Though Florida hasn’t suffered a category 5 storm since Hurricane Andrew hit the north Miami area on Aug. 24 1992, the risk of such a storm has not gone away or even declined. In fact, the state’s risk of destruction may be even greater today.
Not only has Florida’s population grown 35 percent since Andrew, but also building, especially along the coast, has been booming since the 1990s.
According to a Swiss Re report “Hurricane Andrew: The 20 miles that saved Miami,” the total insured losses of an Andrew-type storm with the exact intensity and track would be significantly higher than the $15.5 billion loss in 1992 – totaling about $50 to $60 billion, due to the combination of increased development and asset values.
If a major cat 5 storm makes landfall in downtown Miami, 20 miles north of where Andrew hit, losses to the insurance industry could range from $60 billion to $180 billion, according to Swiss Re models.
Catastrophe modeling firm AIR Worldwide estimated that if an Andrew-like hurricane were to strike just south of Miami and just 8 miles north of Andrew’s landfall in the city of Homestead, total insured losses would exceed $138 billion, with losses in Florida exceeding $127 billion.
While these numbers are merely speculative, two of the authors of the Swiss Re report say the worst-case scenario should be kept in mind, especially since hurricane risk in Florida has not decreased in the last 25 years.
“It’s more important than ever before to assess present day hurricane risk and to really think deeply about if current insurance purchasing decisions are sufficient for if there was another Andrew,” said Megan Linkin, Swiss Re Natural Hazards expert.
Many in the industry also fear that history could repeat itself when it comes to the industry’s ability to survive a catastrophic event like Andrew.
Hurricane Andrew was the costliest disaster in U.S. history, according to the Insurance Information Institute (I.I.I.), and the industry was not prepared to pay the large number of claims. Andrew was responsible for the failure of at least 16 insurers between 1992 and 1993.
Before Andrew, insurers writing in Florida were not looking closely at their exposure in hurricane prone regions. It was the first cat 5 hurricane in Florida since the Labor Day Hurricane of 1935, and only the third cat 5 storm to make landfall in the U.S., giving everyone who lived there and insurance companies a false sense of security.
“I can’t tell you how many companies told me they were driving street by street to count how many homes they insured because they didn’t know – it was a free-for-all,” said Robert Reynolds, owner of the Morris & Reynolds Agency in Miami.
His agency office was completely destroyed by Andrew and as many as 2,000 of his clients were affected.
Industry experts have expressed concern that Florida’s young crop of domestic insurers would not be able to survive another Andrew. Just last week, one of them, Sawgrass Mutual, was placed under administrative supervision by the Florida Office of Insurance Regulation and will begin winding down its operation, which includes 20,000 homeowners policies.
That announcement during peak hurricane season fuels concerns about whether Florida insurers are adequately assessing their risk.
“The industry has come a long way and is doing a much better job than they were doing before Andrew, but some companies still ignore other facts around them,” said Karen Clark, founder and CEO of Karen Clark & Co.
Clark says catastrophe models, which became widely used after Andrew, are a reliable way of assessing an insurer’s exposure, but the probable maximum loss (PML) models insurers rely on don’t always tell the whole story.
“Models can’t prevent events from happening, but they are supposed to prepare companies for what could happen. Unfortunately, people are using models like they are this magic box and putting out this magic PML number and all they have to do is manage [exposures] to the PML. That’s not correct,” she said.
Clark says the 1-in-100-year PML model can underestimate an insurer’s exposure because it does not tell the whole story of a company’s potential losses. She said one big piece that is missing from this metric is a company’s exposure concentration. She worries about some of the newer Florida domestic companies in particular.
“The insurance industry has gotten into this mode where as long as they are managing to their PML the world is fine and they don’t have to do anything else…small domestics form who don’t have a lot of capital but they are buying reinsurance up to this PML,” she said.
KCC developed the Characteristic Event methodology back in 2012 that provides insurers with information on what size losses different return period events could cause, and where losses from 100-year events could exceed an insurer’s 100-year PML.
“Traditional cat modelers are very stuck in the original model paradigm of ‘this is what we provide’ versus thinking about what additional types of information might be useful to model users,” Clark said. “The more info the models can give the better.”
Clark says carriers should look at more than just a PML model in assessing their exposure, but it may take “some disruption to get widespread industry adoption.”
Joe Petrelli, president of Demotech, which rates most of the Florida domestic insurers, says it reviewed the characteristic event model from KCC, as well as conditional probable maximum losses calculated by JLT Re and peak loss scenarios calculated by Aon, among others, prior to assigning its 2017 ratings.
“Keep in mind that a PML of one in 100 years attempts to measure a 1 percent probability of exceedance over the portfolio of an insurer,” Petrelli said. “In contrast, peak loss scenarios, characteristic events, conditional PMLs and other measures focus on the reality that the probability of exceedance is not uniform across the portfolio of an insurer.”
In rating its companies, Demotech follows at least 19 criteria in reviewing a company’s reinsurance backing, including identifying and confirming each company’s utilized model currently approved by the Florida Commission on Hurricane Loss Projection Methodology.
Petrelli added that Florida OIR performs catastrophic stress test analysis annually on all Florida insurers, utilizing the final reinsurance program purchased by the carrier as the baseline.
“In each year that the OIR has performed its independent analysis, carriers reviewed and rated by Demotech have passed the catastrophic simulations utilized by OIR,” he said.
This said, Petrelli said that Demotech believes that the raw dollar damages associated with such a catastrophic event should be viewed in conjunction with the likelihood of the selected event.
“For example, in 2017, JLT Re indicated that the annual probability of a Category 5 event hitting Miami-Dade area was 0.2 percent; two chances in 1,000 that it could happen and 998 chances in a thousand that it would not happen,” he said. “Putting this annual probability into human terms, an 82-year-old person is more likely to live to be 83 years old than a category 5 event is likely to strike Miami-Dade.”
Marla Schwartz, an atmospheric perils specialist for Swiss Re, said insurers should stick to the one common theme when using models – no model is perfect but they are still a very useful and important tool to the industry.
“It is hard to say if they are sophisticated enough or what they are missing,” she said. “That’s where risk assessment and underwriting experience comes into play and the human component.”
Roger L. Desjadon, chief executive officer of Florida Peninsula Insurance and chairman of the Florida Property Casualty Association, said rating agencies and regulators are “very aware of what the industry needs to do and how it needs to do it,” to remain financially stable.
He added that with the experience of past storms, those who regulate and rate the industry understand the importance of securing an environment where companies are at their best.
“They work very hard to make sure the market is what people expect it to be,” he said. “And they give companies opportunities to make whatever adoptions they need to make in order to perform in stressful situations.”
A.M. Best, which also rates many Florida insurers, said in a recent A.M. Best TV spot on Hurricane Andrew that its insurers with A.M. Best ratings “possess the financial strength to honor policyholders’ claims,” in the event of another catastrophe.
Still, Reynolds said with what he witnessed after Andrew 25 years ago, the worry about insurer solvency is always in the back of his mind.
“Using what was then ‘state of the art technology’ many in our industry thought their predictions would protect their firms and clients but unfortunately in many cases we learned that they got it wrong and it was a real awakening,” he said. “We can’t have that happen today, the values and risk is exponentially larger now.”
Reynolds said he hopes he doesn’t see a storm like Andrew happen again in his lifetime, “but to me it is a real concern – is the industry prepared? Insurers and reinsurers say they are and we have to trust that they know what they are doing, but with loss projections in the hundreds of millions the industry can’t afford to get it wrong.”
Swiss Re’s Linkin says a big hurricane hitting the Miami area is only a matter of if, not when.
“It has happened before, and one will hit Miami again,” she said. “It’s imperative that anyone who does business or lives in the area is financially and physically prepared for major hurricanes to strike.”