The fragility of Florida

How dodgy insurance is being used to systematically dump risk on the books of the entire country

Insurance makes the mortgage market work. You can't get a mortgage from a bank (these days, likely a nonbank) without insurance, because the lender needs some kind of guarantee that you'll pay them back. Your pledge to the bank that it can have your house if you don't repay the loan has to be accompanied by property insurance that will preserve the collateral value of the property that secures the mortgage.

But what if the insurer is unreliable and isn't able to pay claims following a major storm? Then everything happens at once: properties are damaged at the same time that the insurer becomes insolvent, borrowers default on their loans, and lenders have losses on their books. Multiply those losses, and you get tremendous economic pain to entire financial systems as credit freezes and markets plunge. 

The 2008 financial crisis was triggered by the insolvency of nonbanks heavily involved in dodgy mortgages. You probably remember that credit rating agencies were at the heart of the problem: they gave high ratings to mortgage-backed securities that were bought by nonbanks (and investors around the globe). Those MBSs became worthless and the whole house of cards tumbled.

Today, a similar story appears to be playing out in Florida, where 10 percent of the nation's homeowners live: a dubious credit rating agency is giving high "financial stability" ratings to flimsy insurers. A whole series of players is going along with those ratings, using them to check boxes and issue mortgages. And then those actors are swiftly shifting the resulting mass of risk to Fannie Mae and Freddie Mac. The GSEs are accepting these inflated financial stability rankings and buying these smelly mortgages. The strategy is, must be, to have the feds bail out the situation when storms strike and the casino suddenly closes its doors. 

We know that this summer will be dangerous for Florida and Florida real estate. Here's a use of scare-quotes from this week's New York Times that is genuinely scary:

The consequences of this credit-rating box-checking will be felt far beyond Florida. In the subprime mortgage crisis, the problem was that the borrowers weren't credit-worthy. Now the problem is that the properties aren't credit-worthy, but it is in no one's interest to blow the whistle.

This alarming story is spelled out in an elegant recent paper by Pari Sastry, Ishita Sen, and Ana-Maria Tenekedjieva called "When Insurers Exit: Climate Losses, Fragile Insurers, and Mortgage Markets." The authors look carefully at public data up through 2018 and conclude that although large, well-capitalized insurers have steadily pulled out, smaller, riskier insurers kept issuing policies in risky Florida areas.

These shakier insurers were incentivized in part by payments Florida made to them to take on policies the state's insurer-of-last-resort, Citizens Property insurance Corporation, issued. And most of them were the beneficiaries of grade inflation by a rapidly-growing rating agency called Demotech. Both the mortgage originators and the GSEs accept Demotech's financial stability rankings for these insurers, and the GSEs are now stuck with a steaming pile of risk.

I think we can assume that this situation has only gotten worse since 2018. But the authors had to stop there because the insurers convinced a court that their data was all a trade secret and shouldn't be released to the public. This is a common move; picture the industry as ATMs with lawyers layered on top. Avoiding disclosure is easy.

Back to the narrative. You may have heard of the big three credit rating agencies: AM Best, Moody's, and S&P. Demotech is probably new to you. It entered the market in the 1990s and specializes in evaluating the financial stability of regional and specialty insurers, mostly in Florida. It now issues most of the insurer credit ratings in Florida. 

Following Hurricane Andrew in 1992, and the huge numbers of insurer exits and insolvencies that storm caused, Florida set up its public insurance option, Citizens. Floridians swarmed toward Citizens, which ballooned. Florida then began paying insurers to take policies off Citizens' books. 

Meanwhile, however, the big insurers like State Farm had quietly been leaving high climate-risk counties in Florida. That didn't mean that the wheel stopped spinning. New, substantially lower-quality insurers with financial stability ratings from Demotech rapidly gained market share in Florida, growing in large part by absorbing those Citizens' policies. That was superficially great news for Florida, because if Citizens goes under Florida bears all the liability—meaning the state’s residents will have to bail it out.

According to the study's authors, the Demotech-rated insurers are lower quality in many ways: they have far less capital, focus on lower-value homes, aren't geographically- or product-diversified, and tend to enter rehabilitation (restructuring) at higher rates. But Demotech gives them grades that are higher than the ratings traditional credit ratings agencies would give for insurers comparable characteristics. Grade inflation is rampant at Demotech: "The vast majority of these insurers would likely be rated 'junk' if they received their ratings from a traditional rating agency rather than Demotech," say the study authors.

Alarmingly, the authors find "suggestive evidence of higher regulatory forbearance over time" when it comes to Demotech-rated insurers. State regulators aren't looking harder at these riskier insurers. In fact, they appear be getting less scrutiny than more traditional insurers. This makes sense from Florida’s point of view: insurance has to remain “available” and “affordable” for Florida’s real estate market to keep spinning.

Here’s the key to this story: The GSEs are accepting the "financial stability ratings" provided by Demotech as part of their own box-checking enterprise. Lenders in Florida counties with larger shares of Demotech insurers are selling on mortgages to the GSEs at higher rates. (“Securitization shares in a county strongly covary with the market share of Demotech insurers," in the authors' words.) This makes complete strategic sense. If you know what you've got as collateral is risky, you'll do your best to shift your risk elsewhere.

Again, this is great for Florida: it can keep collecting property tax, keep people believing in the dream, keep encouraging them to get and keep mortgages, while shoving the risks of this behavior off its own books by steadily "depopulating" Citizens' policy holdings—by shifting them to these private sector, higher-risk insurers (hate government, love the private sector). All the while, the state is actually, ultimately, shifting its risk to the federal government.

The GSEs are exposed, apparently willingly, to these ever-increasing risks. This means that some portion of the $12 trillion residential mortgage-backed-securities market in the US (second in size only to the market for US Treasuries) is also exposed to this risk.

Why is this happening? What makes the flywheel stop? Inevitably, a storm, or a series of storms, will cause a series of painful, disruptive reactions that will harm vulnerable homeowners first and everyone else later. The fat tails of low but extreme risks are getting more likely all the time. Surely it would be better to put all the risk on the table, change the GSE guidelines, and plan carefully for a 10-year transition to a safer financial story. But that seems as dreamlike as .... Florida real estate.