If you want a tidy illustration of why building a plan around monthly economic data is a bad idea, this week supplied one for free.
July retail sales fell 0.6%, the Commerce Department reported — well short of the 0.1% increase economists expected. The number landed a week after a jobs report showing the economy lost 23,000 jobs in July, and alongside moderate inflation figures. Three data points, one story: the economy that looked to be accelerating in the spring was cooling.
Then August came in strong, rebounding from July. And on Wednesday the Atlanta Fed revised its GDPNow estimate for third-quarter growth up, to 5.1%.
What was actually soft, and what was a calendar quirk
The July weakness looks less alarming under the surface, and this is where careful reading pays. Sales at nonstore retailers — mostly online shopping — fell 2.2%, pulling down the overall total. Many analysts attribute that drop to a scheduling change rather than to consumer distress: Amazon.com (AMZN) held its Prime Day event in June this year instead of July.
Move a shopping holiday and you move the sales into a different month. The economy didn’t change; the calendar did.
One piece was harder to explain away. The “control group” — the everyday categories economists track most closely because they strip out the noisiest items — fell 0.4%, its first decline this year. And a separate University of Michigan preliminary survey showed consumer sentiment reversing recent gains, with shoppers pointing at persistent price increases and the Iran war.
Why the Fed raised through the soft patch
This is the part that clarifies Wednesday. If you were building the case to wait, July’s data was your evidence: soft sales, negative payrolls, moderate inflation. Goldman Sachs (GS) economists made exactly that argument, calling the economic case for raising weak.
The committee went the other way, and the reasoning is instructive. The officials who wanted to move weren’t arguing about what the last three months showed. They were asking what would give them confidence that price growth will be muted in six — and the assumptions underpinning the wait-and-see case had stopped holding. Energy prices were supposed to have receded. A softening job market was supposed to take pressure off wages. Neither happened.
The household version of this lesson
Sentiment surveys and single-month prints are genuinely useful for economists and genuinely useless as a trigger for changing a retirement allocation. July said one thing, August said the opposite, and anyone who rearranged a portfolio on the first print paid twice.
The durable signal in all of this is not the monthly number. It is that shoppers are citing prices and a war as the reasons they feel worse — two things that show up in a budget long before they show up in a brokerage statement.
Which makes this a good week to look at the household cash-flow page rather than the market page. Fifteen minutes, a recent statement, and an honest look at what the last twelve months of prices did to the monthly number you actually live on.