Rising FHA Delinquencies: Why None of the Usual Suspects Are to Blame
From mid-2024 to February 2025, Federal Housing Administration (FHA) mortgage (insured loans intended to make homeownership more accessible with low down payments) serious delinquencies (90+ days past due) jumped from 3.7% to 4.8% bringing the overall numbers on par with 2017. Meanwhile, GSE (Government Sponsored Enterprises, Freddie Mac and Fannie Mae) delinquencies from mid-2024 toQ1 2025 were basically flat. Housing experts scrambled to find explanations.
In my work for Urban Institute, I conducted a study with Laurie Goodman, Ted Tozer, and Jung Hyun Choi to track down the source of this increase in the FHA delinquency rate, which resulted in an issue brief we published, "Why Has the Number of FHA Delinquencies Increased?” This article is based on that study. When we first approached this issue, the most likely culprit seemed obvious as one or more of the items on the list of usual suspects: relaxed lending standards (including higher debt-to-income [DTI] ratios and competition from government-sponsored enterprises [GSEs]), slowing home price appreciation, or improved loss mitigation encouraging strategic defaults.
But according to our study, none of these explanations hold water. In fact, the answer turned out to be much simpler than expected.
Debunking the Conventional Wisdom
When we began our study, we looked at typical reasons why delinquency, or failure to pay, rose. One common reason is high debt-to-income ratios, or DTIs, which are a household’s total debt payments divided by their gross monthly income. When DTIs are high, people can become less able or less likely to pay their mortgage payments. Also, lending from GSEs has opened up additional options for borrowers (GSEs have increased their high loan-to-value (LTV) share within the agency mortgage space), particularly the highest quality borrowers, indicating a deterioration in the FHA book of business. That is, the GSEs skim off the best loans with the strongest credit characteristics.
To test whether the two leading theories — rising DTI ratios and increased GSE competition for better borrowers — could explain the delinquency surge, we conducted a simulation analysis. We grouped delinquencies from the 2017 book of business by FICO score (one source of credit score), LTV ratio, and DTI ratio. We then simulated what the delinquencies would look like if every year had the same delinquency rate as 2017. Although the lending composition has shifted, the effects offset, and the simulated default rate does not vary much from year to year.
Comparing the simulated results with the actual results in Table 1 shows, as expected, that delinquencies among loans originated in 2019 and 2020 were much higher than predicted because the pandemic occurred in the first year following origination. The pandemic’s impact was much diminished by the 2021 book of business, and delinquencies began to fall. But delinquencies on loans originated in 2022, which should have been less affected by the pandemic, were higher than those originated 2021, and the 2023 book of business did not show much of a decline from 2022. Comparing the actual serious delinquency rates with the 2017 simulated results, the actual rates were more than double for 2022 and 2023. The 2024 actual numbers cannot be used for the comparison, as we do not have a full year of data.
| Origination year | FHA | Fannie Mae | ||
|---|---|---|---|---|
| Actual serious delinquency rates in the first year | Predicted serious delinquency rates in the first yeara | Actual serious delinquency rates | Predicted serious delinquency rates | |
| 2017 | 1.95% | 1.95% | 0.39% | 0.39% |
| 2018 | 1.75% | 2.15% | 0.27% | 0.44% |
| 2019 | 6.45% | 2.14% | 3.00% | 0.35% |
| 2020 | 5.93% | 1.96% | 0.95% | 0.27% |
| 2021 | 4.13% | 2.05% | 0.40% | 0.31% |
| 2022 | 4.82% | 2.14% | 0.62% | 0.38% |
| 2023 | 3.99% | 1.92% | 0.38% | 0.33% |
| 2024 | 2.54% | 1.83% | 0.05% | 0.33% |
a Using 2017 first-year default rates.
Note: FHA = Federal Housing Administration.
Source: Urban Institute calculations using eMBS data and Fannie Mae loan level data.
When we applied the same simulation analysis to Fannie Mae loan level data as a test of reasonableness, we see the same effects, though the differences were not as pronounced for non-FHA loans based solely on borrower profiles.
While local markets vary on the direction of home prices, most markets across the United States have stopped growing as quickly as they did during 2021 and early 2022, when low interest rates fueled steep price increases. To test whether slowing home price appreciation could explain the rise in FHA delinquencies, we conducted a logit regression analysis using Fannie Mae loan-level data focused on high LTV lending (FHA data wasn’t available for this specific analysis). While slower home price appreciation is theoretically linked to higher default rates over time — since borrowers with substantial equity can sell rather than default while borrowers with less equity do not have that safety net and may be forced to default — we hypothesized this wouldn't significantly impact first-year delinquencies because equity accumulation takes time to occur.
We examined the relationship between various loan and market characteristics and the probability of serious delinquency within the first year, using home price appreciation measured at the three-digit zip code level as a key variable alongside borrower DTI ratios, FICO scores, and various loan characteristics with established correlations that we could benchmark against.
The numbers in Table 2 show how different factors affect mortgage defaults. When you see a negative number, it means that factor helps reduce defaults — the bigger the negative number, the stronger the protective effect. When you see a positive number, it means that factor increases default risk. The asterisks tell you how reliable each finding is: three asterisks means we're very confident in the result, while one asterisk means we’re less certain.
| Variable | 2017 | 2018 | 2019 | 2020 | 2021 | 2022 | 2023 |
|---|---|---|---|---|---|---|---|
| HPA | -3.382*** | -2.290** | 4.290*** | -4.507*** | 0.596*** | -1.627*** | 1.764* |
| (-4.140) | (-1.940) | (17.320) | (-25.000) | (3.150) | (-3.490) | 1.680) | |
| DTI ratio | 0.029*** | 0.025*** | 0.045*** | 0.048*** | 0.037*** | 0.023*** | 0.034*** |
| (9.100) | (7.020) | (38.720) | (31.040) | (17.170) | (9.130) | (7.920) | |
| FICO score | -0.015*** | -0.019*** | -0.014*** | -0.010*** | -0.015*** | -0.018*** | -0.019*** |
| (-29.490) | (-31.270) | (-66.060) | (-38.640) | (-37.990) | (-39.450) | (-25.830) | |
| Controls | Yes | Yes | Yes | Yes | Yes | Yes | Yes |
✳ ✳ ✳ p < 0.01; ✳✳ p < 0.05; ✳p < 0.1.
Notes: DTI = debt-to-income; HPA = home price appreciation. HPA is an annual estimation at the zip code level. Values in the brackets are t-statistics.
Source: Urban Institute calculations using Fannie Mae loan-level data and ICE Mortgage Technology home price data
The results confirmed our hypothesis. Looking across all years, credit scores (FICO) consistently show negative numbers with three asterisks, meaning higher credit scores reliably prevented defaults every single year. DTI show the opposite pattern — positive numbers with three asterisks across all years, meaning higher debt loads consistently increased default risk. But home price appreciation (HPA) tells a completely different story. The numbers flip between positive and negative from year to year with no clear pattern; sometimes rising home prices seemed to increase defaults, other times they seemed to reduce them, and sometimes they barely mattered at all. This indicates no clear relationship between local price trends and first-year mortgage defaults.
As we continued our study, some suggested that improved loss mitigation, or forbearance options (the ability of borrowers to temporarily pause or reduce payments during periods of financial hardship), were encouraging strategic defaults. The data, however, shows the opposite. Ginnie Mae securities — which package FHA loans and other government-backed mortgages — show that only 0.84% of these loans are currently in forbearance. A loan in forbearance counts as delinquent. Most loans in forbearance would likely have gone delinquent in any case. The question is what proportion of the loans in forbearance would have kept paying had forbearance not been an option. If we assume 25% of the loans would have paid had forbearance not been available, the delinquency rate would have increased by about 0.21% (25% of 0.84%). Recall from our simulation exercise that FHA delinquencies for 2022 and 2023 originations are more than double what would have been expected based on the 2017 experience — that is, they were over 4%, versus a 2% expectation. Based on our assumptions above, increased forbearance could not explain more than 10% of the increase.
Moreover, mortgages have become the highest priority debt for households — more likely to be paid than auto loans or credit cards — making it unlikely that borrowers would choose to default on their mortgages first because they have options not to pay.
Even if we assume that one-quarter of these FHA borrowers could actually afford their payments but chose forbearance anyway, this “strategic” behavior would explain less than 10% of the overall increase in FHA loan delinquencies we’ve observed. More importantly, research shows that mortgages have become the absolute top priority for American households — people are more likely to skip car payments or credit card bills before missing a mortgage payment. This makes it highly unlikely that FHA borrowers would strategically default on their home loans simply because temporary relief options exist.
Where Does This Leave Us?
After systematically ruling out the sophisticated explanations that housing experts initially suspected, we arrived at the most straightforward answer: borrowers are simply experiencing greater financial stress. Inflation has consistently outpaced wage growth, eroding families’ ability to build the emergency savings needed to weather unexpected financial shocks. This financial pressure extends far beyond mortgages — delinquency rates have climbed simultaneously across auto loans, credit cards, and buy-now-pay-later products, suggesting a broad-based affordability crisis rather than housing-specific issues.
Additional stressors compound the problem: an unprecedented frequency of natural disasters, insurance premiums that have skyrocketed in many markets, and property tax increases that particularly burden newer homeowners who haven’t had time to build equity cushions.
Despite these troubling trends, several factors prevent this from becoming a systemic threat to housing finance stability reminiscent of the market crash in the mid-2000s .The improved loss mitigation tools that some critics have blamed for encouraging defaults have actually strengthened the system’s resilience, reducing the transition rate from serious delinquency to foreclosure by 46%. More fundamentally, the FHA’s Mutual Mortgage Insurance Fund (which ensures FHA mortgages similarly to private mortgage insurance, or PMI, for other types of loans) maintains an 11.47% capital ratio, much higher than the statutory requirement, indicating there is no systemic financial weakness.
While many want to blame the usual suspects we examined, the true villain perpetrating this rise in FHA delinquencies is actually broader economic pressures on American households rather than fundamental problems with mortgage lending or policy. While the trend warrants continued monitoring and concern for affected families, the robust institutional safeguards now in place are acting as heroes, insulating the housing finance system from systemic risk.
