Irma, Harvey, And Janet: Define ‘Bad’

Blame Amazon (Prime Day payback). Or maybe blame Harvey. Or maybe just blame an economy that’s not really on solid footing.

Whatever you want to blame, August retail sales disappointed.

Treasurys pared losses and the dollar hit a new low for the day after the release, although yields and the greenback are still up handily from the lows hit last Friday ahead of Irma and ahead of an expected North Korean missile launch which finally materialized overnight.

Judging by stocks, the econ wasn’t bad enough to be “good” where “good” means it would materially impact the Fed’s reaction function and bolster the dovish case.

That brings us back to the question folks have been asking all week: will Irma and Harvey end up derailing a December hike by either denting the data or else making it so noisy as to be unreliable?

Goldman is out addressing that question for the umpteenth time, and although we’ve been over this before, it’s worth quickly running through the bank’s latest.

They begin by noting that between Harvey and Irma, the damages are expected to be more than $100 billion, before parroting the consensus line that “near-term growth hit from natural disasters tends to be followed by an equal or larger growth boost from reconstruction, and Fed officials seem to agree.”

More interesting is the following table which takes a look at which econ indicators contain sufficient granularity to allow Fed officials to spot hurricane distortions:

Here’s Goldman:

As shown in Exhibit 1, several labor market and housing indicators include geographic granularity, making it easier to identify storm effects. For example, we estimate that Hurricanes Harvey and Irma could depress September payrolls growth by as much as 100k. While that would normally add considerable uncertainty to the growth outlook, the release of state-level payrolls and unemployment data two weeks later should clarify whether job growth remained stable outside of Texas and Florida. Similarly, a potential decline in August and September homes sales and housing starts would be less concerning if the weakness were concentrated in the South.

Of course “tells” are just “suggestive” (i.e. not definitive) but the bank does note that jobless claims offer a decent read:

The increase in Texas jobless claims over the last two weeks has already exceeded the average during the top 10 most damaging post-war disasters, as shown in Exhibit 2. And Florida claims seem likely to rise meaningfully in next week’s report, given that over 2 million Floridians remained without power as of Thursday afternoon. This combination of elevated jobless claims and near-record hurricane damage suggests that growth is indeed likely to slow over the next 1-2 months, as has occurred historically on average (see Exhibit 2, right panel).

In the end – and this is the punchline – “disaster capitalism” wins the day. Because as Goldman reminds you, “transcripts from the first meeting after Hurricane Katrina show that participants expected only a temporary drag on growth [and] in fact, the Fed staff revised down its growth projection for 2005 but revised up its forecast for 2006 by slightly more, similar to our expectations for Harvey and Irma, as shown in Exhibit 3″:

So bad news (hurricanes) is good news for the economy (rebound tied to the rebuilding effort) which means Harvey and Irma aren’t likely to influence the Fed’s calculus for December (although a fiscal crisis might).

But wait a minute. Bad news is also good news for markets but for the opposite reason. The worse the news, the less likely it is that the Fed can hike. But as shown above, to the extent this particular breed of bad news (natural disasters) is good news for the expected economic payback down the road, it might actually raise the odds that the Fed continues on its hiking cycle undeterred, which would mean “bad” hurricane news is actually not “good” news for markets.

Now define “bad.”

Because we can’t do it anymore.