Another one of those wonky Federal Reserve papers was released last week, which made the claim that government-backed mortgage insurance promoted a faster recovery post Great Recession.
I had to read it about four times to understand it, that’s a government working paper for you, but the takeaway is basically that areas of the country with lots of FHA loans fared better than areas with other types of loans, such as those backed by the GSEs, or worse, portfolio loans and private-label securities.
The researchers looked at things like unemployment rates, home sales, and home prices prior to and after the financial crisis and found that counties with a high concentration of FHA loans performed best.
For example, by 2008, unemployment rates increased 26% in counties that had a low FHA share in 2005, compared to just a four percent increase in counties where the FHA share was high.
A year later in 2009, the unemployment rate had risen by 106% and 58%, respectively, for those two groups.
Once the crisis ended, around the end of 2012, the unemployment rate remained 79% higher in the low FHA-share counties but only 49% higher in high FHA-share counties.
At the end of 2014, the unemployment rates were still 30% higher in the low-FHA share counties compared to 2005 levels, and just 19% higher in the high-FHA counties.
Areas with loans backed by Fannie Mae and Freddie Mac (GSEs) also did fairly well, with smaller declines in home sales and home prices during and after the housing crisis when compared to portfolio mortgages and private-label securities.
Things Can Get Ugly When Banks Do Whatever They Want
The researchers discovered that areas with lots of portfolio mortgages (those held by banks that set the underwriting criteria) and private-label securities (those sold on the secondary market to private investors) performed the worst.
Counties with lots of portfolio mortgages and PLS saw significant increases in unemployment, along with higher delinquency rates and foreclosure completions.
As of the end of 2008, unemployment rates had risen just seven percent in counties with a low PLS share in 2005, compared to a 25% increase in counties with high PLS shares.
A year later, the unemployment rate climbed 69% and 106%, respectively. At the end of 2012, it remained 45% higher in areas less dependent on PLS and 76% higher in areas heavily reliant on PLS.
If you’re wondering why the counties with heavy government mortgage participation did better, consider the fact that government underwriting guidelines are set at the national level and not subject to much change.
Conversely, banks that make their own mortgages with their own rules can change them at will as the market demands it. So if their max LTV was 90%, and they see a market for 100% cash out refinances, they may go for it (hint: they did go for it). The consequences can be particular nasty, though only after huge profits are taken.
This explains why the counties with a high FHA-share did better relative to the counties with loans underwritten by the likes of Countrywide and other antagonists.
It’s not to say that government mortgage lending doesn’t have its flaws…the FHA got into a lot of trouble when it allowed down payment assistance providers to run rampant. And there are still plenty of folks who believe their low down payment and low credit score combo equates to high risk.
They’ve since made it a little tougher to get a loan, though you can still put down just 3.5% on a home with a 580 credit score…that’s low.
The one flaw with this paper might be that the mortgages being written during the most recent crisis were truly awful and hopefully will never be replicated. Fat chance, right?