The Centers for Disease Control and Prevention this week provided new data on the spread of COVID-19 that diverges sharply from past reports and is something health and workers compensation insurance providers will want to incorporate into their claims projections.
In its Morbidity and Mortality Weekly Report, the CDC says that between June and August the virus was most prevalent in people between the ages of 20 and 29, accounting for more than 20 percent of all confirmed cases. It went on to say that “across the southern United States in June 2020, increases in percentage of positive [COVID-19] test results among adults aged 20-39 years preceded increases among those aged ≥60 years” by between four and 15 days.
Most of the workforce
“This has profound implications for claims made against health insurance and workers comp,” says Dr. Steven N. Weisbart, CLU, Triple-I’s senior vice president and chief economist. “Early in the pandemic, COVID-19 was most common among adults age 70 or older – people who are mostly retired. Now, the CDC says, more than 50 percent of confirmed cases during the referenced period were among people between 20 and 49. This is the segment of the population that makes up most of the workforce and tends to have health and life insurance.”
They also are the most mobile portion of the population, more likely than the elderly and infirm to spread the infection to co-workers, friends, and family before they know they have it.
Indicating how significant the shift has been, Weisbart points out that in May the most affected age group was still 80 and older, with a case incidence of 4.04 per 1,000 population. In August the most affected age group was 20-29 (case incidence: 4.17 per 1,000).
“By August,” Weisbart says, “the case incidence of the 80-plus group was down to 2.61 per 1,000.”
Expanded workers comp coverage
The ultimate impact of the pandemic on workers compensation is still not clear. It generally doesn’t cover illnesses like a cold or flu because they can’t be tied to the workplace. Before the pandemic, the National Council on Compensation Insurance (NCCI) says, at least 18 states had policies that presumed firefighters’ and other first responders’ chronic lung or respiratory illnesses are work-related and therefore covered.
Since the pandemic, some states have extended coverage to include health care workers and other essential employees. A common approach is to amend state policy so COVID-19 infections in certain workers are presumed to be work related. This puts the burden on the employer and insurer to prove the infection was not work-related, making it easier for workers to file successful claims.
Dr. Steven N. Weisbart, CLU, Triple-I Senior Vice President and Chief Economist
The Federal Open Market Committee (FOMC) of the Federal Reserve Board recently spelled out its objectives and strategies for at least the next several years—describing a financial framework they will maintain longer than the timeframe they typically describe. The length and parameters of this framework will have significant impact on the property/casualty industry.
The FOMC says it will hold short-term interest rates near zero, likely for several years—perhaps to 2023, quite possibly longer. Insurers don’t invest much in short-term instruments – to the extent that they do, it’s to have cash available to pay claims. They primarily invest in intermediate- and longer-term bonds and similar fixed-rate interest-paying instruments that provide steady income, which, together with premiums, covers claims and operating expenses. Insurers raise and lower premiums – partly in response to changes in investment income – to sustain profitable operations.
Because yields on these investments generally track short-term rates, the FOMC’s action will likely keep longer-term rates exceptionally low for several years more.
One signpost the Fed will use to decide when to raise rates is when inflation, as measured by the Personal Consumption Expenditure (PCE) deflator, is sustained at over 2 percent such that the average inflation rate including recent years equals 2 percent. To appreciate what this means, consider Figure 1.It shows that, since 2012, the PCE deflator has been below 2% (vs. same month, prior year) most of the time. The average over this span was 1.40%. But the Fed might not go back that far to calculate its long-run average. For example, since 2017 the PCE deflator averaged 1.69%. If the deflator averages 2.4% from now through 2023, the average from 2017 through 2023 will be 2.01%.
Figure 1
Rates falling since the 1980s
Based on the FOMC’s new framework, intermediate- and longer-term interest rates will, at best, remain at their current historically depressed levels for several years. One consequence of this is that bonds insurers hold to maturity and roll over will be reinvested at lower rates than they currently yield.
Prevailing interest rates have been generally falling since the early 1980s. Figure 2 shows this decline since 2002, as proxied by the yield on constant-maturity 10-year U.S. Treasury notes (the blue line), and its effect on the portfolio yield for the P/C insurance industry over the last two decades (the gold bars).
Figure 2
P/C insurers invest mainly in bonds, but not just U.S. Treasury securities. They also invest in corporate and municipal bonds, both of which generally yield higher rates than U.S. Treasury bonds because they are riskier. Yields on corporate and municipal bonds will likely loosely track Treasury yields.
P/C insurers also receive investment income from dividends on common and preferred stock they hold. These dividends are likely to be affected by corporate profits, which might be depressed for at least as long as the current recession lasts.
A shift to shorter maturities?
How will the insurers respond to these persistent conditions? If recent behavior is any guide, they are likely to shift to shorter-maturity bonds to retain the flexibility to switch back to longer-term, higher-yielding investments when rates eventually rise again. Figure 3 shows this pattern of shortening maturities during the years since 2009 as prevailing rates fell. From 2009 to 2019, the percent of bonds with one-to-five-year maturities rose from 36% to 41%, but those with 10 or more years of maturity fell from 19% to 11%.
Figure 3
What’s notable about this strategy is that – since shorter-term bonds yield less than longer-term bonds – the shift results in an even lower portfolio yield than the industry would have achieved if maturities were unchanged over this time span. It sacrifices near-term opportunities for the flexibility to eventually seize longer-term gains.
If the insurers continue this strategy, the shift to shorter-term bonds, combined with continued low interest rates, could lead to a scenario over the next five years that looks like Figure 4, which includes 2015-2019 yields for historical context.
Figure 4
Of course, future portfolio yields might be different from this scenario. For example, insurers might realize significant capital gains or losses. The portfolio yield in 2012, for example, was nearly two percentage points above the U.S. Treasury 10-year yield that year due to realized capital gains.
On the other hand, if interest rates rise, low-yielding bonds that are available for sale would suffer unrealized capital losses, which would be a direct reduction in policyholder’s surplus.
In a typical year the industry posts capital gains of $5 billion to $10 billion, but any number outside this range would affect the portfolio yield for that year. Capital losses also could result from investments affected by bankruptcies or other business setbacks caused by the recession. Impaired bonds would have to be accounted for on the balance sheet.
The U.K. High Court last week issued a ruling involving business-interruption claims against policies issued by eight insurers. Jason Schupp of the Centers for Better Insurance says the ruling is a “mixed bag” for U.K. insurers and policyholders and has little relevance for their U.S. counterparts.
In the U.K. case, Schupp writes, “the fundamental theme running through the insurers’ defense was that the policies only covered localized outbreaks, not global pandemics.”
“More to the point for U.S. property/casualty insurers,” says Michael Menapace, a professor of insurance law at Quinnipiac University School of Law and a Triple-I non-resident scholar, the U.K. case involved disease coverage – “an affirmative coverage not included in most U.S. commercial property policies.”
U.S. business interruption disputes so far have turned on two key policy features:
U.S. business-interruption coverage almost always requires property damage to trigger a payout.
Nearly all U.S. COVID-19-related court cases have involved policies that specifically exclude viruses.
“The U.K. court did not address either the question of property damage or the applicability of a virus exclusion,” Schupp writes.
As Menapace put it in a recent blog post about U.S. business-interruption cases, “Policy language controls whether COVID-19 interruptions are covered…. The threshold issue [for U.S. insurers] will be whether the insureds can prove their business losses are caused by ‘physical damage to property’.”
Rivers swollen by Hurricane Sally’s rains have devastated parts of the Florida Panhandle and south Alabama, and the storm’s remnants are forecast to spread the flooding to Georgia and the Carolinas.
Many of the properties damaged will doubtless be found to be uninsured, compounding homeowners’ misery.
A well-known coverage gap
The flood insurance protection gap has been well documented. A recent Triple-I paper – Hurricane Season: More Than Just Wind and Water – states that “about 90 percent of all natural disasters in the United States involve flooding” and cites experts strenuously urging everyone to buy flood insurance.
“Any home can flood,” says Dan Kaniewski — managing director for public sector innovation at Marsh & McLennan and former deputy administrator for resilience at the Federal Emergency Management Agency (FEMA). “Even if you’re well outside a floodplain…. Get flood insurance. Whether you’re a homeowner or a renter or a business — get flood insurance.”
Dr. Rick Knabb — on-air hurricane expert for the Weather Channel, speaking at Triple-I’s 2019 Joint Industry Forum — is similarly emphatic.
“If it can rain where you live,” he said, “it can flood where you live.”
Despite such warnings, even in designated flood zones, the protection gap remains large. A McKinsey & Co. analysis of flood insurance purchase rates in areas most affected by three Category 4 hurricanes that made landfall in the United States — Harvey, Irma, and Maria — found that as many as 80 percent of homeowners in Texas, 60 percent in Florida, and 99 percent in Puerto Rico lacked flood insurance.
To make matters worse, a recent analysis by the nonprofit First Street Foundation found the United States to be woefully underprepared for damaging floods. The report identified “around 1.7 times the number of properties as having substantial risk,” compared with FEMA’s designation.
“This equates to a total of 14.6 million properties across the country at substantial risk, of which 5.9 million property owners are currently unaware of or underestimating the risk they face,” the foundation says.
A more recent Triple-I analysis, conducted in advance of Hurricane Sally, found that flood insurance purchase rates in the counties most likely to be affected by the storm were “remarkably low.”
“In Taylor County, Ga., for example, just 0.09 percent of properties are insured against flooding,” Triple-I wrote.
NOT covered by homeowners insurance
Flood damage is excluded under standard homeowners and renters insurance policies. However, flood coverage is available as a separate policy from the National Flood Insurance Program (NFIP), administered by FEMA, and from a growing number of private insurers, thanks to sophisticated flood models that have made insurers more comfortable writing this once “untouchable” risk.
Invest in resilience
If it seems as if you’ve heard me beat this drum before, you’re right. I take flood and flood insurance very personally.
After Hurricane Irene flooded my inland New Jersey basement in August 2011, destroying many irreplaceable items, it was my flood insurance that enabled me to have a French drain and two powerful pumps installed that have since kept my historically damp basement bone dry – even during Superstorm Sandy the following year.
Perhaps the most emotionally compelling data point invoked by those who would compel insurers – through litigation and legislation – to pay business-interruption claims explicitly excluded from the policies they wrote is the property/casualty insurance industry’s nearly $800 billion policyholder surplus.
Many Americans hear “surplus” and think of a bit of cash they have stashed away for emergencies. And when you consider that nearly 40 percent of Americans surveyed by the Federal Reserve said they would either have to borrow or sell something to cover an unexpected $400 expense – or couldn’t pay it at all – that number may sound like overkill.
Not as much as you think
But policyholder surplus isn’t a “rainy day fund.” It’s an essential part of the industry’s ability to keep the promises it makes to policyholders. And although a number like $800 billion may raise eyebrows, when we look more closely at its components, the amount available to cover claims turns out to be considerably less.
Insurers are regulated on a state-by-state basis. Regulators require them to hold a certain amount in reserve to pay claims based on each insurer’s own risk profile. The aggregation of these reserves – required by every state for every insurer doing business in those states – accounts for about half the oft-cited industry surplus.
Call it $400 billion, for simplicity’s sake.
Each company’s regulator-required surplus can be thought of as that company’s “running on empty” mark – the point at which alarms go off and regulators start talking about requiring it to set even more aside to make sure no policyholders are left in a lurch.
By extension, $400 billion is where alarms begin going off for the entire industry.
It gets worse – or better, depending on your perspective.
In addition to state regulators’ requirements, the private rating agencies that gauge insurers’ financial strength and claims-paying ability don’t want to see reserves get anywhere near “Empty.” To get a strong rating from A.M. Best, Fitch, S&P, or Moody’s, insurers have to keep even more in reserve.
Why do private agency ratings matter? Consumers and businesses use them to determine what insurer they’ll buy coverage from. Also, stronger ratings can contribute to lower borrowing expenses, which can help keep insurers’ operating costs – and, in turn, policyholders’ premiums – at reasonable levels.
So, let’s say these additional reserves amount to about $200 billion for the industry. The nearly $800 billion surplus we started with now falls to about $200 billion.
To cover claims by all personal and commercial policyholders in a given year without prompting regulatory and rating agency actions that could drive up insurers’ costs and policyholders’ premiums.
Which brings us to today.
Losses ordinary and extraordinary
In the first quarter of 2020, the industry experienced its largest-ever quarterly decline in surplus, to $771.9 billion. This decline was due, in large part, to declines in stock value related to the economic recession sparked by the coronavirus pandemic.
Nevertheless, the industry remains financially strong, in large part because the bulk of insurers’ investments are in investment-grade corporate and governmental bonds. And it’s a good thing, too, because the conditions underlying that surplus decline preceded an extremely active hurricane season, atypical wildfire activity, and damages related to civil unrest approaching levels not seen since 1992 – involving losses that are not yet reflected in the surplus.
Insured losses from this year’s Hurricane Isaias are estimated in the vicinity of $5 billion. Hurricane Laura’s losses could, by some estimates, be as “small” as $4 billion or as large as $13 billion.
And the Atlantic hurricane season has not yet peaked.
The 2020 wildfire season is off to a horrific start. From January 1 to September 8, 2020, there were 41,051 wildfires, compared with 35,386 in the same period in 2019, according to the National Interagency Fire Center. About 4.7 million acres were burned in the 2020 period, compared with 4.2 million acres in 2019.
In California alone, wildfires have already burned 2.2 million acres in 2020 — more than any year on record. For context, insured losses for California’s November 2018 fires were estimated at more than $11 billion.
And the 2020 wildfire season still has a way to go.
All this is on top of routine claims for property and casualty losses.
Four billion here, 11 billion there – pretty soon we’re talking about “real money,” against available reserves that are far smaller than they at first appear.
No end in sight
Oh, yeah – and the pandemic-fueled recession isn’t expected to reverse any time soon. Economic growth worldwide remains depressed, with nearly every country experiencing declines in gross domestic product (GDP) – the total value of goods and services produced. GDP growth for the world’s 10 largest insurance markets is expected to decrease by 6.99 percent in 2020, compared to Triple-I’s previous estimate of a 4.9 percent decrease.
If insurers were required to pay business-interruption claims they never agreed to cover – and, therefore, didn’t reserve for – the cost to the industry related to small businesses alone could be as high as $383 billion per month.
This would bankrupt the industry, leaving many policyholders uninsured and insurance itself an untenable business proposition.
Fortunately, Americans seem to be beginning to get this. A recent poll by Future of American Insurance and Reinsurance (FAIR) found the majority of Americans believe the federal government should bear the financial responsibility for helping businesses stay afloat during the coronavirus pandemic. Only 16 percent of respondents said insurers should bear the responsibility, and only 8 percent said they believe lawsuits against insurers are the best path for businesses to secure financial relief.
The COVID-19 pandemic continues to depress economic growth around the world, with nearly every country experiencing declines in gross domestic product (GDP) – the total value of goods and services produced. GDP growth for the world’s 10 largest insurance markets is expected to decrease by 6.99 percent in 2020, compared to Triple-I’s previous estimate of a 4.9 percent decrease.
Forward-looking growth proxies, such as interest rates, government spending, equity markets, and commodity prices, are sending mixed to negative messages about growth into 2021.
Against this backdrop, Triple-I experts report, the global insurance industry has continued to issue new policies, service existing ones, and process and pay claims. While the final numbers on the extent of the pandemic and recession’s impact on the industry won’t be clear until 2021-2022, early indicators point to flat premium growth in 2020 globally and to significant differences in how the pandemic, monetary policy, and the recession are affecting insurers in the United States versus abroad.
In its Global Macro and Insurance Outlook for the third quarter, published this week,Triple-I noted that global central banks kept benchmark interest rates mostly on hold in the third quarter at an average of 0.6 percent, reflecting the limits imposed by near-zero interest rates policies.
Concerns about lower long-term interest rates are increasing as global central banks have pushed rates even lower during the pandemic, the report says. In a recent survey, about 33 percent of U.S. insurers said they assume flat long-term benchmark rates, while 50 percent reported having changed, or say they are in the process of changing, their investment strategy. These changes are likely to accelerate now that the U.S. Federal Reserve officially changed the focus of its monetary policy and central banks around the world follow.
Interest rates matter because insurers get the bulk of their profits from investment earnings. U.S. insurers, in particular, rely on fixed-income financial instruments like corporate and government bonds. If lower interest rates put pressure on insurers’ investment earnings, they will have to compensate by raising premiums paid by policyholders or adjusting their risk profiles to reduce claim payouts.
“COVID-19 and lower economic activity continue to hinder premium growth in property, workers compensation, and auto,” the report says, “while a recent survey indicates that COVID-19 led to a reduction in life premium.”
The report says it’s too early to determine whether increasing demand for warranty, indemnity, and cyber coverage and a surge of interest in captive insurers will make up for downward pressure on premium growth across the industry.
The Los Angeles office of the National Weather Service predicted prolonged, potentially record-setting heat and dangerous weather conditions throughout California this summer – and, some experts expect it to continue for some time beyond.
“If you like 2020, you’re going to love 2050,” said Michael Gerrard, director of Columbia University’s Sabin Center for Climate Change Law, in a recent Los Angeles Times article.
These conditions can only exacerbate this year’s atypical wildfire activity in the state. So, it should be no surprise that California is grappling with how to stop insurers from abandoning fire-prone areas, leaving countless homeowners at risk.
“Years of megafires have caused huge losses for insurance companies, a problem so severe that, last year, California temporarily banned insurers from canceling policies on some 800,000 homes in or near risky parts of the state,” The New York Times reports. “However, that ban is about to expire and can’t be renewed, and a recent plan to deal with the problem fell apart in a clash between insurers and consumer advocates.”
Insurers are widely expected to continue their retreat.
“The marketplace has largely collapsed” in high-risk areas, said Graham Knaus, executive director of the California State Association of Counties. “It’s a very large geographic area of the state that is facing this.”
California, where regulations lean toward consumer protection, is particularly challenged. The state doesn’t let insurers set premiums based on what they expect in future damages. They can only set rates based on prior losses. They also aren’t allowed to pass along reinsurance costs to policyholders – costs that are expected to rise as fire risks worsen.
State lawmakers introduced a bill to let insurers writing coverage in wildfire-prone areas incorporate climate predictions and other costs into their rate requests in return for making coverage more available and offering discounts to people who take steps to reduce their home’s vulnerability.
Insurers supported the change, as did the counties association and the union representing firefighters. But the bill faced strong opposition from consumer groups, who ultimately prevailed. Last month, the state senate stripped most of the provisions from the bill and directed the insurance commissioner to review the current rules and report back to the legislature in two years.
The legislature ended its session without acting on the revised version. Insurance Commissioner Ricardo Lara said his focus now is working with high-risk communities to reduce their wildfire risk enough that insurers will keep offering coverage without big rate increases.
“If Californians do our part to protect homes from wildfire,” Lara said, the industry should respond by agreeing to insure those homes.
Janet Ruiz, Triple-I’s director of strategic communications, said, “Insurers in California are working with legislators and the California Department of Insurance to find solutions to keep homeowners insured in wildfire risk areas. The industry supports mitigation efforts, the California FAIR Plan, and the proposed IMAP program.”
This blog post has been updated based on new information received since it was first published on September 4, 2020.
Hurricane Laura may have caused as little as $4 billion of insured damage or as much as $13 billion, according to early estimates.
Property information, analytics and data provider CoreLogic said residential and commercial insured losses from Hurricane Laura in Louisiana and Texas will come in at between $8 billion and $12 billion. Most of the property damage occurred in southwest Louisiana, where Laura made landfall early as a Category 4 hurricane with 150 mph winds.
Catastrophe risk modeler RMS puts the range between $9 billion and $13 billion. This includes wind and storm surge losses of between $8.5 billion and $12 billion, inland flood losses of $100 million to $400 million, and National Flood Insurance Program (NFIP) losses of $400 million to $600 million.
Catastrophe risk modeling firm Air Worldwide said it expects losses related to Laura to fall in the $4 billion to $8 billion range. The combination of the storm’s track through less-populated areas and its relatively small “Rmax” – the distance from the center of the storm to the location of the maximum winds – should keep insured losses down somewhat, the company said.
Cat risk modeler Karen Clark & Co. estimates insured onshore property losses from wind and storm surge will likely amount to $8.7 billion in the U.S. and $200 million in the Caribbean. Its estimate includes the privately insured wind and storm surge damage to residential, commercial, and industrial properties and automobiles but not losses covered by the NFIP or losses to offshore assets.
All estimates are subject to change as more information becomes available.
Under the best of circumstances, the Atlantic hurricane season is a challenging time. Despite improved forecasting and analytical tools, pre-storm communication, and engineering, hurricane-related losses continue to climb.
But the 2020 season hasn’t come during the best of circumstances. This extremely active season arrived on the heels of a pandemic that hasn’t ebbed, accompanied by civil unrest and atypical wildfire activity that could draw attention and resources away from preparation and post-storm aid.
And, as if that wasn’t enough, it falls in the middle of what is arguably the most contentious, chaotic U.S. election year in modern history.
To say these new variables complicate resilience would be a gross understatement in a year whose (to use the technical insurance phrase) “general weirdness” would be difficult to overstate.
So, in a paper published today we review the current state of hurricane resilience – how forecasting, modeling, preparation, and mitigation have evolved – and how the insurance industry is working to help communities bounce back from hurricanes.
Demographics more than climate change
Nine of the 10 costliest hurricanes in U.S. history have occurred since 2004, and 2017, 2018, and 2019 were the largest back-to-back-to-back insured property loss years in U.S. history. Many would instinctively chalk up such numbers to climate change. But a careful look at the data suggests climate change isn’t the predominant driver of losses.
U.S. Census Bureau data indicate that the number of housing units in the United States increased most dramatically since 1940 in areas that are most vulnerable to weather-related damage. They also show that new homes are bigger and more expensive than in past decades.
Bigger homes full of more valuables, more cars and infrastructure in disaster-prone locales – these, more than climate trends, seem to be the dominant factors driving losses.
Not more, but wetter
Hurricanes may not be more frequent or significantly more intense due to climate change, but they seem to be getting wetter. Inland flooding has caused more deaths in the United States in the past 30 years than any other hurricane-related threat.
Early in the 2020 season, Tropical Storm Cristobal made landfall along southeastern Louisiana and triggered flash flooding as far inland as northwest Wisconsin.
“As the atmosphere continues to warm, storms can hold more moisture and bring more rainfall,” said Triple-I non-resident scholar and Colorado State University atmospheric scientist Dr. Philip Klotzbach. This trend could be exacerbated if, as some experts expect, storms begin traveling more slowly, adding to the moisture they would pick up from the ocean and drop over land.
Our paper also looks at the evolution of hurricane modeling and forecasting, as well as developments in preparation and mitigation.
Better data and improved modeling have made private insurers comfortable writing coverage, like flood insurance, that was previously considered “untouchable” and enabled the creation of entirely new types of insurance products.
But challenges remain. Experts disagree as to which models are best, and the proprietary nature of these models can make it hard for regulators to determine whether filed rates based on them are unfairly discriminatory.
Hurricane preparation and damage mitigation have benefited from improved communication and public planning.
“Many people still don’t evacuate the way they should,” says Todd Blachier of Church Mutual Insurance, “but states like Louisiana, Florida, Alabama and Mississippi have gotten much better in terms of shutting down inbound roads and creating one-way egress to facilitate evacuation.”
He says officials are acting much more quickly and communicating more effectively, thanks in large part to improved information from the National Oceanic and Atmospheric Administration (NOAA) and other resources.
One area in which improvements could boost resilience is building codes and standards. A recent Federal Emergency Management Agency (FEMA) study quantified the losses avoided due to buildings being constructed according to modern, hazard-resistant codes and standards. In California and Florida — two of the most catastrophe-prone states — FEMA found adopting and enforcing modern codes over the past 20 years led to a long-term average future savings of $1 billion per year for those two states combined.
Damage reports from Hurricane Laura are coming in. Lake Charles, La., was especially hard hit, and there are reports of a chemical leak nearby.
Laura made landfall near Cameron, La., as an “extremely dangerous” Category 4 hurricane. The National Hurricane Center (NHC) said in an early-morning bulletin that Laura had weakened to a Category 3 hurricane, with rapid weakening forecast, and the storm has since been downgraded to a Category 1 as it heads northward.
“Lake Charles experienced severe wind damage but seems not to have seen the amount of storm surge that was feared,” said Triple-I non-resident scholar and Colorado State University atmospheric scientist Dr. Philip Klotzbach.
Despite the downgrade,
the hurricane still had sustained winds of more than 100 mph. Heavy rain is
predicted to be widespread across the west-central Gulf Coast, with five to 10
inches falling over a broad area, and locally up to 18 inches, leading to flash
flooding.
The storm is now tracking inland
across western Louisiana with damaging winds and is an inland flood risk as far
north and east as Arkansas and the Ohio and Tennessee valleys.
“The threat of tornadoes today and
even tomorrow also exists as the storm recurves into the Tennessee
Valley,” Klotzbach said.