Do you know where the next mortgage defaults will occur in your city? What cities and zip codes are most at risk during this recession? In every recession, there are two predictors of default/loss, collateral and equity. What does the data tell us today? Since 2008, there have been some major changes in financing, gone are the old days of subprime now replaced by government programs. Why is this change important? How can we predict where future mortgage defaults will likely occur?
Stay tuned for future blogs where I will focus on Georgia and Colorado specifically as these are our primary markets. When breaking the data down at the state level some very interesting trends begin to emerge.
What changed in financing?
Prior to 2008, the FHA market share was around 4% of the total mortgage market, fast forward to 2019, the share of FHA loans is now around 3.5% of the market. As the subprime mortgage market imploded in 2008 and 2009, these borrowers now are financed by FHA which has led to the explosive growth (source HUD).
Why are borrowers choosing FHA?
FHA rates are between ½ to ¾% of a percent higher than conventional loans. On a 400k house this adds up to around 3k/year. Why are borrowers choosing to go FHA versus your traditional conventional mortgage? It is easier! FHA allows down payments as low as 3.5%, credit scores down to 580, and higher debt to income ratios.
FHA loans are considerably riskier for defaults
According to 2019 data from Core logic, mortgage defaults for FHA loans is over 3 times more likely than a conventional mortgage. Why? FHA loans are riskier loans due to decreased equity and decreased credit and income requirements. FHA is a subsidized “subprime” lender sponsored by the federal government and tax payers.
Why track FHA loans?
With default rates on FHA loans substantially higher than traditional mortgages it is important to see where the majority of these loans are being made. The larger the amount of FHA loans in a particular area is a flag for increased mortgage defaults. 36% of all FHA loans originated in 2001 ended up defaulting (Housing wire). If this occurred in a particular city/zip this would no doubt have an impact on surrounding properties.
What was in the data?
FHA publishes data on origination’s on both purchases and refinances into FHA products. They also break down refinances on refi from an FHA loan to another FHA loan along with Conventional to FHA.
I found it interesting as to why would someone refinance from a conventional mortgage to an FHA mortgage? As mentioned above FHA rates are considerably higher so the only explanation is that the borrower can no longer qualify for a conventional loan. This is a key indicator of financial duress as there is no other plausible reason to refinance into a higher rate product from a conforming product.
What is the new data telling us today?
To see areas more at risk for default, I focused on the amount of mortgages (dollar value) where borrowers were refinancing from a conventional mortgage to an FHA mortgage as this is a leading indicator of possible financial duress. Below are the top ten cities
| City | Refi_FHA | Refi_Conv_Curr | Grand Total | |
| 1 | LAS VEGAS | 73438286 | 11330627 | 84768913 |
| 2 | PHOENIX | 52804347 | 6918181 | 59722528 |
| 3 | CHICAGO | 48077722 | 7070988 | 55148710 |
| 4 | LOS ANGELES | 43794101 | 8461984 | 52256085 |
| 5 | AURORA | 41176118 | 5037890 | 46214008 |
| 6 | SAN ANTONIO | 40018338 | 1200926 | 41219264 |
| 7 | SAN DIEGO | 32338100 | 5853472 | 38191572 |
| 8 | SACRAMENTO | 32599827 | 5451302 | 38051129 |
| 9 | MESA | 29820424 | 3182550 | 33002974 |
| 10 | MIAMI | 27422433 | 4420549 | 31842982 |
I then went back and looked at it by zip code as city level data many times does not tell the full story
| City | State | Zip Code | Refi_FHA | Refi_Conv_Curr | Grand Total | |
| 1 | Fontana | CA | 92336 | 10962934 | 2658728 | 13621662 |
| 2 | Roseville | CA | 95747 | 6698204 | 2435690 | 9133894 |
| 3 | Beaumont | CA | 92223 | 11932119 | 2408200 | 14340319 |
| 4 | Santa Maria | CA | 93454 | 1116032 | 2315368 | 3431400 |
| 5 | Long Beach | CA | 90805 | 9621501 | 2312594 | 11934095 |
| 6 | Arlington | WA | 94509 | 8510425 | 2176214 | 10686639 |
| 7 | Bronx | NY | 90808 | 334015 | 2083854 | 2417869 |
| 8 | Cerritos | CA | 98223 | 2586546 | 1882882 | 4469428 |
| 9 | Eagle Mountain | Utah | 10469 | 3976127 | 1835569 | 5811696 |
| 10 | Duarte | CA | 90703 | 522582 | 1773598 | 2296180 |
Do you want to know how your city/zip ranks, I created pivot tables based on the FHA data; if you click on FHA conv state you can filter by state, city, zip, etc.. Here is a link to the excel sheet I used, the original data was from the FHA website and I utilized the most recent period (May 2020).
Copy of FHA_SFSnapshot_May2020
What do these locations have in common?
I did a map of the top ten. Each of the cities are suburban or “exurban” locations near major cities from LA, Seattle, Salt Lake City, etc… (the exception is the Bronx, NY, I assume this is on the list due to the huge impacts from the coronavirus shutdown). With each of these locations houses are less expensive and inventory more plentiful. Furthermore, this area has considerable supply of newer tract developments which increases inventory.
Summary
As Mark Twain famously said, history doesn’t repeat itself but rhymes. In our current crisis, I’m starting to see some very distinct rhymes from the last real estate crisis. Although it doesn’t appear that real estate will have nearly the fall that it did in 08, there will still be defaults and stress in the market. The locations highlighted by analyzing the conventional loans to FHA loan refinance are eerily similar to the last crisis that hit many suburban and exurban areas considerably harder than other areas.
Additional Reading/Resources
- https://www.hud.gov/program_offices/housing/rmra/oe/rpts/sfsnap/sfsnap
- https://www.consumerfinance.gov/data-research/mortgage-performance-trends/download-the-data/
- https://www.huduser.gov/portal/ushmc/fi_FHAShareVol.html
- https://www.hud.gov/program_offices/housing/rmra/oe/rpts/fhamktsh/fhamktqtrly
- https://www.corelogic.com/blog/2019/01/mortgage-delinquency-rates-for-all-loan-types-continue-to-fall.aspx