This post first appeared on the Climate Risk Disclosure Lab’s website.
The physical effects of climate change are wreaking havoc across the United States as extreme weather events are increasing in severity and frequency. The 2020 summer wildfires in the western U.S. broke “almost every record there is to break,” and there has been a notable increase in storm surges, sea level rise (“SLR”), heavy precipitation events, and chronic floods. The continental U.S. was hit by twelve named storms in 2020—the most ever in a single season. Importantly, these natural disasters are increasingly threatening areas beyond officially recognized high-risk zones. As noted in a 2019 letter to Federal Reserve Bank of Chicago, “climate scientists have expressed concerns that this recent wave of devastating storms is not an anomaly, but rather part of a long-term trend.” These extreme weather events, boosted by climate change, have made clear the devastating potential of climate-related disasters. But, as indicated by BlackRock, the financial implications of these events have been “notoriously hard for investors to grasp.” Investors in mortgaged-backed securities (“MBS”) in particular, are increasingly concerned with the vacuum of information on climate risks.
The mispricing of SLR and wildfires poses a substantial danger to the U.S. mortgage market. Large disasters can cause abrupt devaluations of property and mortgage values, representing a substantial risk for lenders who retain mortgages on their books. Moreover, “the correlation of such natural disaster risk across loans in a mortgage pool” poses a substantial source of aggregate risk for holders of MBS. The mispricing of disaster risks and poor disclosure standards also means that home buyers risk overpaying for properties that could be struck by an unexpected extreme weather event over the course of a 30-year mortgage. In January 2021, the Federal Housing Finance Agency acknowledged that climate change and natural disasters threaten to disrupt to the national housing finance market and began seeking information to better identify and assess potential climate-related impacts on Fannie Mae, Freddie Mac, and Federal Home Loan Banks.
This article starts by discussing the recent surge in climate-related disasters and the effects that these events have on regional real estate markets. It first assesses the increase in large wildfires in the western United States, and then turns to the consequences of SLR, flooding, and storm surges. It then evaluates the potential downstream effects that these extreme weather events, coupled with poor disclosure standards, could have on the U.S. mortgage market. The article finds: (1) extreme weather events can cause rapid increases in delinquency and forbearance rates, as well as prepayment rates, (2) these weather events can also increase the number of risky mortgages being securitized and transferred to the secondary market, and (3) the increased threat of extreme weather disasters can and should undermine property values.
Human-caused climate change is contributing to a surge in wildfires in the western United States. A warmer climate dries out vegetation and deepens summer draughts, fueling longer and more extreme fire seasons. As seven of the 10 most destructive fires in California’s history have occurred in the last five years, it is clear that climate change has made forests more susceptible to bigger, more destructive wildfires. The 2018 National Climate Assessment noted that “in a world without climate change,” only half as much forest in the western U.S. would have burned between 1984 and 2015. According to California’s fire-fighting agency, CalFire, about a quarter of the state’s homes are currently at “high risk” from wildfires. One study found that, overall, nearly 60 million homes were within less than a mile of a wildfire between 1992 and 2015. Importantly, in recent years, wildfires have “leapt from wildlands and into dense suburban neighborhoods where residents never considered them a threat.”
As indicated in the Federal Reserve’s most recent Financial Stability Report, the price of real-estate-linked assets in these regions would accurately reflect wildfire risks if market participants had “perfect information.” But an analysis by NPR found that “most wildfire-prone states have no requirements for disclosing fire risk to someone who buys or rents a home.” According to the study, only California and Oregon require some level of wildfire risk disclosure—but even in those states, “disclosures amount to only a few lines of text, buried in the hundreds of pages buyers typically receive when closing on a new home.” The Oregon disclosure requirement, for example, does not contain the word “wildfire.” Rather, one line in the seven-page disclosure document filed during a sale specifies whether a “property is in the ‘forestland-urban interface,’ a potential wildfire zone mapped by state officials.” A 2012 study surveyed homeowners in a fire-prone area in Colorado and found that nearly half of the respondents “did not know that their homes were located in an area at risk of wildfire.” As homeowners in western states “systematically underestimated wildfire risk,” the vast majority of them fail to undertake mitigating actions “to decrease the risk of losing their home,” such as purchasing insurance or taking fire-proofing precautions.
Enhancing wildfire disclosure requirements would help homebuyers make more informed risk-mitigating decisions. But there is also a growing need for more robust information about wildfire risks. As the United States is set to experience “increased fire risk” and longer fire seasons, many wildfire maps are outdated and inaccurate. The California Department of Forestry and Fire Protection (“CDFFP”) has not delineated at-risk fire areas since 2007. As highlighted by Bloomberg News, California has experienced 15 of the 20 most damaging fires in its history since that time. Further, wildfires before 2007 mainly affected forest and open grasslands, and only a small number of houses that stood at the edge of the wildland-urban interface. But in recent years, fires have encroached further into suburban areas, threatening homes that lie outside of official at-risk zones. When the existing maps were made in 2007, these urban areas were marked as “vegetation-free”—but in reality, they are “pocked with trees, shrubs, and backyard landscaping that can dry up and create a conduit for flames.” Moreover, NPR found that the majority of existing wildfire maps fail to incorporate “added risk from climate change as hotter temperatures fuel more extreme fires.” Though the U.S. Forest Service released new interactive maps in 2020 that show community risk nationwide, “the agency says the maps are not fine-scaled enough to use for individual properties.”
California has been working on a new model to delineate more granular data regarding severe fire hazards. According to Bloomberg, there are 2.2 million homes in the current “severe fire zone.” But, as noted by David Sapsis, head of risk mapping at CDFFP, “that number could increase dramatically with new maps.” Such a change would surely result in reduced prices for homes within the “new” severe fire zone.
Most importantly, the rise of large-scale wildfires has led to massive losses for insurance companies, forcing them to raise their premiums or abandon newly-delineated fire-prone areas altogether. Many insurers have concluded that their premiums are set too low to cover the increased losses from wildfire. As a result, fire insurance rates are “skyrocketing,” and many companies are simply abandoning customers or refusing to write new policies on homes in particularly risky areas. According to the New York Times, Graham Knaus, the executive director of the California State Association of Counties, concluded that the marketplace for insurance in high-risk areas “has largely collapsed.” The number of households turning to the state’s “high-risk insurance program,” a costly and bare-bones alternative for homeowners unable to get private coverage, “has increased by more than 50 percent between the start of 2019 and June 2020, to almost 200,000.” In December 2019, California’s insurance commissioner, Ricardo Lara, called the state’s insurance dilemma a crisis. In an attempt to stop insurance companies from retreating, Lara imposed a one-year ban preventing “insurers from dropping customers in or alongside ZIP codes struck by recent wildfires.” The statewide moratorium ended in December 2020 and cannot be renewed.
Floods & Storms
Floods are the most common natural disaster in the United States. Climate change is boosting heavy precipitation events, hurricane-force winds, changing weather patterns, storm surges, and SLR—all of which contribute to increased flood risk. According to a federal report released in 2018, because of SLR, “more and more cities are becoming increasingly exposed end evermore vulnerable to high-tide flooding,” which is expected to increase exponentially in “frequency, depth and extent” along U.S. coastlines. Freddie Mac’s Economic & Housing Research group noted in 2016 that the increase in SLR is likely “twice the level of previous estimates,” highlighting that coastal properties are becoming increasingly exposed to unexpected losses. The agency concluded that the impacts of climate change on coastal regions are inevitable. Moreover, Freddie Mac’s then-chief economist Sean Becketti concluded in 2016 that losses from flooding are “likely to be greater in total than those experienced in the housing crisis and the Great Recession,” noting that unlike the financial crisis, “homeowners will have no expectation that the values of their homes will ever recover.” Most concerning, flooding is increasingly threatening areas outside of known at-risk areas.
As of 2018, “42% of the U.S. population resides in a coastal shoreline county,” and, despite the intensifying of climate-related disasters, that number continues to rise. By 2045, it is estimated that 300,000 existing coastal homes will be at risk of flooding regularly. And according to the U.S. Climate Resilience Toolkit, nearly 9 million people live in areas “that have at least a one percent chance of flooding in any one year.” But flood risk is also increasing in non-coastal cities, where the threat of climate-related disasters “might be less obvious to residents.” Over 75,000 properties in Chicago are “highly vulnerable” to flooding over the next 30 years, according to a 2020 analysis by the First Street Foundation. Notably, that number dwarfs the official estimate from the U.S. Federal Emergency Management Agency (“FEMA”), which has identified only 1,500 at-risk Chicago properties.
FEMA, the agency responsible for determining “the objective risk of flooding,” is tasked with ascertaining whether a property is located in an official floodplain. Houses located in a 100-year floodplain, known as a “special flood hazard area” (“SFHA”), have a 1% chance of being flooded each year. Over the course of a 30-year mortgage, “this translates into a 26% chance of flooding.” When a homebuyer applies for a mortgage, the lender is required to consult FEMA-maintained “Flood Insurance Rate Maps.” If the home is located in a floodplain and if the loan is from a federally backed or regulated lender, the borrower must obtain a flood insurance policy and maintain it throughout the life of the mortgage. Flood insurance is typically provided to mortgage owners in SFHAs through the National Flood Insurance Program (“NFIP”), which is managed by FEMA. But a growing body of research shows that FEMA’s floodmaps are extremely unreliable predictors of future flood risks.
Recent data released by the First Street Foundation shows that nearly 15 million residential and commercial properties are currently at “substantial risk” for flooding. After adjusting for future environmental factors, the Foundation’s model, which was developed by over 80 of the “world’s leading hydrologists, researchers and data scientists,” identified an additional 1.2 million properties that will be at substantial risk by 2050. That brings the total number of at-risk properties to 16.2 million. By contrast, FEMA has concluded that only 8.7 million properties are located in an SHFA.
Moreover, according to FEMA estimates, only 13 million people live within a 100-year floodplain. However, a 2018 study published in Environmental Research Letters shows that nearly 41 million Americans—over three times FEMA’s estimate—“live within the 1% annual exceedance probability floodplain.” The study, which used high-resolution flood models that incorporated updated river, elevation, and rainfall data, as well as revised population density maps, concluded that the total value of assets at risk from flood damage in the floodplain is $1.2 trillion. The study concluded that communities in South Dakota, Nebraska, and New Mexico could see a five-fold increase in flood exposure by the end of the century—it may triple or quadruple in Florida and Texas. The study did not take into account the effects of climate change, like SLR and more extreme precipitation events, meaning that absent substantial mitigation measures, even more Americans may be at risk of flooding.
The Department of Homeland Security’s Office of Inspector General released a report in 2017 showing that nearly 60% of FEMA flood maps are outdated. According to the National Resources Defense Council (“NRDC”), despite FEMA’s requirement to assess its maps every five years, many of the agency’s maps have not been updated in decades. Moreover, FEMA’s maps typically fail to incorporate the effects of climate change into their assessments. FEMA does not take into account how erosion, SLR, or changing precipitation patterns might alter flood risk—instead, they rely on “historical data to determine future flood hazard projections.” The NRDC concludes that this practice leads “officials to designate areas as being ‘safe’ for development today even when they are at risk of serious floods tomorrow.” As concluded by Carolyn Kousky, the executive director of the University of Pennsylvania’s Wharton Risk Management and Decision Processes Center, FEMA’s flood maps are only a “snapshot of today’s flood hazards in a community.” As such, they are poor indicators of future flood risk and are insufficient to effectively guide development and land-use decisions.
In 2017, FEMA recognized that properties outside of its designated high-risk zones are also susceptible to flooding. According to the agency, one-third of the federal Disaster Assistance Funds distributed following flooding events have gone to property owners who live outside of its designated flood zones. And BlackRock estimated that 80% of the commercial properties in Miami and Houston that were damaged during Hurricanes Irma and Harvey lay outside of FEMA flood maps. Further, when Hurricane Sandy struck in 2012, “many of the maps for areas that were flooded had not been updated in nearly 30 years,” and the resulting floods “covered an area 65% larger” than the FEMA-designated flood-vulnerable area.
The fact that such a large number of at-risk properties lay outside official FEMA zones means that many property owners are likely underinsured against such extreme weather events. In fact, the number of National Flood Insurance Program policies in force has been steadily declining since 2009. As flood insurance is only required for mortgage owners whose properties are located within FEMA-designated flood zones, not only do non-designations “give owners a false sense of security that might influence” their decision to get flood insurance, they also remove “the legal obligation” to do so. Critics of FEMA maps also highlight that flood-maps present flooding as a binary risk—a property is either inside the SFHA and at risk, or outside of the SFHA and safe. Studies show that many homeowners are “myopic or just fail to pay attention to low-probability risks and not seek out the necessary information.” When Hurricane Harvey hit in 2017, for example, less than 20% of homeowners in the eight most affected counties had flood insurance.
The risks associated with the vacuum of reliable flooding information are exacerbated by poor disclosure standards which vary by state. Many states do not require information about flood risk or past flooding events to be disclosed to potential homebuyers. According to the NRDC, twenty-one states have “no statutory or regulatory requirements for a seller to disclose a property’s flood risks or past flood damages.” And many of those states, as noted by NPR, “are among the most vulnerable to rising seas and climate-driven extreme rain, including Florida, Virginia and Massachusetts.” An article published in the Florida Bar Journal in October 2020 explains that Florida’s current common law effectively “expects buyers to perform inspections beyond what is customary, to research and analyze flood maps and records, and to knock on the doors of their prospective neighbors in order to evaluate their flood risk.” And in Virginia, sellers are not required to make any representations with respect to whether the property is located in an official FEMA floodplain. There, prospective buyers are “advised to exercise whatever due diligence they deem necessary” to determine whether their property is at risk of flooding and to determine whether flood insurance is required. Moreover, in 27 of the 29 states that dorequire disclosure of flood risks, NPR found that buyers do not “receive information about whether a house is prone to flooding” until after they have made an offer on the house.
Nor do real estate websites like Zillow and Trulia, which have become indispensable for home buyers, include information on flood risks in their listings. This past August, Realtor.com became “the first site to disclose information about a home’s flood risk and how climate change could increase that risk in the coming decades.” Home listings on the site now include both FEMA designated flood zones and privately assembled flood information. If other real-estate websites stood up to home sellers—who are concerned that disclosing flood risk may make their home less desirable—and disclosed flood risk information, millions of buyers would be protected from overpaying for homes.
Potential Implications for the Mortgage Market
In their most recent semiannual Financial Stability Report, the U.S. Federal Reserve, for the first-time, recognized climate change as a threat to the U.S. financial system. The report paid particular attention to the devastating potential of climate-related risks on the real estate market. The Fed noted that residential and commercial properties will be increasingly “subject to acute hazards such as storm surges associated with rising sea levels and more intense and frequent hurricanes.” As extreme weather events become more frequent, “the expected value of exposed real estate may decrease, which may in turn pose risks to real estate loans, mortgage-backed securities, the holders of these loans and securities, and the profitability of nonfinancial firms using such properties.” The agency recognized that there is great market uncertainty regarding the “timing and severity of future climate-related flooding,” which has led to an “opacity of asset exposures.” This lack of climate-related information may cause “investors in real-estate-linked assets [to] react abruptly to new information about a region’s exposure to climate-related financial risks,” incentivizing fire sales by leveraged financial and nonfinancial firms.
The Fed’s Stability Report echoes warnings raised by the Commodity Futures Trading Commission’s Climate-Related Market Risk Subcommittee (“the Subcommittee”) in its comprehensive climate risk report released in September 2020. The Subcommittee recognized that “the value of real estate is closely linked to the value of the land it is built on,” leaving real estate prices vulnerable to the physical risks associated with climate change. Noting that “climate risk already appears to affect real estate values,” the Subcommittee warned that “these effects likely will increase as physical risks become more frequent and severe.” As most residential real estate in the United States is purchased with a mortgage, the downstream consequences of these extreme weather events could be extremely destabilizing to the U.S. mortgage market. Natural disasters can cause losses for mortgage lenders and Fannie Mae and Freddie Mac (“the GSEs”) in two principal ways: in very short periods of time, extreme weather events can lead to increased mortgage delinquency rates, or, if households have adequate insurance, an increase in prepayment rates. Further, evidence suggests that lenders in high-risk areas are not tightening credit—instead, they are likely increasing securitization and thereby transferring climate risks to the secondary market.
Default and Forbearance
According to the Climate-Related Market Risk Subcommittee, wildfires and flooding typically lead to increased residential mortgage default rates. In 2016, Freddie Mac’s Economic & Housing Research group highlighted the unique risk of mortgage delinquencies following extreme weather events: according to the agency, a significant share of mortgage holders continued to make their payments during the 2007–09 housing crisis even when their home’s value fell below the balance on their mortgage (meaning the borrower was figuratively underwater). But, in the case of extreme weather events, it is “less likely that borrowers will continue to make mortgage payments if their homes are literally underwater.” As a result, “lenders, servicers and mortgage insurers are likely to suffer large losses” after such disasters.
Mortgage data from Black Knight McDash shows that in the months following Hurricane Harvey, delinquency rates nearly doubled in areas where flooding was most severe. An analysis by CoreLogic shows that “three months after 2018’s Hurricane Florence made landfall, serious delinquency rates had doubled in major metros affected by the storm.” And a 2020 study that assessed the effects of fires on the U.S. mortgage market between 2000 and 2018 concluded that “mortgage default and foreclosure increase in the event of a wildfire.”
Importantly, according to Politico, Fannie Mae found that mortgage defaults are more likely to occur when flood damage affects homes outside of the FEMA flood-zones, where homeowners are not required to have flood insurance. As mentioned above, the number of flood insurance policies in force has been in steady decline in recent years, and insurers in western states have been aggressively increasing premiums and decreasing coverage in fire-prone areas. Extreme weather disasters that damage uninsured properties can significantly increase default risk. Affected homeowners with insufficient insurance may find themselves unable able to cover losses. Absent a disaster, when property values are increasing, default risk is generally low, as mortgagors unable to make mortgage payments can opt to sell the property and pay off the mortgage. But following extreme weather events, property values in the affected areas often decline drastically, making it unlikely that the proceeds from a sale would be sufficient to cover outstanding mortgage balances. Moreover, following abrupt decreases in value, foreclosure may be insufficient to make the lender whole in the case of default. Disaster risk thus affects lenders’ mortgage payoffs “over and above the other drivers of default,” such as job loss or divorce.
Extreme weather events can also lead to an increase in forbearance rates for government-backed mortgages, which can cause liquidity strains for mortgage servicers. Homeowners with a federally-backed mortgage whose homes are affected by natural disasters are often entitled to forbearance and other forms of payment relief. Homeowners affected by a disaster are typically eligible to reduce or suspend their mortgage payments for up to a full year, during which time they will not incur late fees and foreclosures are suspended. Fannie Mae, Freddie Mac, and the Federal Housing Administration offered loan forbearance for most homeowners affected by Hurricane Harvey for 12 months.
The risks associated with mortgage defaults and forbearance are exacerbated by the COVID-19 pandemic. Though forbearance is a normal “tool in the toolkit” for the GSEs and mortgage lenders, the pandemic has led to an unprecedented rise in mortgage forbearance rates. As of September 2020, nearly 7% of residential mortgages were in forbearance, representing approximately 3.5 million homeowners. By contrast, less than 4% of residential mortgage balances were delinquent at the end of 2019. An additional surge of homeowners unable to pay their mortgage due to an extreme weather event could result in significant losses for mortgage servicers, who are contractually responsible for advancing loan payments to secondary market investors, regardless of whether the borrower is making payments on time.
Natural disasters can also lead to an increase in prepayment rates. Prepayment risk is a unique concern for mortgage lenders and holders of MBS. In the most basic mortgage-backed security structure, pass-through participation certificates entitle investors to a “pro-rata share of all principal and interest payments made on the pool of mortgage loans.” The GSEs guarantee these payments in exchange for monthly fees, and “protect investors from defaults on the underlying mortgages.” Homeowners with a fixed-rate mortgage are obligated to make pre-determined monthly payments at a fixed rate, consisting of interest and principal. These mortgages can typically be prepaid at any time, without penalty. A pool of mortgages thus generates monthly “cash flows consisting of scheduled interest, scheduled principal, and possibly prepaid principal.” Mortgage servicers are responsible for collecting these monthly payments from homeowners then forwarding them to the GSEs—these payments are then “passed-through” to the holders of MBS. The mortgage servicers retain a monthly fee, known as a “servicing strip,” based on “a percentage of the outstanding mortgage balance at the beginning of the month.”
When mortgagees prepay their mortgage, the principal value of the underlying security shrinks, as does the value of the servicer’s servicing strip. Prepayment reduces the mortgage value for holders of MBS, as they experience both an early return of principal and a decrease in interest income. Mortgage prepayments are typical when interest rates are declining, as homeowners refinance their mortgages at a lower cost. During periods of declining interest rates, MBS holders enjoy higher-yielding securities in a “low-rate environment.” But when mortgagees refinance and prepay their mortgages, principal is returned to investors who are forced to reinvest the funds at the new lower rate.
After extreme weather events, homeowners can use disaster insurance to pay down their mortgages early. As reported by Politico, research by Fannie Mae economists and Carolyn Kousky shows that homeowners inside the 100-year floodplain prepay their mortgages faster after severe or moderate flood damage than affected homeowners whose properties lay outside of official zones. Following Hurricane Katrina, for example, insurance payments from the NFIP led to overall reductions in household debt. Kousky concludes that mortgage prepayment rates can increase after natural disasters because insured homeowners use “insurance proceeds to pay down debt,” or because “the knowledge of forthcoming insurance payments means they can more quickly sell their homes to investors.” Full insurance in disaster-prone areas could thus substantially lower the value of MBS.
The lack of climate-related information and the potential for extreme-weather events can also lead to an increase in securitization volumes. In the years following natural disasters, mortgage originators are more likely to approve mortgages that can be securitized. Further evidence shows that while lenders “have not meaningfully tightened credit” in areas exposed to SLR, they have likely “increased securitization to transfer their climate risk to the housing GSEs.” The securitization of risky mortgages allows mortgage originators, especially local lenders who may have a more granular understanding of risks, to transfer climate-related risks to the secondary market. This dynamic echoes what occurred with subprime mortgages leading up to the Great Recession, where lenders engaged in an originate-to-distribute model of lending that allowed them to pocket the fees and retain none of the risk.
A paper published by the National Bureau of Economic Research in October 2020 suggests that “in the aftermath of natural disasters, lenders are more likely to approve mortgages that can be securitized.” The authors found a “statistically and economically significant increase in securitization volumes at the conforming loan limit” in the years following large natural disasters, concluding that the probability of securitization by mortgage lenders increases by up to 19.3 percentage points in areas affected by “billion-dollar” disasters. Another study shows that concentrated lenders not only expand their lending growth in recovery locations after extreme weather events, they also increasingly sell those loans on the secondary market. Overall, in the first year following a natural disaster, denial rates for conforming mortgages declineby nearly 3% on average. Three years after the disaster, denial rates drop by nearly 9%. Evidence also suggests that conforming loans originated after an extreme weather event perform worse than similar mortgages originated before the disaster. Mortgages originated one year after a disaster are 3.6% more likely to face foreclosure, and the probability of foreclosure is nearly 5% higher for those originated in the third year after a disaster.
Local mortgage lenders concentrated in high-risk areas are securitizing mortgages at higher rates than non-concentrated national or regional lenders. One study finds that, in coastal regions with high risk of flooding and SLR, non-concentrated lenders are more likely to accept and retainmortgages than concentrated lenders, who are more likely to sell those mortgages on the secondary market. The study advances the “Retention Hypothesis,” which suggests that local lenders, whose business is concentrated in a relatively small number of markets, likely retain fewer mortgage assets on their books than larger lenders with diversified portfolios when those mortgages are collateralized by properties in high-risk areas. Another study shows that in 2009, local banks sold off 43% of their mortgages in vulnerable zones, representing about the same share of mortgages sold by lenders in non-vulnerable areas. By 2017, loan sales by local banks in at-risk areas “had jumped by one-third, to 57%, despite staying flat in less vulnerable neighborhoods.” Though regional lenders may be more vulnerable to regionally concentrated physical risks, as their assets “tend to be more geographically concentrated than the loans of larger banks,” the Retention Hypothesis shows an increase in liquidity among concentrated lenders in SLR-exposed areas following extreme weather events.
One explanation for the increase in securitization following extreme weather events is that local lenders have “superior locally sourced information” about the physical exposure of asset collateral due to their “longstanding and diverse relationships in the communities that they serve.” This unevenness of scientific knowledge and locally sourced “soft information” can result in determinations of risk on the secondary market that may not accurately reflect long-term asset performance or credit loss. This risk is especially true for mortgages securitized and sold to the GSEs. As of 2020, the GSEs account for around 24% of the total volume of exposed loans sold to the secondary market by concentrated lenders within two miles of at-risk coastal regions. The GSEs “do not price predictable regional variation in default risk, including climate risk,” nor do they adjust their securitization rules or guarantee fees in response to flood risk information. Local loan officers have discretion over the characteristics of the mortgages sold for securitization, but the GSE’s guidelines for securitization do not rely on this on-the-ground information. As a result, the GSEs fail to “take into account local climate risk as accurately as the local loan officer with better knowledge of the future distribution of house prices.” In 2019, the GSEs guaranteed over $6.8 trillion in home mortgage debt—without pricing flood risk in their guarantee fees. Concentrated lenders with granular, localized knowledge can thereby securitize loans with the GSEs “at prices that are completely independent of current or future flood risk.” Thus, while these lenders may have more robust information and more sophisticated risk assessment capabilities relative to homebuyers, their incentives are not aligned to internalize SLR risk, as they can “exploit the mis-pricing of risk at the federal level.” Further, absent regulatory standardization, lenders are disincentivized to “share information and cooperatively standardize assessment methodologies” for fear of losing market share and their monopolistic pricing advantages. This market advantage for concentrated lenders will likely become even more significant as SLR and coastal flooding becomes more common.
Finally, not only does the ability to transfer climate risks to the GSEs increase moral hazard, it also potentially weakens “the discipline brought about by the mortgage finance industry in fostering climate change adaptation.” As the GSEs rely only on a “finite vector of observable loan, borrower, and collateral characteristics,” mortgage lenders may not have an incentive to collect the full range of private climate-related information in the underwriting process. But if mortgage lenders could not sell mortgages to the GSEs regardless of climate risk, they would have a greater incentive to fully understand and accurately price those risks.
Extreme weather events can cause drastic declines in property values. Freddie Mac’s Economic & Housing Research Group recently concluded that the housing market is beginning to respond to damages from extreme weather events, and a recent report from the National Bureau of Economic Research shows that climate change has been eroding real estate markets in coastal regions for over a decade. Not only can the physical effects of wildfires and floods cause direct property damage, evidence also shows that increased market perception of future disaster risks leads to a decrease in real estate values. In addition, the annual volume of home-sale transactions has been falling in low-lying coastal areas since 2013, with prices declining a few years later.
A study by First Street Foundation shows that increased tidal flooding between 2005 and 2007 led to $14.1 billion in home devaluations across eight states. The study concluded that 820,000 homes are now worth less than they would have been without increased flooding. The small coastal town of Ocean City, New Jersey, for example, “has lost $530 million in potential increases in property value since 2005.” And according to a 2020 McKinsey report, flood risk has resulted in a total devaluation of $5 billion for at-risk residential properties in Florida when compared with prices for unexposed homes.
Freddie Mac’s Economic & Housing Research Group studied the effect of Hurricane Harvey on housing prices, and found that homes inside the SFHA zone in Harris County Texas sold for 2.3% less than those outside the area even before Harvey. After the storm, that discountrose to 5.5%, and the prices of homes inside the 100-year floodplain fell by 4.6% compared to similar homes outside of the floodplain. The group concluded that this discount is a signal that the market perceives the incremental flood risk prior to a hurricane and prices that risk accordingly. Homebuyers update their flood risk perception following hurricanes, and discount rates rise relative to the pre-hurricane discount. This trend indicates that recent experience with flooding “leads to a perception of increased flood risk.”
These findings are consistent with a study published by the University of Colorado Boulder, which shows that market perception of disaster risk affects property values. The authors conclude that the SLR exposure discount varies depending on the “degree to which inhabitants are worried about the effects of climate change: with more worried areas impounding a significant discount and unworried areas demanding no concessions for SLR exposure.” Evidence suggests that the SLR exposure discount is greatest in markets with sophisticated participants, who are more likely to have a greater understanding of potential flood risk and to incorporate that information into their investment decisions. The study concludes that the “absence of a current house price discount in less sophisticated market segments raises the possibility of a large wealth shock to coastal communities unless strategies are undertaken to mitigate the effects of SLR.”
Climate change will likely increase the number of homes subject to persistent “nuisance flooding,” and the market’s perception of flood risk may correspondingly increase. This threatens to seriously erode property values for entire markets. As an increasing number of properties are exposed to risk of severe and frequent flooding, and as those risks are becoming more apparent to consumers, homeowners may see resale prices drop significantly, even falling to zero if there are no prospective buyers. If market perception of disaster risk affects the value of real estate, it follows that increased disclosure requirements for climate risks could cause an abrupt revaluation in exposed segments of the housing market. Research suggests that in states where flood-risk disclosure is the most stringent, the “flood zone discount” for homes may be up to 4% higher than the national average. Another study found that housing values in flood-prone areas are comparatively lower in states where disclosure laws require flood risk information to be made available to homebuyers beforethey make a closing bid.
One study that assessed the relationship between SLR exposure and changes in housing and mortgage markets from 2001-2020 showed that home-sales in at-risk areas have been consistently contracting for years. The demand for at-risk coastal properties has fallen sharply since 2013, at rates “above and beyond any adjustments currently being made by lenders, insurers, or the GSEs.” The study, which focused on the coastal Florida market, indicates “a sharp decline in transaction volume since 2013 in housing markets most exposed to SLR.” By 2018, the most-SLR-exposed census tracts in Florida had approximately 18% lower transaction volumes compared to their 2001-2012 annual average. For similar homes in low-risk coastal tracts, the transaction volume rose steadily between 2013-2018. Though transaction volumes began to decline notably in high-risk areas in 2013, the prices of homes did not begin to decline until several years later. From 2013-2016, housing prices increased in both exposed and non-exposed areas, “with a relative decline in the high-risk tracts only beginning to emerge in 2018.” This lead-lag relationship, which is “consistent with dynamics at the peak of prior real estate bubbles,” suggests that “prospective buyers have become more pessimistic about climate change risk” than prospective sellers. The authors conclude that sellers likely remain optimistic about their home’s value “because they overestimate demand by extrapolating from recent price increases,” while systemically underestimating the SLR risk discount demanded by increasingly anxious buyers. As a result, prices are set higher than most prospective buyers are willing to pay.
As climate risks become more salient, market participants are becoming more aware of SLR risk. As a result, prospective buyers will become increasingly wary of at-risk coastal markets, and transaction volumes and property values in at-risk areas will continue to decline.
Extreme weather events can lead to heightened mortgage delinquency, forbearance, and prepayment rates. The mispricing of SLR and wildfires can lead to sudden decreases in mortgage values, having devastating downstream effects for financial market participants, including banks that hold mortgages on their balance sheets and investors in MBS. Moreover, an abrupt increase in mortgage defaults in flood- and fire-zones could cause massive losses for the GSEs—losses that could ultimately be borne by taxpayers. An increase in unexpected—or rather, mispriced—natural disasters can also cause sudden downward adjustments in property values.
Nuisance flooding and increases in storm surges and wildfires threaten to cause rapid property re-valuations in areas that have failed to properly account for those risks. But without robust disclosure requirements for climate-related information, that risk can be passed along to buyers who have inadequate information to demand an accurate price discount. Moreover, if the GSEs fail to adjust their securitization rules and guarantee fees, mortgage originators with superior information will continue to arbitrage climate-risks and pass them on to the secondary market. Absent public policy aimed at incorporating these risks into asset values, participants will continue to play “hot-potato” with overvalued houses and mortgages, hoping to pass them along to unknowing buyers before disaster-risk is accurately—and abruptly—priced.
Importantly, a sudden increase in the measuring, pricing, and disclosure of climate-related risks could cause an overcorrection amongst market participants. Though there is currently a weak capitalization of disaster risk in housing prices, values decline notably following an increase in perception of those risks. Lenders who become weary of disaster risk may demand higher rates for longer-term mortgages, or may retreat from the 30-year mortgage altogether—putting homeownership “out of reach” for many Americans. Sudden home devaluations could also rapidly strand property values, as well as the “large portfolios of assets” that they back.
Public policy should thus balance the need for the accurate incorporation of climate-related risks into asset values with the danger of causing an abrupt over-correction in the market. The GSEs have historically been “circumspect in discussing the potential threats to mortgages on homes” from the increasing risk of climate-related natural disasters. This hesitation is likely a “signal that the two companies are wary of spooking the housing market.” But the increase in the securitization rates of risky mortgages and the steady decline in transaction volumes and sale prices are smoke signals, indicating that the housing market is already reacting to climate-driven disaster risks. This smoke is a sure sign of fire.
Charlie Wowk is a research assistant with the Climate Risk Disclosure Lab