The damage to American consumers’ financial histories caused by the COVID-19 pandemic will harm their ability to purchase homes for years to come. To date, this challenge has been addressed by government programs and proposals meant to safeguard traditional FICO credit scores from the missed payments and high credit-card balances that are a part of the current economic downturn.These proposals ignore the rapidly expanding fintech mortgage sector. Fintech lenders supplement traditional financial metrics like credit scores with a vast array of ‘big data’ sources to judge consumer creditworthiness. Policymakers must ensure fintech lenders do not penalize consumers for recent damages to their data profiles, especially since consumer data collection has proliferated during the pandemic, often unevenly across the lines of race, age, and income level.
Government responses and media attention focused on the economic crisis’ impact on future homeownership have focused narrowly on protecting traditional FICO credit scores. The House of Representatives’ Protecting Your Credit Score Act, passed in June, seeks to increase consumer access to credit reports. The CARES Act includes a provision ensuring a person’s credit score won’t drop if they have a repayment agreement in place. Senators introduced a bill ensuring negative financial information from recent months would not be reflected on credit reports. And Vice President Biden’s housing platform proposes the creation of a public credit reporting agency that has been lauded for its potential to protect vulnerable consumers’ credit scores during times of economic turmoil.
Yet, no measures have been forwarded to protect consumers seeking loans at fintech – short for financial technology – mortgage lenders, which today include some of the nation’s largest mortgage lenders such as Guaranteed Rate, Better Mortgage, and Quicken Loans’ Rocket Mortgage. This new class of technology-driven financial institutions is characterized by online applications and near-instant underwriting decisions that deemphasize the importance of credit scores. Research suggests lower-income and non-white borrowers are more likely to originate loans from fintech lenders relative to brick-and-mortar lenders, meaning the fintech market may play a role in serving historically underserved groups.
Fintech lenders supplement information from credit histories and other traditional financial metrics by using machine learning algorithms to sift through immense amounts of consumer data in an attempt to link a loan applicant’s profile to various loan default risk factors. These links are not always intuitive; social media friend network characteristics, educational attainment, merchant transactions, utility usage, travel behavior, and internet search history can all be used as predictors of loan default despite being seemingly unrelated to creditworthiness.
For many, the pandemic has changed elements of these ‘digital footprints’ in ways that are likely to be associated with higher credit risk by fintech algorithms. Web searches for unemployment benefits or stimulus funds, grocery shopping at lower-end markets, and exhibiting anxiety on social media would all likely be online behaviors that a fintech algorithm would associate with mortgage default risk. Given the virus’s well-documented uneven spread by race, age, and occupation, these effects will disproportionately disadvantage already vulnerable groups.
Consumer data profiles have not only been altered by the pandemic; they have also been expanded. Individuals volunteer a wealth of health information in exchange for COVID-19 tests and in contact-tracing processes, and new proposals for the collection of physical health records related to the disease allow these data to be shared without an individual’s consent. These expanded data sources suffer from privacy and security concerns. Reports have noted the vulnerability of databases cataloging recipients of recent government stimulus payments, concerns over the anonymity of individuals’ geospatial locations tracked by contract tracing apps, and insufficient protection of user data on video conferencing platforms. These new and expanded sources of data and others like them will be packaged and sold to those who value them, and their exploitation by fintech lenders is not difficult to imagine.
Policymakers must not turn a blind eye to the pandemic’s impact on the data used in fintech lending. Like credit scores used by traditional lenders, consumers’ online data profiles are sure to be damaged, and borrowers should not be denied access to loans or given subprime credit by fintech lenders due to economic factors well beyond their control. Lawmakers and regulators such as the Consumer Financial Protection Bureau should draft and implement legislation preventing fintech lenders from incorporating consumer data reflecting the pandemic’s economic impact into their future lending decisions. This would require fintech lenders to reveal the sources of their underwriting data and be held accountable for unfair lending outcomes, a long-overdue step that would aid in Fair Housing monitoring more generally. Recent proposals go even further, calling on regulators to encourage fintech lenders to produce outcomes that improve consumers’ financial health.
To be sure, requiring that fintech lenders reveal their consumer data sources compromises the proprietary nature of their algorithmic underwriting systems. However, data inputs comprise only a small piece of the loan underwriting process for fintech lenders, and their trade secrets can be preserved while regulators ensure consumers are protected.
In this way, the economic divisions revealed by the pandemic could present an opportunity for policymakers and regulators to promote equitable housing outcomes for protected classes of borrowers that long outlast COVID-19, turning a devastating crisis into a catalyst for fairness and ensuring the individuals and communities most impacted by the shutdown are not shut out of the wealth-building opportunities of homeownership.
Addressing the fintech lending market would also represent a much-needed recognition of the importance of consumers’ online behavior in shaping their financial outcomes. The pandemic will leave its mark, not only on society’s physical and emotional well-being, but also on its digital health. Without a swift intervention, fintech mortgage lending outcomes will reflect this damage by deeming some borrowers financial risks, simply because their struggles against a health risk were tracked online.
This article is part of The Bottom Line, a series exploring scalable solutions for problems related to affordability, inclusive economic growth and access to capital. Click here to subscribe to our Bottom Line newsletter. The Bottom Line is made possible with support from Citi.
Tyler Haupert is a doctoral candidate in Urban Planning at Columbia University's Graduate School of Architecture, Planning and Preservation. Follow him on Twitter at @tyler_haupert.
Gayatri Kawlra is a doctoral candidate in Urban Planning at Columbia University's Graduate School of Architecture, Planning and Preservation. Follow her on Twitter at @gaya__tree.