Liquidation Event?

Right so on Friday, we brought you “Are Hurricanes Good Now?,” a piece in which we asked whether it makes sense to believe that over the longer-term, hurricanes are actually a boon to economic growth.

Intuitively, that both does and doesn’t make sense. That is, it makes sense to the extent that someone has to clean everything up and it makes no sense to the extent that it exemplifies the broken window fallacy.

But I’m not going to launch into a diatribe on economic theory here. Suffice to say the consensus is that while there will invariably be a drag on near-term growth, that will be recouped in the months after the initial disaster. Bill Dudley was out parroting that on Friday.

Well, needless to say folks are worried about the insurance industry. Or at least they were until Friday when it the rebounded – and bigly.

The S&P 1500 Insurance Index had its best day since early last year, bouncing from the recent doldrums.

“A weaker Irma and widespread oversold conditions (some stocks had the lowest 14-day RSI in years) sparked a round of buying,” Bloomberg notes.

“Property and casualty insurers may face about $50 billion in losses from Hurricane Irma after the storm system appeared to change direction, down from previous expectations of as much as $125 billion if the eye of the hurricane passed over Miami,” Wells Fargo analyst Elyse Greenspan said in a note on Friday.

Whatever, right? There’s no way to assess this with any degree of confidence ahead of time.

Some folks have raised the question of what it would mean for yields if insurers or reinsurers are forced to sell assets to pay claims. This is a potentially important question although again, it’s more thought experiment than anything else at this juncture.

Well, that issue comes up in Deutsche Bank’s latest Fixed Income weekly piece in a subsection called “Should we talk about the weather?”

First, the bank takes a look at how yields, the curve, and stocks performed in the lead up to and the aftermath of landfall for significant storms. To wit:

We looked at the behavior of T10 yields, the 5s30s curve, and the SPX around the first continental US landfall of the five largest hurricanes since 2000, by insured losses. We index changes in these variables to their value on the first US landfall date, and look at the 20 trading days preceding first landfall, and the 30 days following. All the usual caveats apply: there were other market-relevant events coinciding with some of these large storms, and of course the sample size is small. We note that the 2008 experience shortly preceded the most acute phase of the financial crisis, and that the 2011 experience more or less coincided with the credit downgrade of US Treasuries and the debt limit fiasco that occurred that year. For that reason, we remove the obvious outliers and average across the remaining four storm experiences.

From that, Deutsche notes that “the moderate pop in yields and the curve roughly 15-20 trading days after first landfall” caught their interest.

“These hearken to the anecdotal experience that large insured losses could drive some insurers or reinsurers to liquidate assets (presumably largely long duration bond holdings) in order to free funds to service claims,” the bank writes, adding that “there is less apparent effect for the SPX, but taken together with the behavior of the level of rates and curve slope, the modest depression in SPX at roughly the same post-storm horizon could also be consistent with some asset liquidation.”

Will this be material? Who knows. But it’s something worth considering as you attempt to make a list of all the possibly relevant variables for your Irma-market equation.