Back with another edition of “what happens in Florida after the FHCF fails”! (Last year’s version here.)
The good news is there were no losses last year, so the FHCF rebuilt some of the surplus lost in recent years.
The bad news is they still don’t have enough money to pay their potential claims this year and are, once again, assuming they can raise state debt after a loss to pay claims (but once more, with no explanation of how they will pay all the debt back = buyer beware!).
In simple terms, the FHCF is insolvent. It does not have the assets to pay its guaranteed obligations.
It only continues to exist because a) the state lets it and b) primary insurers need to believe it will survive or they will collapse themselves.
It is not unlike the Lloyd’s Central Fund, except without the supervision of Lloyd’s to make sure underwriters aren’t taking unacceptable risk.
Additionally, Lloyd’s syndicates are at least somewhat diversified (there is a lot of overlap, but not 100%) while the FHCF is a central fund for insurers all writing the exact same risk. It’s ludicrous!
So what will take the FHCF from an illusory appearance of stability to an ugly end? That’s easy. One more Irma or Ian sized storm.
A quick heads up before I get into the details. This is part one of two. I split it up due to length.
Today, I’ll provide an update on what has changed since last year and where the risks lie. Next week, I’ll do a numerical breakdown showing how FHCF’s assumptions don’t add up.
FHCF’s Capital Structure
Let me do a quick reminder on how the FHCF works. It provides $17B of reinsurance to home insurers in the Florida market. This is above a $10B industry retention.
The problem is it doesn’t have $17B to pay claims. It only has $10B.
What happens if it incurs $17B in losses? It waves a magic wand conjures up $7B of hypothetical bond buyers to bail it out.
Nobody knows if these bond buyers exist (more on that later). Even if it can raise the $7B, it is uncertain whether it can pay the lenders back (no, assessments alone won’t do the trick).
Would you buy collateralized retro from a market that said “pay us premium for $170M of limit, but we’ve only got the funds for $100M of limit right now, but trust us, we can get the rest later”?
No, of course not. But every Florida insurer blindly pays their premium to FHCF and assumes they’ll make good if there is a claim.
Willfully Blind
So why do insurers go along with this charade? Desperation.
The FHCF offers “cheap” reinsurance. Without it, many insurers couldn’t afford to buy enough reinsurance to be viable.
The FHCF is all that stands between survival and runoff. Thus, they must pretend it will survive long enough to pay them. There are no better options than lying to one’s self.
The dilemma is, without the FHCF, there would be a full blown capacity crisis that could only be fixed by substantial further premium increases to homeowners.
But by sustaining the FHCF, Florida creates a bigger ultimate loss for its taxpayers (the eventual bailout) and still ends up in the same place – a full blown capacity crisis.
Modeled Loss Events
So what does it take to end the FHCF as we know it? Another Hurricane Irma or Ian, which is between a 1 in 15 to 1 in 20 industry event (though they both happened within 5 years!).
How do I know this? Because the FHCF kindly volunteers this information.
A 1 in 15 event wipes out all their surplus (a roughly $30B industry event) and a repeat of Ian nearly wipes out all their bonds too. (Note, Irma cost FHCF $7.8B so is in between these two points.)
While FHCF will argue they can continue on after a loss like this, I will show in part II that this is unrealistic.
They also show in a separate document that the gross AAL for home insurance claims is $4.7B and this results in a $1.1B expected annual claim to FHCF.
Some interesting notes on that. First, you may ask, I thought the industry retention is $10B. How can there be $1B of losses from a $5B event?
Because each company has its own retention. The $10B applies if a storm hit the entire state at once.
But since some carriers are overweight certain regions, their retention may be reached at a storm below the industry retention (and conversely, some carriers will get paid nothing for storms that go over $10B).
What is interesting about this is FHCF collects $1.5B in premium, but tells you it’s expected loss is $1.1B. When you add on interest and other expenses, you will see this leaves little free cash flow for debt maturities.
Also, it is quite laughable to suggest AAL is $1.1B when they incurred $18B in claims in just five years! If you go back 20 years, that grows to $27B.
When you update the older storms for inflation, it would be at least $36B. And that’s without any mega storms like an Andrew to drive up the total.
So it seems fair to say the AAL is at least twice what FHCF reports which means they are not charging nearly enough premium to break even. Part II will demonstrate the consequences of this.
Raising Debt
So what is FHCF’s answer to the seeming inability to pay expected claims? Don’t worry, they say. We’ll just borrow the money!
This sounds great except for a whole host of reasons.
– they need to issue more debt than the market can bear.
– they assume it can be done at a normal, rather than distressed, interest rate.
– they don’t ever explain how they can pay it back when it matures.
– oh, and they implicitly assume there won’t be another storm for a long, long time.
Other than that, it’s a perfectly coherent plan!
Let’s address a few of these issues. So FHCF’s plan for paying claims when they are out of money is to raise new debt.
That’s right, they think they can go to borrowers and say “we’re flat broke, we don’t know how we can pay you back, but lend to us anyway.”
To perpetuate this charade, every year the FHCF pays a bunch of banks for their guesstimate of how much debt they can issue after a large loss that exhausted their capital.
And every year, heroically, the banks come back and say FHCF can raise well over $10B at reasonable rates.
This, of course, is absurd. I don’t know if FHCF believes this or not, but surely the banks have all their excuses ready for when the day comes about “market conditions have changed, blah blah blah”.
The biggest fail in these estimates is the banks don’t account for the cost of distress. They assume FHCF can raise debt at the same interest rate after a giant loss as before one. This ignores reality.
Bond buyers, even municipal ones, aren’t complete idiots. FHCF is not explicitly backed by the state of FL (though bond buyers are betting it implicitly is). It can fail.
Thus, when the risk of failure is higher, they will not buy its bonds unless they are paid a lot more in interest.
This higher coupon requires FHCF to charge a higher assessment, which means less of the assessment is available to pay off the debt when it matures. As I will show in part II, that limits the ability of assessments to keep FHCF solvent.
Beyond that, the FHCF never addresses a key issue anywhere in its report. While it assumes it can borrow as needed to pay any immediate claims owed, it never explains how it can pay off the debt when it matures, nor how it will borrow more if (when!) there are future cat losses.
These are pretty essential risk management considerations that get swept under the rug. There is an implicit assumption assessments can solve everything, but they never prove the numbers on that.
And for good reason. Because the math doesn’t work.
The Failed Refinancing
But before I get into the heavy math next week, let me finish Part I with a pretty important data point that surprisingly didn’t get reported in any of the trade press.
The FHCF has $1.25B of bonds maturing next year. They, wisely, tried to pre-fund that maturity earlier this year, before hurricane season.
They sought to raise $1.5B to give themselves some cushion. Guess what happened?
The bond markets balked. They could only raise $1B. Oops!
And this was just a refinancing. Imagine what happens if this was after they wiped out their remaining surplus from this August’s Hurricane Isaac (they have bad luck with the I storms)!
Yet, they assume they can raise $8B before the following June and $14B within two years. How???
Even if they can raise the money, and even if they announce large assessments after, they will still be in the same financial position next June – one storm away from being out of money, as I’ll show in part II.
Raising debt after a giant loss only buys time. It doesn’t solve the problem.
And if, instead, there are no bond buyers post the next storm? The FHCF doesn’t address this, but I will.
Insurers won’t get paid and there will be no FHCF coverage offered next year.