A Journey to More Equitable Mortgage Standards

A key part of Fannie Mae's mission is setting standards for single-family mortgages. We do this to ensure mortgages are safe and sustainable both for homeowners and for Fannie Mae. In addition, these standards are a large reason behind the mortgage finance market's ability to provide a steady, stable source of funding for borrowers.

At the heart of these standards is evaluating the many risks associated with both borrowing and lending a large amount of money for a long period of time. Remember that most mortgages have a 30-year term, and for many people it is the largest financial transaction they will ever undertake.

The standards we set are designed in part to maximize the chances that a home purchase is sustainable over time, helping homeowners attain all the benefits that homeownership offers – which in turn reduces financial risk to Fannie Mae.

Economic cycles and housing market conditions are among the many factors that lead lenders and Fannie Mae to either "ease" or "tighten" credit for risk management reasons and ensure long-term homeowner success. For instance, we adjusted our risk assessment approach earlier in 2023 to account for uncertainty in the U.S. economic environment and the potential for a mild recession.

But beyond this standard risk management practice, something new is at work. New, technology-driven capabilities are allowing us to challenge our past approaches to make a better, more inclusive risk assessment without compromising a mortgage's long-term soundness.

The primary tool for ensuring credit standards are met is Desktop Underwriter®, or DU®, the leading automated mortgage underwriting system for more than 20 years. DU automates Fannie Mae's Selling Guide for lenders, allowing them to consider a wide range of information about a mortgage applicant and a property in assessing risk and ensuring the loan is eligible to be sold to Fannie Mae.

One of the great challenges (and opportunities) facing the U.S. mortgage system is finding new ways to assess mortgage credit risk safely and soundly, while ensuring that as many people as possible have access to a mortgage.

We have embraced that challenge at Fannie Mae. We continually look for ways to reduce barriers to housing access by improving DU's ability to assess risk and evaluate creditworthiness. In recent years, Fannie Mae introduced a number of innovations and changes to our mortgage standards that are designed to serve the evolving needs of consumers with increasingly diverse needs and aspirations.

Thanks in large part to advances in our ability to access, use, and analyze data, empowered by massive increases in computing power and an occasional dose of machine learning, Fannie Mae has been able to reimagine borrower eligibility in important ways that can increase access to credit in a manner that we believe is safe and sound both for Fannie Mae and the borrowers it ultimately serves.

The most recent examples of this were outlined in a recent White Paper by my colleagues Walter Scott, Stephen Schwartz, and Angela Threadgill. They include changes to the maximum loan-to-value (LTV) ratio on loans for single-family properties with 2-4-unit residences, and the removal of "number of borrowers" as a risk factor in DU.

I encourage you to read their thorough analysis of what made these changes possible and why they believe the changes can make mortgage credit more accessible to more people without jeopardizing long-term homeownership success. But here are a few takeaways worth highlighting.

2- to 4-unit residential properties
Single-family homes that are divided into apartments (from two to four units) are one of the most stable and affordable sources of rental housing in many U.S. cities. Our analysis of 2021 housing data shows that one-out-of-six rental units is in a 2-4-unit building, and the median rent in these units is $900 a month, making them among the most affordable rental housing available. By increasing the allowable LTV ratios for loans on these properties, we are reducing the up-front costs (by way of smaller down payments) for their purchasers and owners. By reducing down payment burden, Fannie Mae is not only helping to create and preserve affordable housing, we are also empowering more borrowers to own income-producing properties. As many people in cities such as Baltimore, Chicago, or Boston can attest, living in and managing small, multi-unit properties has been a traditional wealth-building strategy for many minority and immigrant households. Black and Latino/Hispanic people comprise more than 30 percent of the those who own 2-4-unit buildings, and comprise more than 44 percent of the people who rent them.

Number of borrowers on a mortgage loan
In the last 20 years, Desktop Underwriter looked at the number of borrowers on a loan as part of its evaluation of risk. Put simply, having one borrower on a loan was an important predictor of higher risk. Going forward, this risk factor will not be considered in Desktop Underwriter. This is possible because we have advanced DU's ability to assess the risk of default on any loan by evaluating a wider range of factors and data, including a broader view of an applicant's credit profile, loan application information, and property value attributes. As a result, considering the number of borrowers on a loan no longer provides meaningful value in our risk evaluation. This matters because in those same last 20 years, the number of homeowners and potential homeowners who are single heads of households, including single female heads of households, has grown substantially. With this change to our risk assessment, the fact that these consumers are borrowing on their own will not be a detrimental factor in loans underwritten through DU.

Both of these changes could have broad implications for underserved communities, including (but not limited to) Black and Latino families. In short, we expect both changes will have a positive impact on populations of first-time home buyers with less access to generational wealth, which is a key driver of homeownership disparities among different racial and ethnic groups in the United States.

Improving access to credit safely
These improvements are the latest milestones in a technology- and data-driven journey that has allowed us to make DU refinements that balance Fannie Mae's need to manage risk with its mission to make mortgage credit more accessible. A few recent examples:

• In 2021, Desktop Underwriter was the first large-scale automated underwriting system to consider on-time rental payments as part of its credit risk assessment.
• In 2021, for loans with more than one borrower, we determined that using the average median credit score of both borrowers was a better indicator of credit risk than just using the lowest median credit score, which we had used prior to this change.
• In 2022, we developed the capability in Desktop Underwriter to assess a borrower's eligibility, even if that borrower had no established credit score, through access to bank account data to determine a borrower's ability to make a monthly mortgage payment.

These innovations were made possible by our never-ending quest to harness technology and data to assess the creditworthiness of individual borrowers more completely and more equitably. The mortgage system can address uneven credit opportunities that have edged underserved populations out of homeownership for many decades, and do it without sacrificing the fundamental soundness of a mortgage or the safety of that same system.