Does Borrower and Broker Race Affect the Cost of Mortgage Credit?
Brent Ambrose, James Conklin & Luis Lopez
Review of Financial Studies, forthcoming
We test for pricing disparities in mortgage contracts using a novel data set that allows us to observe the race and ethnicity of both parties to the loan. We find that minorities pay between 3% and 5% more in fees than similarly qualified whites when obtaining a loan through the same white broker. Critically, we find that the premium paid by minorities depends on the race of the broker. We also examine recent policy changes around broker compensation rules that may not only reduce these price disparities but may also limit access to credit for minorities.
Effect of Predictive Algorithms on Home Prices and Racial Biases in the Housing Market
MIT Working Paper, April 2020
In this paper, I investigate how predictive algorithms change market outcomes in the housing market. By focusing on the effect of the Zestimate home valuation on housing markets, I investigate to what extent the market participants rely on the Zillow’s estimated market value when making the transactions and how it interacts with other information sources. I collect detailed property transaction information for 120,482 properties sold between May 2017 and May 2019 in the Greater Philadelphia Area and the preliminary results suggest that sale prices tend to be higher for those properties that have a higher estimated home value on Zillow.com. Evidence shows that people use this statistic as a summary of the transactional data and rely more on it when it is harder to process the information. Moreover, the Zestimate helps improve racial disparities in the real estate market by providing all sub-populations with less biased information.
Partisanship in Loan Pricing
Ramona Dagostino, Janet Gao & Pengfei Ma
Indiana University Working Paper, September 2020
We document a strong effect of lender partisanship on corporate loan pricing. Using novel data on the voter registration records of bankers in charge of originating large-scale corporate loans, we find that bankers who are registered with a different party from the one represented by the president of the United States ("misaligned bankers") charge 7% higher loan spreads compared to bankers affiliated with the same party as the president. This effect is not explained by bankers’ innate characteristics, borrower fundamentals, or bank-level policies, but is consistent with misaligned bankers having a more pessimistic economic outlook. Despite charging higher interest rates, misaligned bankers do not seem to generate higher revenue than aligned bankers.
Does Increasing Access to Formal Credit Reduce Payday Borrowing?
Sarah Miller & Cindy Soo
NBER Working Paper, September 2020
The use of high cost “payday loans” among subprime borrowers has generated substantial concern among policymakers. This paper provides the first evidence of substitution between “alternative” and “traditional” credit by exploiting an unexpected positive shock to traditional credit access among payday loan borrowers: the removal of a Chapter 7 bankruptcy flag. We find that the removal of a bankruptcy flag on a credit report results in a sharp increase in access to traditional credit and raises credit scores, credit card limits, and approval rates. However, despite meaningful increases in access to traditional credit, we find no evidence that borrowers reduce their use of payday loans, and our confidence intervals allow us to rule out even very small reductions in payday borrowing. Furthermore, we find evidence that flag removals increase the use of other alternative credit products such as online subprime installment loans. These results indicate that marginally improving access to less expensive formal credit is insufficient to meaningfully shift borrowers away from high cost subprime products. We discuss likely explanations for this including increased marketing of subprime products associated with the flag removal, the imperfect substitutability between cash and credit for low income borrowers, and an insufficiency in the size of the increase in credit access associated with the flag removal.
The Housing Boom and Bust: Model Meets Evidence
Greg Kaplan, Kurt Mitman & Giovanni Violante
Journal of Political Economy, September 2020, Pages 3285-3345
We build a model of the US economy with multiple aggregate shocks that generate fluctuations in equilibrium house prices. Through counterfactual experiments, we study the housing boom-bust around the Great Recession, with three main results. First, the main driver of movements in house prices and rents was a shift in beliefs, not a change in credit conditions. Second, the boom-bust in house prices explains half of the corresponding swings in nondurable expenditures through a wealth effect. Third, a large-scale debt forgiveness program would have done little to temper the collapse of house prices and expenditures but would have dramatically reduced foreclosures and induced a small, but persistent, increase in consumption during the recovery.
Public Disclosure and Consumer Financial Protection
Yiwei Dou & Yongoh Roh
NYU Working Paper, July 2020
The U.S. Consumer Financial Protection Bureau has accepted complaints about banks’ financial products and services since 2011 and has released the complaint database to the public since 2013. We analyze the effectiveness of this public disclosure in protecting mortgage borrowers. We find a greater reduction in mortgage applications to banks that receive more mortgage complaints in local markets after the disclosure. The effect is stronger in areas with more sophisticated consumers and higher credit competition, and for banks receiving more severe complaints. The number of monthly mortgage complaints per bank exhibits faster mean reversion after the publication of the database. Our findings suggest that the public disclosure of banks’ provision of inferior products and services enhances product market discipline and consumer financial protection.
Low‐Income Homeownership and the Role of State Subsidies: A Comparative Analysis of Mortgage Outcomes
Erik Hembre, Stephanie Moulton & Matthew Record
Journal of Policy Analysis and Management, forthcoming
Between the late 1970s through 2013, state Housing Finance Agencies (HFAs) financed nearly $300 billion in mortgages to low‐ and moderate‐income first‐time homebuyers. Descriptive evidence indicates that HFAs help households retain their homes at higher rates than similar households purchasing homes in the private mortgage market. Using a matched sample of HFA originations between 2005 and 2014, we estimate a multinomial logit model of mortgage default (or foreclosure) and prepayment. We find that HFA borrowers are about 30 percent less likely to default or foreclose on their mortgages than otherwise similar non‐HFA borrowers. We find that 37 percent of this HFA effect can be explained by HFA origination and service delivery practices including direct servicing and homeownership counseling.
Is COVID Revealing a CMBS Virus?
John Griffin & Alex Priest
University of Texas Working Paper, August 2020
Commercial loan valuations crucially depend on accurate loan income, but we find that underwritten income is commonly overstated when compared to actual property income. Consistent with these differences being purposeful, income overstatement varies widely and consistently across originators, is priced by originators, predictable ex-ante, and accompanied by clear inflation of past financials. Income overstatement is highly predictive of pre- and COVID-period loan distress, even after controlling for loan characteristics and geographic and time fixed effects. We document appraisal aggressiveness and abnormally low capitalization rates that are also related to current distress. Overall, aggressive CMBS underwriting practices appear to be revealing themselves in the current crisis.