Will the FHCF survive 2023?

Informed Group CEO and industry commentator Ian Gutterman re-examines the perilous financial position of the Florida Hurricane Catastrophe Fund (FHCF)…

Last year, I wrote after Hurricane Ian that the FHCF may collapse due to the size of the loss. While the fund has not gone bust yet, it is perilously close.

What follows is an update on the FHCF’s financial condition, its ability to handle further losses and a projection of its future cash flows. The story is not pretty.

Punchline: The FHCF needs multiple loss-free years to survive. If you are an insurer paying premium in, you should ask yourself how you feel about providing bailout funds for cover that may not exist.

FHCF 2023 claims paying capacity

By statute, the FHCF is supposed to be able to pay $17bn in claims. Newsflash: it can’t!

It only has $3.7bn of surplus, down from $12.7bn last year (and actually that’s $3.7bn projected at YE23). It also has $3.5bn of previously issued bonds which gives it $7.2bn of “capacity” (though no way to pay back the bonds!).

This leaves it $9.8bn short! The FHCF assumes it can raise $8.6bn in new debt after a loss. Much more on that later, but even if so, it is still $1.2bn short. In other words, it can’t pay $17bn in coverage and may pay much less than that if bond markets don’t cooperate.

How do insurance companies buying FHCF cover feel about that? They appear to be nonplussed. They continued to pay normal premiums to the FHCF ($1.5bn statewide) without any rebate for the potential shortfall.

What happens if the FHCF can only raise another $3.5bn of debt? A full loss would leave insurers getting less than two-thirds of what they expected. That may be the difference between them remaining solvent or not!

Has any insurer disclosed what their “surprise retention” could be if there is a large hurricane and the FHCF can’t fully pay? Has Demotech incorporated this risk into their ratings?

Nobody seems to want to discuss it. The approach across the market seems to be wilful ignorance.

Loss modelling

So what type of loss would it take to cause a crisis? Good question. Thankfully, the FHCF has provided us with some answers.

To wipe out the $3.7bn of surplus (so $12.8bn loss including industry retention) takes a $13.3bn personal lines loss, which is around a $20bn industry loss. In this case, the FHCF only has bond funding available and would likely be declared insolvent (more details later).

Another Irma would also wipe out the $3.5bn of existing bonds as well. A repeat of Ian would leave the fund needing to raise $2.8bn to pay claims with nothing left to continue on next year, so it would need to raise $20bn to be a going concern.

To say the situation is precarious is an understatement.

The no-storm solution

But what if there are no hurricanes? Wouldn’t the FHCF rebuild its surplus? Yes, but slowly.

Premium is $1.5bn a year. The fund has historically operated with surplus above $10bn so you would need five loss free years to get there (after accounting for operating costs and interest). You would also have to refinance the $3.5bn of debt.

Could Florida go five years without a loss? Certainly, but this is now necessary. Otherwise, the FHCF is reliant on the debt market.

I should also mention all this math presumes the $10bn estimate for Ian holds. Given most Florida storms develop adversely, this is not an insignificant risk.

Bonds to the rescue?

One of the FHCF’s grave mistakes is assuming that if it has losses greater than its current $7.2bn of funds it can bridge the difference by raising debt.

This makes two huge assumptions:

This is more optimistic than assuming there won’t be any hurricanes for the next five years!

When the FHCF did its $3.5bn offering in 2020, it had $11bn of surplus. That makes it a lot easier to borrow because the risk of attachment was much lower.

When you only have $3.7bn of surplus (and have upcoming maturities on the $3.5bn outstanding debt), the odds of the bonds experiencing losses are much higher. Thus, the coupons should be much higher and the rating lower. This makes it harder to borrow as much.

Yet, the FHCF has indicated no change in its potential borrowing capacity since last year (it’s actually up slightly from $8.0bn to $8.6bn!). It also suggests it can still borrow at the same rate (5 percent) even with increased risk and increased rates across the economy.

Unless it has kidnapped bond buyers and forced them to buy bonds in return for their freedom, this seems nigh impossible.

The “process” the FHCF uses to make these fundraising estimates is to ask five large banks what they think they can raise and take the average.

You can imagine JP Morgan isn’t putting its top bankers on this project. They seem to take the same estimates as the year before and roll them forward with small changes. Not a lot of critical thought here and thus the accuracy of the estimates is likely very low.

What will be the bank’s excuse when they’re wrong? “We couldn’t foresee the change in market conditions.”

The bond math

So how much debt can the FHCF raise? It’s hard to say precisely, but let’s go through some of the math to figure out what might make sense.

First, as noted, the FHCF collects $1.5bn in premium annually. It also has the ability to charge assessments. However, that is politically unpalatable and it didn’t even add any assessments for this year.

The assessable base is >$70bn and it can in theory go as high as 6 percent but that would never fly in Tallahassee. I assume a 1 percent assessment is more realistic which would add $700mn a year.

Thus, in total, inflows would be $2.2bn. What about outflows?

First, there are claims. Let’s be optimistic and assume no new losses for the moment.

Next, there is interest. This will depend on how much money is raised. The FHCF estimates it can raise at 5 percent, though I suspect it will be far higher. We’ll return to this.

Finally, there are debt maturities. Bondholders eventually expect to be paid back. This is where the big problem lies.

Scenario #1: $10bn at 5.3%, no losses

Let’s assume it raises $10bn at (generously) 5.3 percent for seven years. How does the claims-paying ability change if there are no losses?

We start with $7.2bn of surplus plus the old bonds. Adding the $10bn takes it to $17.2bn. Add on the next seven years of premium and there is $27.7bn gross. This sounds fantastic!

But we have to subtract things too. Interest would go to $600mn a year. That takes off $4.2bn. And, of course, debt comes due – $13.5bn of total debt would need to be paid leaving the FHCF with $10bn of surplus and no debt.

This is a good outcome, but it is reliant on no losses for seven years! And even then, the FHCF still would have to raise additional debt to get from $10bn to $17bn.

Scenario 2: $5bn at 10.5%, $20bn storm

However, let’s change a few assumptions. One, there is a $20bn hurricane this year which wipes out the $3.7bn in surplus. Two, the bond raise is at a more realistic 10.5 percent given the greater risk and, three, it is only able to place $5bn.

In this case, I assume the 1 percent assessment, so there are $15.4bn of inflows over the next seven years.

However, there is no surplus left, just debt – the $5bn new and the existing $3.5bn. This creates a few problems.

First, the fund has only half its mandated capacity for 2024 so insurers can’t get fully paid. How would they feel about paying full premium for half the coverage? Not so good, I’d imagine.

Next, even if the following six years have no storms, they still don’t get all the way back to $17bn before the debt has to be paid off. If there is no way to get back to $17bn, why should insurers keep paying premium?

What other rosy assumptions could you make? You can argue for higher assessments. You can assume it can raise more debt in future years, but, remember, you’re also assuming no future storms!

The math just doesn’t work.

Scenario #3: $40bn storm (Hurricane Ian repeat), debt markets shut

I’m not even going to do this one! You can figure it out by now. That would be a $10bn loss to the fund leaving it with -$6.3bn of surplus and no way to pay off the $3.5bn of current debt. It would be impossible to raise debt in that scenario.

It’d have to turn the lights out immediately and insurers would get paid pennies on the dollar for their claims.

Conclusion: Pray!

What’s the lesson from these scenarios? The only way insurers might get paid if there’s a storm this year is if it’s a small one. There is no protection for larger events! So why are insurers still buying the coverage?

Willing Ponzi participants

The FHCF has become a Ponzi scheme. Its only way to survive is to take premium from insurers and pay nothing out!

Ironically, the insurers willingly engage in the deception because if they don’t pretend the FHCF will pay them, then they can’t afford reinsurance and they will have to shut their doors.

Everyone is bluffing each other! And they are all holding losing hands. Those who will pay the price are ordinary Floridians who don’t get fully paid on their claims because they end up in the state guaranty fund after the next hurricane.

Insurers are supposed to have enough capital to pay their potential claims before they take on (or renew) business. Any insurer in the state of Florida that is dependent on the solvency of the FHCF to remain solvent themselves isn’t honouring that promise.

If the only way for the insurer to win is if the wind doesn’t blow, then it’s heads the insurer wins, tails the policyholder loses. That isn’t how insurance is supposed to work.

So it’s nice that the state passed some reforms that might reduce future severity after a storm, but you know what they didn’t address? The elephant in the room…

Local insurers won’t have the reinsurance proceeds to pay claims after a loss which means this hurricane season is Florida roulette.

Ian is an insurance industry investor and the founder of Informed Group, a start-up homeowners insurer seeking to change the way people buy home insurance. His regular Nominal Returns blog can be found here.