Introduction

At this point, perhaps we should retire the letter “I” from use in naming hurricanes: Hurricanes Irma (2017), Ida (2021), and Ian (2022) are the three costliest U.S. storms of the past decade. The first and last of these events made landfall in southwestern Florida, and bookend five years of extreme turbulence in the Florida property market.

Florida insurers enjoyed a dozen years of mostly smooth sailing leading up to Irma; by contrast the runup to Ian was marked by rising rates, company insolvencies, and hastily assembled legislative reform packages. As insurers sat down to work through Ian claims, the Florida market faced a real chance of significant market failure and major challenges to insurance availability in the state.

Now, nine months after Ian, we can review that storm’s aftermath and the latest set of efforts to stabilize Florida’s homeowners market. While additional reforms have laid a path toward sustainability, we believe that some of the most challenging days may still be ahead for Floridians looking to secure affordable cover, with the market likely to see appreciable continued rate increases for years to come even in the absence of further hurricanes.

Florida property insurance: Market basics

Florida is—in the best of times—a challenging environment in which to write property insurance. After the wave of insolvencies from Hurricane Andrew (1992), most nationwide insurers have avoided writing homeowners business in the state, limited their capacity in Florida, or set up standalone subsidiaries to silo Florida property risk. What arose in their place was an extensive cohort of Florida-specific property carriers. These carriers feature local expertise but typically lack the risk diversification that offers long-term stability to insurance companies.

Instead, Florida property insurers are highly reliant on purchasing reinsurance, both from the state Florida Hurricane Catastrophe Fund (FHCF) and the private market. Demotech, the primary ratings agency for Florida property insurers, requires insurers to have reinsurance to cover losses from the 1-in-130-year hurricane, as well as having cover for two hurricanes each at the 1-in-100-year loss level.¹ With most Florida carriers maintaining relatively little surplus to provide protection past the end of their reinsurance tower, this has always meant that an event that was worse than these thresholds could wipe out large portions of the market in one blow.

However, both Irma and Ian are generally understood to be less severe than these doomsday scenarios.^2,3 Yet we’ve seen market turmoil anyway, with companies representing over 16% of market share filing for insolvency post-Irma, and the potential for more to come.

Clearly, there is a missing element here that was not accounted for in the models. To better understand what it was, we turn next to the courts and the Florida judicial system.

Adjusting claims in the courts: The wild development tail of Hurricane Irma

With an estimated insurance industry loss of $27 billion, Irma was clearly a major challenge to Florida insurers.⁴ However, an interesting trend can be observed in Figure 1 above. From 2017 to 2020, only one insurer with greater than a 0.5% property market share by written premium (WP) at the time of Irma (Florida Specialty) went insolvent. Since 2021, though, nine insurers have gone insolvent, including three of the 10 largest property insurers in Florida (FedNat, United Property & Casualty, and St. John’s).⁵

Typically, homeowners insurance is what actuaries call a “short-tailed” line of business, with most claims being paid and settled relatively shortly after a loss event. In theory, insurers should have had a strong grasp on their losses within a year of a hurricane, perhaps at most two. What accounts for the delayed reaction and wave of insolvencies after Irma?

Actuarially, the answer is that Florida companies saw an unprecedented upward progression of estimated Irma losses as companies were forced to revalue their claims higher and higher over time. This was driven both by claims that had been closed and were subsequently reopened at later dates, as well as new claims that were first reported years after the hurricane event. A review of anonymized insurer loss data indicates that it was not uncommon for primary insurers to witness significant development on their ultimate loss estimates from the middle of 2018 through 2022. Figure 2 shows the ultimate loss development for a representative sample of four Florida primary insurers, with the ultimate loss at June 2018 (nine months after Irma) indexed to 1.00 and other ultimate loss estimates expressed as a ratio to the June 2018 figures.⁶ In this sample even the “best” performing companies witnessed their ultimate Irma loss estimate double from June 2018 to 2022, while one company’s ultimate loss estimate increased nearly fourfold.

But the driver of increased ultimates on Irma wasn’t strictly speaking actuarial—it was legal. Three practices allowed in Florida were primarily to blame:

• Assignment of benefits (AOB): AOB provisions in insurance policies allowed insureds to transfer their insurance benefits to third parties, enabling contractors to directly bill and be reimbursed by the property insurer.
• One-way attorney fee structures mandate that the winning side does not pay attorney’s fees while the losing side does, and allow lawsuits to be filed without up-front costs.⁷
• Contingency fee multipliers allow juries to multiply the fees awarded to plaintiff’s attorneys for rare or exceptional cases that are considered difficult to win or otherwise challenging, as a way to provide extra compensation for what is presumed to be additional work.

These practices were originally set up as mechanisms for consumer protection; in theory, assignment of benefits allows contractors who have greater familiarity with insurers to negotiate fairer settlements, while one-way attorney fees and contingency fee multipliers aim to level the playing field between policyholders and the deeper pockets of insurers.

In practice, they made it incredibly easy, cheap, and profitable for contractors and attorneys to sue insurers—and a major, state-spanning hurricane proved a fruitful breeding ground for increasingly extreme behavior. Insurers accumulated more late-reported and reopened claims for Irma than for any prior natural disaster on record, as word spread across the state that even minor roof damage could profitably be argued after the fact to justify full roof replacements from insurers (and significant attorney fees for plaintiff attorneys).

The resulting statistics are almost unbelievable: In 2021, with many Irma claims still open, the state had 76% of all insurance litigation despite having only 7% of total property insurance claims.⁸ One Florida carrier (now insolvent) reported approximately 3,500 litigated claims newly reported in 2021, an exponential increase from around 200 filed in 2016.⁹

What can we learn from the mess that Irma turned into? First, it’s worth noting that, even prior to Irma, there were warning signs regarding AOB that—had they been addressed at the time—might have helped avoid the full severity of the situation. A 2018 Florida Office of Insurance Regulation (FLOIR) report relying solely on data collected prior to Irma warned that “loopholes in the way [AOBs are] being used in the marketplace are driving up costs for homeowners across the state due to unnecessary litigation associated with certain AOB claims,” and warned about insurer nonrenewal or withdrawal, dramatically increased premiums, and lower consumer choice should AOB abuse continue unabated.¹⁰

Two other factors have continued to exacerbate the litigation situation. Immediately following Irma, legal rulings for property insurance lawsuits (see Joyce v. Federated National Insurance 2017) lowered the burden for awarding contingency fee multipliers to the point where they would be applicable to nearly all cases. In addition, Florida received an influx of third-party litigation funding, which allows plaintiffs’ attorneys to more aggressively take on cases due to a consistent flow of investment to fund the case expenses. In this environment, precedents that allowed for small abuses during “business as usual” turned into five-alarm fires when fueled by the claim loads of a major hurricane.

Knock-on effects of a disaster: Reinsurance costs and repopulated citizens

The slow-developing turmoil from Irma had two major knock-on effects in the market. First, as insurers left the market either by choice or insolvency, many homeowners were forced into Florida Citizens, the state-run residual insurer. In the wake of Katrina, Rita, and Wilma in the mid-2000s, Citizens policies had increased from around 800,000 to nearly 1.5 million. Following takeouts by private companies and the relative calm witnessed from 2005 to 2017, Citizens policies decreased back to 500,000. Now Citizens policies have shot back north of 1 million in 2022, with an estimated 1.7 million expected in 2023.¹¹

In addition, reinsurers—the backbone of the market—faced growing losses in their excess towers and either left the Florida market entirely or required significant price increases to remain. Some reinsurance layers more than doubled in price from their pre-Irma levels. In a market where so many insurance dollars are immediately dedicated to purchasing reinsurance, spiraling reinsurance costs leave two options. The first is to purchase less reinsurance. However, in the Florida market, a company can only cut a limited amount from its reinsurance tower before it risks falling afoul of the minimum coverage levels required by rating agencies.¹²

The other option is to pass the cost to the consumer. Unsurprisingly, those policies that were retained in the private market faced significant premium increases post-Irma; according to data from S&P Global Market Intelligence, the average cost of Florida primary homeowners insurance increased 16% from 2021 to 2022 alone, and 45% from 2017 to 2022.¹³

In the summer of 2022, Demotech placed 17 different Florida insurers on a watch list for potential downgrades, eventually downgrading four in August. While this drew the ire of various Florida politicians, the major effect was to spotlight the fragility of a market that by many accounts seemed to be teetering on an edge.

One month later, Florida suffered what will likely end up being the costliest natural disaster in the state’s history.

Hurricane Ian basics

Hurricane Ian made landfall near Cayo Costa, Florida, on September 28, 2022, as a Category 4 storm with sustained winds of 155 mph (just 1 mph shy of Category 5 strength). Ian caused catastrophic damage to the Fort Myers and Cape Coral metropolitan areas, with Sanibel Island and Fort Myers Beach particularly hard-hit, and cut a swath of damage up the Florida peninsula through Orlando before exiting into the Atlantic Ocean near Jacksonville.

Immediately after landfall, it was clear that Ian would be an expensive event. Just how expensive was remarkably unclear at first. The major industry catastrophe modelers’ initial estimates showed a wide range (see the table in Figure 3), with estimates ranging from $28 billion to $74 billion.

These estimates reflect a greater than usual uncertainty for a hurricane event. While some of this uncertainty can be attributed to nationwide economic factors such as heightened inflation, the major differences across estimates can mostly be attributed to uncertainty regarding just how bad the Florida litigation environment would be for Ian claims.

At any point in these ranges, however, Ian will be a market-shaping event. In fact, we have already begun to see its immediate impacts, as reflected through the recent work of Florida’s government.

Governments, regulation, and responses

Many of Florida’s problems are intractable—the state can’t move away from its hurricane exposure or prevent a changing climate all by itself—but by 2022 the situation had deteriorated enough that the litigation climate in the state had become a major public policy issue.

In late May 2022, during the reinsurance renewal season, the Florida legislature convened a special session that produced two signed pieces of legislation, Senate Bills 2D and 4D. These bills, particularly 2D, took moderate steps toward claim reform, including prohibiting contractor solicitation for AOBs, increasing the threshold for contingency fee multipliers, requiring disclosures for contractor solicitation for roof damages, and allowing insurers to include separate roof deductibles.¹⁸ Among other things, the bills also authorized an additional temporary reinsurance fund (the Reinsurance Assistance to Policyholders Program, or RAP) to supplement the existing FHCF reinsurance structure.

It took Ian’s landfall, however, to produce more appreciable reforms. A second special session was convened in December 2022 and resulted in Senate Bill 2A, which passed on December 14. The bill included the following significant changes:¹⁹

• Repealing one-way attorney fee provisions for property insurance
• Prohibiting AOB of any residential or commercial property insurance policy as of January 1, 2023
• Establishing a second temporary reinsurance program, Florida Optional Reinsurance Assistance (FORA) for 2023
• Increasing the threshold for bad faith lawsuits
• Reducing claim reporting deadline for policyholders from two years to one for new or reopened claims and from three years to one and a half years for any supplemental claims
• Requiring homeowners to be on a private insurer if they get a quote within 20% of the Citizens quoted price

Many of these provisions directly address the problems that have been identified in the Florida property market. FORA provides much needed cover at the lowest layers of companies’ reinsurance towers.²⁰ The repeal of AOB and one-way attorney fee provisions combined with limitations on lawsuits should make it more difficult to sue insurers and thus reduce loss adjustment expenses (LAE) for insurers and claim severities. The reduction of claim reporting deadline is intended both to tamp down on frivolous lawsuits and to clarify loss exposures for insurers and reinsurers alike.²¹

What is not included in Senate Bill 2A is equally as notable. The law does not apply retroactively. Claims from Irma (not to mention a host of other smaller storm impacts in the last five years) remain open, with benefits already assigned to contractors and lawsuits filed by law firms. AOBs are still allowed for policies issued in 2022—the very policies that were in force at the time of Hurricane Ian.

In addition, the contractors and law firms that relied on AOB lawsuits as a significant source of income are likely not to give up that stream without a fight. There is no mention of third-party litigation funding reform, another significant contributor to incurred legal expense costs in Florida and elsewhere.²² FORA is funded for a single year only (and RAP for two), raising the risk that their future (and thus reinsurance capacity in Florida) will be contingent on their success during the upcoming 2023 and 2024 hurricane seasons.

What comes next?

Despite the efforts of its government, the Florida property market is still in trouble. How much trouble will depend on the impacts of the next couple of hurricane seasons. Despite a recent reinsurance renewal that was described in the press as “orderly,” we believe that another major event would place significant solvency pressures on many remaining private market companies.²³ Even in the absence of a major event, there are two questions to watch closely throughout the remainder of the year.

1. Will Ian development turn into another Irma?

To date, there are several reasons to believe that Ian will not in fact prove to have the highly litigious claim tail that upended the market following Irma.

• First, Ian development will benefit moderately from some of the post-Irma legislative improvements (such as reducing the claims reporting window from three years to two years post-event), although as noted above we do not expect to see benefits from the December 2022 reforms.
• Second, Ian’s footprint of high winds, while extreme near landfall, was generally less broad than Irma’s and impacted fewer households in general. For the sake of monitoring loss development, full-limits home losses are easier to reserve for, and we expect to see relatively fewer minor Ian claims that turn into protracted legal battles when compared to Irma.
• Finally, one of the side effects of the Irma claim situation is actually beneficial for Ian—with so many households in Florida receiving full roof replacements after the 2017 storm, the housing stock as a whole should be better prepared for Ian as a subsequent event.

On the other hand, when tracking Ian incurred development to date, we have seen the event significantly outpacing Irma to date. According to FLOIR, Irma had approximately $5.9 billion of reported Florida loss two months following the storm. Ian surpassed that figure at two months, and has $16.7 billion of reported Florida loss as of June 2023. As Figure 4 indicates, Ian has lived up to the initial high estimates and is well outpacing Irma loss amounts at similar maturities.²⁴

In addition, it is worth keeping a close eye on Florida companies’ Ian reserve estimates following the June 1 reinsurance renewals. For companies that may be under some amount of solvency pressure, getting past the key renewal date may allow for more of an opportunity later this summer to take a deep assessment of Ian claims. Irma development accelerated greatly in the period from nine months after the event to 24 months after the event. If Ian trends down even a moderately similar path, we will likely begin to see it emerge more clearly in the upcoming months.

2. How long will it take for the market to reach rate adequacy?

Due to its position as a government-owned entity and the only available insurer for many Florida properties vulnerable to hurricane risk, Citizens rates garner a significant amount of public attention. The company has pushed for rate increases in recent years because it considers itself to still be well short of actuarial rate adequacy. According to Citizens, rates as of June 2023 remain 58.6% below what they consider to be actuarially sound even with last year’s legal reforms. This is better than the alternative—Citizens similarly indicated that the rates would have been 88.3% below actuarially sound rates without the reforms—but it is still woefully inadequate.²⁵

It is one of the basic principles of insurance that a market cannot get away in the long run with charging less than the actuarially indicated amount of premium for the risk. Currently, the Florida government is implicitly subsidizing policyholders through both actuarially inadequate Citizens rates, in addition to providing favorable reinsurance backstops through two temporary facilities.

It is possible for governments to subsidize private insurance markets for extended lengths of time (the National Flood Insurance Program provides a clear example of this). However, the Florida statehouse is most likely hoping that the private market will step up to relieve the government of these obligations, similar to the cycle of rebirth the private market experienced after the 1992 and 2004-2005 hurricane seasons.

At this point, it is unclear to us whether this path exists in the short term. Florida statute limits the annual rate increases Citizens can pass to its policyholders (the cap for 2023 is 12% and for 2024 13%), making it a multiple-year proposition to dig out of the actuarial deficit it currently faces.²⁶ Consequently, it will remain more difficult for the private market to compete against the rates it offers. It is also currently unclear where private appetite will come from, as there have been few notable post-Ian primary company formations or capital raises to date, and the process of beginning a new insurance entity takes an appreciable amount of time.

To its credit, the legislature has already passed bills that begin to address many of the state’s most fundamental issues. But the road back to sustainability is a long one, and likely will be paved by several years of significant rate hikes for Florida property owners. For now, Florida will continue to furiously triage its fragile insurance market, and hope that both the seas and the courts will provide calmer passage in the years ahead.

---

¹ Warren, R.M. (Fall 2021). Horizontal Catastrophe Reinsurance Coverage. The Demotech Difference, p. 10. Retrieved July 10, 2023, from https://www.demotech.com/pdfs/demotech_difference/flexpaper/2021_Fall/docs/Fall-2021-Magazine.pdf.

² While return period estimates for individual storms are imprecise and dependent on location, Irma would likely have been comfortably within Demotech limits (indeed, one estimate put Irma at a 1-in 15- to 20-year event), while Ian return period estimates may be close to Demotech requirements for the most heavily impacted areas and comfortably below them in other areas.

³ See the Irma estimate at https://ams.confex.com/ams/33HURRICANE/webprogram/Paper340515.html; see the Ian estimate for different areas at https://www.ajg.com/gallagherre/-/media/files/gallagher/gallagherre/assessment-of-damage-caused-by-hurricane-ian.pdf.

⁴ Following Irma, Hurricane Michael (2018, $12 billion of insurance industry loss) also impacted the state in the Panhandle region; however, its impacts were much less severe and generally will not be the focus of this paper.

⁵ See the list of liquidated companies at https://www.myfloridacfo.com/division/receiver/companies. The total above includes Prepared Insurance, a subsidiary of Lighthouse that was formally acquired in 2020 before Lighthouse went insolvent in Louisiana in August 2022. Market share information comes from publicly available data at https://floir.com/tools-and-data/residential-market-share-reports. Note that United Property & Casualty Insurance classified its data as “trade secret” in the first quarter of 2017 and is thus not publicly available, but the company’s market share by WP from Q4 2016 would be in the top 10 as of Q3 2017.

⁶ Based on proprietary Irma loss data from four different Florida primary insurers.

⁷ Further, in the state a legal “victory” for the plaintiff was considered to be any increase in the payment from the initial insurer estimate, even if just a nominal amount (e.g., one cent higher).

⁸ Schorsch, P. (December 14, 2022). Insurance industry group applauds passage of one-way attorney fee elimination bill. Florida Politics. Retrieved July 10, 2023, from https://floridapolitics.com/archives/576381-insurance-industry-group-applauds-passage-of-one-way-attorney-fee-elimination-bill/.

⁹ As per claims study conducted on major top-10 Florida insurers post-Irma.

¹⁰ FLOIR (January 8, 2018). Report of the 2017 Assignment of Benefits Data Call. Retrieved July 10, 2023, from https://floir.com/docs-sf/default-source/property-and-casualty/assignmentbenefitsdatacallreport02082017.pdf?sfvrsn=b7b644fd_2.

¹¹ For a chart of historical Citizens figures, see https://www.citizensfla.com/documents/20702/93064/20221231+Business+Overview.pdf, page 2. The 1.7 million estimate comes from https://www.abcactionnews.com/news/price-of-paradise/floridas-citizens-property-insurance-predicts-to-hit-record-with-nearly-2-million-policies-in-2023.

¹² In turn, any company that fails to achieve a minimum rating level would not be able to write homeowners for mortgages requiring adequately rated insurers, effectively cutting them off from most of the market.

¹³ Using aggregate statistics for Florida insurers active in 2021.

¹⁴ Verisk (October 3, 2022). Verisk Estimates Industry Insured Losses to Onshore Property for Hurricane Ian Will Range from USD 42 Billion to USD 57 Billion. Press release. Retrieved July 10, 2023, from https://www.verisk.com/newsroom/verisk-estimates-industry-insured-losses-for-hurricane-ian/.

¹⁵ Moody’s/RMS (October 7, 2022). RMS Estimates US$67 Billion in Insured Losses From Hurricane Ian. Retrieved July 10, 2023, from https://www.rms.com/newsroom/press-releases/press-detail/2022-10-07/rms-estimates-us67-billion-in-insured-losses-from-hurricane-ian.

¹⁶ Wilkinson, C. (September 30, 2022). Karen Clark pegs Hurricane Ian losses at $63 billion. Business Insurance. Retrieved July 10, 2023, from https://www.businessinsurance.com/article/20220930/NEWS06/912352828/Karen-Clark-pegs-Hurricane-Ian-losses-at-$63-billion.

¹⁷ CoreLogic (September 29, 2022). CoreLogic Estimates $28 Billion to $47 Billion in Insured Losses from Hurricane Ian Wind and Storm Surge in Florida. Retrieved July 10, 2023, from https://www.corelogic.com/intelligence/corelogic-estimates-28-billion-to-47-billion-in-insured-losses-from-hurricane-ian-wind-and-storm-surge-in-florida/.

¹⁸ The full text of 2D is available at https://www.flsenate.gov/Committees/BillSummaries/2022D/html/2874. The 4D bill is more focused on building codes and safety inspections.

¹⁹ A full list is available at https://www.myfloridacfo.com/division/ica/2022propertyinsurancechanges. House Bill 837 then brought these changes to all insurance lines.

²⁰ Florida Association of Insurance Agents. Resource Categories: Florida Optional Reinsurance Assistance Program. Retrieved July 10, 2023, from https://www.faia.com/resources-(1)/legislative-summaries/legislative-summaries/2022a-special-session-legislative-summary/property-insurance/property-insurance-florida-optional-reinsurance-as.

²¹ See https://www.capitaliq.spglobal.com/web/client?auth=inherit#news/article?id=75214801 (sign-in required); up from 1.15 million in-force policies at year-end 2022 to 1.25 million at the end of Q1 2023.

²² As of the publication date, two Florida bills regulating third-party litigation funding have been introduced, but have not become law. See https://www.verisk.com/insurance/visualize/florida-and-other-states-take-aim-at-regulating-third-party-litigation-funding/.

²³ For example, see https://www.artemis.bm/news/florida-renewals-june-1-and-all-but-done/: “Orderly remains the key word in any commentary on the Florida reinsurance renewals in 2023, with the market clearing far more efficiently.”

²⁴ From historical FLOIR reporting at https://www.floir.com/office/hurricaneseason/hurricaneirmaclaimsdata.aspx and https://www.floir.com/home/ian. The figures for both events exclude LAE, incurred but not reported (IBNR), and non-Florida losses associated with the storms. Note that FLOIR’s reporting for these events is not done on consistent intervals. The weeks and months figures are those closest to that date in the historical progression. Totals for months without FLOIR updates are linearly interpolated from months with reported loss data.

²⁵ Citizens (June 8, 2023). Citizens Presents 2023 Rate Recommendations to OIR. Press release. Retrieved July 10, 2023, from https://www.citizensfla.com/-/20230608-citizens-presents-2023-rate-recommendations-to-oir.

²⁶ Ibid.