credit score, there’s not enough information to conclude that there is discrimination,” said Warren Traiger, senior counsel at Buckley, who analyzes HMDA data for financial institutions. “They suggest looking further at the data, but they don’t do so themselves.”

The release of the consumer bureau’s HMDA data in past years would spark a mad scramble by banks, mortgage lenders, advocates and policy experts to slice and dice the information, trying to determine whether financial institutions would be called out individually for discriminating against minority borrowers. Financial regulators use the data to refer potential redlining claims to the Department of Justice.

But such referrals and other fair-lending efforts have not been a priority under the Trump administration, sparking a backlash from Democratic lawmakers. CFPB Director Kathy Kraninger has said the bureau is actively enforcing fair-lending laws, but of the 40 DOJ referrals the CFPB has made since 2011 — peaking at 15 in 2014 — only one has come during Kraninger’s tenure.

But some observers say that under HMDA the onus isn’t on the agency to publicly call out discrimination.

“It’s not the CFPB’s job to prove discrimination, but to identify the disparities and make a referral to the Justice Department,” said Maurice Jourdain-Earl, managing director of Compliance Tech, a McLean, Va., software firm that specializes in fair-lending matters. “Their responsibility is to collect the HMDA data and make it available to the public.”

Still, Kraninger is also seeking to limit the data that lenders are required to submit under HMDA.

The CFPB is on the brink of issuing a new rulemaking that is expected to rescind some if not all of the 25 discretionary data fields issued in 2015 under former CFPB Director Richard Cordray. Dodd-Frank mandated 14 additional data fields, including the age and credit score of the borrower, points and fees, prepayment penalties and property value on top of the nine data points collected when HMDA passed in 1975.

Though denial rates for conventional have plummeted from a decade ago, the disparities are nearly identical to those found in 2013 data published by the Federal Reserve in the aftermath of the Great Recession. The CFPB provided only a descriptive analysis of the 2019 HMDA data, whereas the Fed conducted regression analysis. (Dodd-Frank transferred HMDA authority to the CFPB.)

‘A story to tell, but they can’t tell it’

The CFPB even suggested that other entities take up the mantle of analyzing the data for them.

“The Bureau hopes to provide the public with a roadmap for the new HMDA data, as researchers, government agencies, community groups, financial institutions, and others may use these new data for various other purposes,” it said in August.

Some observers say the CFPB missed an opportunity in the report to make a connection between racial disparities in mortgage denial rates and the nationwide focus on systemic racism after the killing of George Floyd.

“They are suggesting there is a story to tell, but they can’t tell it,” said Jason Richardson, director of research and evaluation at the National Community Reinvestment Coalition. “Essentially they are finding differences in denial rates and discrepancies in interest rates that they can’t explain, and they even say that the CFPB needs to do a deeper statistical analysis to explain them, but they’re not going to do that deeper analysis.”

Others say regulators could more adequately address racial disparities by focusing more holistically on the mortgage market, rather than highlighting conventional loans backed by Fannie Mae and Freddie Mac.

The CFPB’s data includes both conventional and nonconventional loans. But Jourdain-Earl questioned why the CFPB appeared to focus on denial rates for conventional loans when roughly 53% of Black borrowers get loans from Federal Housing Administration or the Department of Veterans Affairs.

“The real story here is if you limit the conversation to conventional loans, then you are excluding half of the loans that were made to Black and brown people who mostly got nonconventional government loans,” he said.

While denial rates tend to be a focus in the HMDA data, he notes that many minorities often are not even applying for home loans because of redlining, when no credit is available.

“In census tracts with people of color, there are fewer applications and fewer loans,” Jourdain-Earl said.

Credit score debate

Observers also said they were disappointed by how the CFPB used new data that was included in this year’s HMDA report.

For years, the Fed as well as banks and mortgage lenders claimed that racial disparities were attributed to factors other than discrimination, such as credit scores, debt-to-income ratios or property values — data that either was not collected or not publicly available.

Before 2018, credit scores were not collected in the HMDA data, but scores were added under a Cordray-era rule that went into effect that year.

While experts have waited for the credit score data in particular, the CFPB downplayed its importance.

“Credit scores, though important, are not the only factors used in lenders’ underwriting and pricing decisions,” the CFPB said in the 282-page report released in August. “Analyzing the denial decisions of mortgage underwriting should not be based on bivariate analysis alone that only examines the relationship between the underwriting decision and one single credit risk factor.”

Instead, the bureau suggests a “multivariate statistical regression” analysis is needed to explore the relationship between credit outcomes and a borrower’s credit characteristics, the report found.

Many HMDA experts had expected the CFPB to conduct its own deep dive given that the 2019 data included new and revised data points on creditworthiness that experts said should help explain persistent lending disparities over decades. But the bureau instead called for more data and analysis.

“Such analyses would require additional information, some of which is not available in HMDA data, and further, more sophisticated, analyses may be needed that are beyond the scope of this introductory article to 2019 HMDA data,” the bureau said.

Some experts were surprised that the CFPB claimed such an analysis was “beyond the scope” of its reporting authority given the agency has sole oversight of HMDA data.

“They drive home the point that credit scores by themselves don’t explain denial rates, but the data does show that whites are still less likely to be denied than Blacks even with the same credit score,” said Traiger, the Buckley senior counsel. “While the CFPB goes out of its way not to place the analysis in the context of the social justice issues that have topped everybody’s agenda since George Floyd’s murder, we fully expect advocates and enforcement agencies to dive into these differentials.”

Minorities pay higher interest rates

The bureau’s analysis found that Black and Hispanic borrowers pay roughly 25 basis points more through higher interest rates than non-Hispanic white borrowers for conventional home loans. For conventional loans, Black homebuyers paid a median interest rate of 4.375%, compared with 4.25% for Hispanic borrowers, and 4.125% for white borrowers, while Asians had the lowest median interest rates, 3.99%.

For the FHA, the interest rates are slightly lower, but borrowers pay more over the life of the loan in upfront and annual mortgage insurance premiums that can add more than 1.75% to the interest rate.

“This is disparate impact and it leads to lower wealth overall and a growing wealth gap between Black and white homeowners,” Richardson said. “Justifying it by saying that it’s because Black buyers have less wealth already, which manifests as lower credit scores and less money to buy down their rates, doesn’t resolve the underlying problem that plagues this system. This hurts all homeowners, it reduces tax revenue for local communities and it perpetuates the expanding racial wealth gap.”

Despite record low rates, the CFPB also found that the share of refinance volume to low- and moderate-income borrowers plummeted to 24% in 2019, from 30% in 2018, according to an analysis of the data by NCRC.

The share of refinance loans for Black borrowers fell to 5.3% last year from 6.2% in 2018, the CFPB data found, while Hispanic borrowers’ share dipped to 6.2% from 6.8%. In contrast, Asian borrowers’ share of refinance activity jumped to 5.4% in 2019 from 3.7% a year earlier. Non-Hispanic white borrowers’ share of refinance loans also declined slightly to 61% last year from 63.3% in 2018.

“It’s particularly troubling because the lower interest rates that have resulted in a massive increase in refinance activity are primarily benefiting white and Asian middle- and upper-income homeowners,” Richardson added.





Source link