GERALD HERBERT/AP PHOTO
Damage from Hurricane Ida in Louisiana is part of a record-breaking year for losses from natural disasters.
This article appears in the September/October 2021 issue of The American Prospect magazine. Subscribe here.
This year has already been a blockbuster for natural-catastrophe damages. Payouts owed by insurers hit a ten-year high in the first half of 2021, according to insurance broker Aon. Winter Storm Uri, with $15 billion in damages, drove record losses. Hurricane Ida is also expected to cost insurers billions.
The record-breaking losses are partly due to more frequent severe weather events, but they’re worsened by maladaptation: People are moving toward, rather than away from, exposed areas.
In California, deadly risks have not deterred people from building in the fire-prone zone where houses border grasslands and forests. That’s accelerated by a shortage of affordable housing that has pushed Latino populations in farmwork and the service industry to the hinterlands. Meanwhile, residential real estate in hurricane-prone Florida and Texas is on a tear.
For decades, FEMA has underwritten the boom in property and population around exposed coasts and the high-growth Sun Belt. The agency’s National Flood Insurance Program (NFIP) has long been underpriced and regressive, artificially lowering premiums, masking flood risk, and directing bigger subsidies to richer homeowners. As coastal populations get wealthier, property values have risen, and their political entrenchment deepens.
The program has required multibillion-dollar bailouts just to keep up with current claims. For years, unlikely allies—environmentalists, fiscal conservatives, and ordinary taxpayers—have pressed to reform it. Congress made an attempt with the Biggert-Waters Act of 2012, which brought premiums more in line with risk. But when real estate firms and coastal homeowners stuck with higher premiums balked, those reforms were rolled back two years later. Since then, Congress has only managed short-term reauthorizations.
Ultimately, market price signals aren’t likely to be enough. States will need to fund managed retreat from the riskiest areas.
In October, FEMA will implement a new pricing system called Risk Rating 2.0, the agency’s first effort to make premiums fairer since the gutted attempt a decade ago. (The entire program is due to be reauthorized September 30, when most likely lawmakers will kick the can down the road, but rate raises will go into effect either way.) Billed as “equity in action,” the rating system will raise premiums for most NFIP policyholders.
Lawmakers in coastal states like Sen. Bob Menendez (D-NJ) and Senate Majority Leader Chuck Schumer (D-NY) have already signaled their opposition. Rep. Maxine Waters (D-CA), who chairs the House Financial Services Committee, has said she plans to “oppose any efforts to substantially raise premiums or to otherwise add to the affordability burdens.”
The point of accurately priced flood insurance, though, is precisely to make dangerous homes unaffordable. Adequately high prices would be a proxy for current and future risk, helping home buyers make smarter decisions.
Risk Rating 2.0 is supposed to put the NFIP on track to solvency, and put individual homes on a path to pay their full risk. But annual rate raises are capped at 18 percent—meaning it will take years to price in that risk. And political opposition is just the first hurdle.
A deeper reckoning is just beginning to unfold over long-term flood risk, which current FEMA models don’t even claim to account for. Internalizing that vulnerability, however, would break the basic business model of the insurance industry.
And even as the agency argues that it is finally playing hardball with rich homeowners, FEMA has quietly developed a different strategy: transferring the biggest risks from taxpayers to private insurance and bond markets.
FEMA ADVERTISES THAT ITS new risk rating program incorporates private-sector data sets, more dynamic modeling, and “evolving actuarial science.” But the agency buys its actuarial models from a handful of legacy catastrophe (“cat”) risk firms, such as AIR Worldwide and CoreLogic, which are being overtaken by a new crowd of climate risk modelers.
Jupiter Intelligence, a leading voice in the emergent climate risk services industry, was part of a research syndicate at FEMA that helped develop Risk Rating 2.0. As climate modelers, Jupiter were the black sheep in a room of more traditional “cat risk” experts. They’re disappointed with the final product.
“I think most people inside FEMA would like to be moving much faster than they have the resources or permission to do,” Jupiter CEO Rich Sorkin told the Prospect. “They know that they can’t get ahead of the political process.”
Cat risk modelers feed storm trends and geophysical data into computer algorithms to spit out disaster scenarios. In step with the insurance industry, they focus on expected losses in the next year, rather than the more uncertain long term, and have resisted incorporating climate change predictions, arguing that there are too many unknowns to produce granular models.
Climate risk modelers like Jupiter argue that the uncertainties of future climate scenarios are precisely the reason rates should reflect a longer, more conservative time horizon. New perils may not only be more intense, but also freakishly different from past storms.
Tom Larsen, an actuary with CoreLogic, another firm that helped FEMA develop its new risk model, disputed the advisability of incorporating climate-modified risk into models, given current unknowns in how climate heating will play out. “The most appropriate risk for today is the risk for today, not the risk that’s going to be here in 30 years,” he said. “Climate change is a slow progression, and certainly the risk will change. They [the new rates] are appropriate for now.”
Trying to incorporate long-run projections would make modelers unduly cautious and drive up premiums, Larsen said. “Our job is to try to strip away any conservatism.”
Larsen and Sorkin’s disagreement is part of a brewing clash among modelers of natural catastrophes.
Under pressure, cat modelers have begun paying lip service to long-term risk. But they continue to work under the assumption that if severe weather patterns worsen dramatically, that will become relevant only later in the century, not in the next several years. That assumption is turning out to be wrong.
Catastrophe models are “purpose-built” for the insurance industry, so in order to incorporate long-term risk, they would have to be overhauled, said Nancy Watkins, a leading cat risk expert who is a principal at Milliman, an actuarial consulting firm founded in 1947. She cautioned against that idea.
“Current rates should reflect current risk,” Watkins said. “People would rather buy five one-year policies than one five-year policy that costs more.”
SEAN BOURGEOIS IS in the business of “bringing risk to capital.” Given the global search for yield and the uptick in natural disasters, the insurance trading executive with Tremor Technologies thinks there are many potential buyers for FEMA’s exposure to severe weather.
After purchasing billions in reinsurance, in 2018, FEMA entered the insurance-linked securities market, issuing its first catastrophe bond.
The reinsurance industry provides insurance to insurers, capping an individual company’s risk. Once a threshold is crossed—for example, if an insurance company owes an unusually high sum in payouts—the insurer no longer has to pay claims; it kicks them up to its reinsurer.
Reinsurance firms make this work by being global and extremely diversified, so that they can absorb losses even when a major event wipes out a whole area. But climate change has the effect of correlating risk around the world, making reinsurers’ job harder.
In that environment, cat bonds are encroaching on reinsurers’ turf. The instruments follow a similar logic to reinsurance, securitizing the risk of a peril, such as an earthquake or wildfire, and packaging it for sale to nonspecialist capital market investors. They were inspired by mortgage-backed securities (MBS), the pooled loans that set off the 2008 global financial crisis.
Increasingly, it’s not insurance institutions staffed by actuaries that are buying these bonds, but insurance-linked securities (ILS) funds, which raise large pools of capital from pension and sovereign wealth funds. Low-cost trading platforms help investors buy and sell ILS more quickly than they would be able to if they were investing directly in a reinsurer.
Bourgeois also envisions an increase in “live cat” trading—real-time buying and selling of insurance-linked securities while a storm approaches. Eventually, he hopes Tremor will achieve enough liquidity to help traders adjust their portfolios “as a hurricane is in the water.”
In making their pitch to investors, ILS funds argue that since natural disasters occur randomly, cat bond risks are uncorrelated with swings in financial markets. Asset manager Neuberger Berman has pithily dubbed cat bonds “natural diversification.”
That calls to mind arguments for MBS, where the buffer was also supposed to be regional price differentiation. Neighborhoods might bust, but, the story went, you wouldn’t see the entire market blow up.
MBS turned out to be more linked than expected. Climate change, too, has the effect of tethering once-independent regional risks.
This doesn’t necessarily mean that cat bonds threaten another financial crisis. FEMA’s heavier reliance on reinsurance and cat bonds does, however, help the program continue extending its “duration mismatch” problem: The short-term time horizon of the insurance industry does not match the multi-decade duration of the assets being insured.
Most cat bonds have a maturity period of under five years, and can be rolled over, potentially creating the false impression that insurance costs will remain low. As a result, homeowners can insure buildings in high-risk areas by rolling over indefinitely. This could explain why real estate firms have lobbied FEMA for more cat bond issuances. Nothing in the system cautions against unsustainable development.
Cat bonds have a final, striking resemblance to MBS: their treatment by credit rating agencies. In selecting cat bonds, investors depend on rating agencies like Moody’s, Fitch, and AM Best, a ratings firm that specializes in the insurance industry.
But it’s notoriously tricky to assign credit ratings to cat bonds. Credit rating agencies mostly compare firms’ outcomes, rather than assessing the methodologies themselves.
“We are agnostic to risk models,” said Sridhar Manyem, research director at AM Best, who previously rated insurance companies for Standard & Poor’s. “We’re not telling insurance companies what they need to use.”
Rating agencies may make clear in fine print that they assess whether insurers have enough capital to pay claims, not their actuarial methodologies. But their rating nonetheless implies a level of trust and certainty. There has been some saber-rattling from credit rating agencies after recent losses from storms, but the bonds haven’t been downgraded.
FEMA has entered the insurance-linked securities market by issuing a catastrophe bond.
CHRIS HEIDRICK, AN INSURANCE AGENT on Sanibel, a Florida barrier island, is not too worried about expected rate raises for his clients, most of whom are wealthy enough to pay cash for their homes. What sets him at ease is the pace of change: Even as rates come up, premiums can’t jump by more than 18 percent per year.
Congress may even lower that cap, further slowing the rate at which homeowners feel new prices kick in.
Either way, the rates rich homeowners pay will take years to catch up to new sticker-price premiums. Some could never catch up, since the new rate is itself a moving target, likely to increase in future years.
FEMA has even included a “glide path,” so that the new buyer of a house can pick up on the same path to paying full price as the previous owner, rather than immediately paying full cost. That will keep FEMA subsidies to rich homeowners intact for some time.
Meanwhile, even modest rate raises are sure to make insurance harder to purchase for poorer families living in exposed areas. When premiums become unpayable, low-income households typically let flood insurance policies lapse, opting out altogether.
That dreaded outcome has made modelers wary of pricing long-term risks into coverage.
“If you were to take the most conservative estimate of the risk, you are potentially overpricing it,” Larsen said. “Regulators in the U.S. are really focused on fairness.”
But making insurance prices “fairer,” short-term, can cloud true flood risk.
Where flood insurance is mandatory, there is widespread noncompliance. Just 1 in 3 homes in the highest-risk areas have flood insurance coverage, according to the Wharton Risk Center. Compounding the problem, some of the highest-risk states, like New York, New Jersey, and Florida, have few or no requirements to disclose flood risk.
Some have proposed shifting part of the climate risk burden to the Federal Housing Finance Agency. Home costs, underwritten by mortgage giants Fannie and Freddie, could price in exposure to long-term climate-driven losses.
Ultimately, market price signals aren’t likely to be enough. States will need to fund managed retreat from the riskiest areas.
FEMA has some money to move whole poor communities up coastlines. But this, too, is stacked with political dilemmas. Authorities want to give lower-income communities priority access to grant money for coastal retreat. But pressuring low-income neighborhoods has an ugly history. Done cheaply, it could look like slum clearance or failed “urban renewal” schemes.
Done well, managed retreat from blighted areas is an opportunity to address the housing crisis and build sustainable, beautiful neighborhoods. Instead, lawmakers have cynically deployed arguments for fairness to continue subsidizing the status quo.