We want to kick off this week’s commentary with a brief – but hopefully not cumbersome – historical overview of the Federal National Mortgage Association (riveting topic for a Thursday evening, we know). We’re going somewhere with this., so bear with us.
One of the most well known government-sponsored enterprises (GSEs) is the Federal National Mortgage Association, better known as Fannie Mae. Fannie Mae was created in the 1930s during the Great Depression as part of The New Deal. It’s purpose is to provide liquidity to the United States housing credit market, through the 1) purchase, 2) securitization, and 3) reselling of mortgage bonds on the secondary market in the form of mortgage-backed securities. For years, Fannie Mae had an ‘explicit guarantee’ from the federal government that it would service all obligations — this changed in 1968 when Fannie Mae became a private corporation, albeit while retaining a direct line of credit to the U.S. Treasury.
In the early 2000s, the mortgage market began to gravitate away from GSE securitization towards private-label securitization. This paradigm shift concurred with the advent of exotic mortgage products and derivative securities products that added extra levels of opacity to the housing finance system. Adjustable-rate mortgages started outpacing fixed-rate mortgages in popularity, and underwriting standards began to diminish (for a number of reasons we won’t go into now). GSE’s such as Fannie Mae also began lowering their underwriting standards in order to appease private shareholders thirst for yield, and before anyone (except for a handful of astute investors) were aware of how frothy the market had become, the financial system was on the verge of collapse.
Both Fannie Mae and Freddie Mac required bail outs as a result of the 2007-2008 mortgage crisis. The government’s bailout of Fannie Mae and Freddie Mac cost the American taxpayer approximately $190 billion. The agencies have both returned to profitability, and after over a decade have repaid that bailout balance + over $100 billion additionally (read: Fannie Mae and Freddie Mac (Bloomberg, 2020)).
Risky underwriting practices, blatant non-disclosure and taxpayer-funded bailouts…sound familiar?
Okay, now to this week’s POS commentary target…
Housing Authority of the City of New Haven, Multifamily Tax-Exempt Mortgage-Backed Bonds, Fairbanks Apartments Project, Series 2021
The bonds are being issued to finance the “acquisition and rehabilitation of a low and moderate income multifamily rental housing facility,” called the Fairbanks Apartments. The Housing Authority of the City of New Haven (HANH) then ‘delivers’ that property to Fannie Mae in exchange for a mortgage pass-through certificate (the “MBS”). The loan now will Fannie Mae’s ‘implicit guarantee’ of payment and the HANH is left holding a (theoretically riskless) MBS, which is backed by a 17-year, fixed rate mortgage loan tied to Fairbanks Apartments. HANH will pay out the principal and interest of the MBS bond to its municipal bond investors.
At this point, the climate risk is off the HANH balance sheet. Of course, that doesn’t mean that the climate risk is gone, it just means that risk has been pushed into different institutions. As was the case leading up to the 2008 financial crisis, Fannie Mae is essentially guaranteeing a mortgage loan without having an adequate level of transparency into the underlying risk. The difference between 2021 and 2008 is that this time the underlying systemic risk isn’t exotic mortgage products and irresponsible underwriting and rating practices, but rather the unprecedented risks posed by climate change.
We often focus on the risks posed by coastal flooding in Florida and wildfire in California, or inland flood in North Carolina. The fact is that climate change poses a threat to the entire system, either directly or indirectly. There will be bag-holding at the end of this imminent climate crisis, and while Fannie Mae is long bag-holding exposure, they’re very short on climate disclosure. This HANH mortgage-backed issuance is one example of that imbalance.
Fairbanks Apartments is located in the neighborhood of Fair Haven, which is bounded by both the Mill River on its west side and the Quinnipiac River on its east and south sides, making it a target for inland and coastal flooding. risQ projects that by 2030, GDP impairment within the 6-minute drivetime band surrounding Fairbanks will total 68%, ranking 94th percentile across all municipal-financed housing assets nationally, and in the 99th percentile statewide. This GDP impairment is primarily driven by inland and coastal flood, with severe hurricane-induced flooding, wind and storm surge hazards projected to be much less frequent.
Unsurprisingly, HANH makes zero mention of “climate” or “flood” in its preliminary official statement. One obvious reason for this non-disclosure is the fact that the parties involved in this specific transaction are well insulated from climate risk, if we’re to only consider this specific transaction. Fairbanks’ climate risk is essentially being socialized across the entire financial system. The timely payments of principal and interest for this bond series is effectively guaranteed by Fannie Mae, which is backed by the full faith and credit of the United States federal government. But what does that say about the system as a whole? There’s material climate risk that is unaccounted for, undisclosed, and essentially uninsured, but as long as the obligations are unconditionally guaranteed by the federal government, everyone participating in each debt transaction gets paid.
Someone has to eat this climate risk eventually, and if that turns out to be Fannie Mae, then climate change and climate risk writ large represents a non-trivial shock to the housing sector. Fortunately, there are investors that are starting to bake climate risk into their investment calculus, both for munis (like you, the readers of this piece) and for mortgage-backed securities. There are even investors that have fully internalized the lessons learned from 2008, and are well prepared for an imminent shock to the housing sector. Dave Burt of DeltaTerra Capital, one of risQ’s first clients, has been working in recent years to get ahead of these systemic risks (read: Climate change will break the housing market, says David Burt, who predicted the 2008 financial crisis (MarketWatch, 2019), Citing climate risk, investors bet against mortgage market (Reuters, 2019), Former subprime player claims he can now short the mortgage market for climate and Covid risks (CNBC, 2020)).
There is a natural interdependence between the municipal and housing sectors. Municipal credit depends heavily on the value of its property tax base. risQ has already built the geospatial risk modeling platform for the municipal sector, and with the help of Level 11 Analytics, we’re now able to extend that platform to modeling climate risk for securitized mortgages at the CUSIP-level (read: risQ and Level 11 Analytics Launch Climate and ESG Analytics for Mortgage-Backed Securities (PR Newswire, March 25th, 2021)).
We beat up on municipal issuers for failing to disclose their climate risk. The truth is, they just don’t know any better. Even the issuers that do disclose climate risks aren’t very good at it. We need change at a systemic level now in order to mitigate the fallout from a potential climate crisis later on. Both the municipal and housing sectors and their market participants collectively play an enormous role in that crisis mitigation effort. risQ is just here to enable those stakeholders with information.
Once this bond series issues, we’ll all indirectly own a piece of Fairbanks Apartments, almost in the same way we all own coastal flood risk that is being poorly insured by FEMA’s National Flood Insurance Program (as hammered in this week’s Thoughts from the Top). Hopefully at the end of the day, we won’t all be holding the bag.
Town of Leonville (Water & Sewer), LA (risQ Score 3.2; Notable Peril Scores: 4.1 (Flood))
A short section on page 22 regarding Hurricanes, Flooding and Sea Level Rise, with some acknowledgement that this may impact the assets of the System and may also adversely impact the desirability of the area for residents. That’s all good and fine, but…
Leonville is far inland in Louisiana, so mention of sea level rise and storm surge from hurricane is nonsensical and exemplifies the “copy/paste” lack of any true consideration of climate risk to the issuer. This is a common problem we’ve discussed before with respect to Louisiana issuers who think that any language will do, and inattentive readers who don’t realize how ridiculous the statements actuallt are. Remembering that this is a Water and Sewer system – so excess water from extreme precipitation might be something they want to actually prepare for – there is no mention of actually investing in, or doing anything about, the impending issues that climate change will cause. Keep in mind, a 100 year inland flood has a 29% property loss for Leonville. Flooding due to precipitation from hurricanes has an average property loss of 21%. Today, that’s a 1 in 65 year event, at the very least. Both these events are going to become a bigger problem for Leonville in the 2035 timeframe of the longest maturity debt in this series of bonds, but no mention of investment to mitigate this problem or a plan to do so.
Chula Vista CFD No. 16-1 (Millenia) (risQ Score 4.0; Notable Peril Scores: 4.7 (Wildfire))
A discussion of Natural Disasters on page 60, the second paragraph of which focuses on wildfires and not without reason. There is guidance that the district is not within a designated very high fire hazard area, but this is misleading. Those resources have less geospatial nuance and specificity than is required for a proper wildfire risk assessment. This CFD is very much in a high risk area if you focus on the CFD specifically. There is mention that the CFD is adjacent to uncontrolled brush areas and that they may cause a risk. Never a truer word was spoken, although this could have been conveyed in a far less understated way given the actual risk this – and the broader implications of the district’s location – actually bring with them. Ironically, one of the facilities planned for this district, according to the OS, is a fire station. Probably a good investment as a Wildfire risQ Score of 4.7 is no joke.
Coconino County, AZ (risQ Score 2.4; Notable Peril Scores: 3.9 (Wildfire))
Firstly, no specific mention of wildfire risk and a single generic mention of “natural disaster” buried in Appendix D, so not much to speak of there. Goes without saying, on that basis, that there’s no mention of climate change. For the record, Coconino County is 96th percentile for wildfire risk nationally and had four different fires within its boundary in 2020: Pine Hollow, Thumb and Monument fires, and the biggest, the Mangum fire, which burned a total of 71,450 acres. Wildfires cost money to fight and there is ample research on what wildfires do to municipal finance in general.