Earlier this week we warned that based on the latest Gallup poll of some 15,000 US shopping adults (and thus far more accurate than the Department of Commerce seasonally adjusted, goal-seeked retail sales data), retail spending in August will be a stark disappointment to those once again holding off for a rebound in that all important driver for the US economy - consumer spending. As we showed, August spending was the weakest in nominal dollar terms since 2012...
... and was also the lowest since March, as a result of 4 consecutive months of y/y spending declines.
Earlier today, Bank of America confirmed just that using its internal data, which tracks aggregate spending on credit and debit
cards, showing that consumers reduced spending in August.
And they reduced it substantially: According to the Bank, while the drop was not as pronounced as what Gallup reported (which saw average daily spending slide from $91/day to $89/day), it was still enough to dramatically impact the economy: "our headline measure, retail sales ex-autos, plunged 0.8% mom seasonally adjusted."
The weakness was not focused geographically, and was widespread across the nation:
As BofA notes, "there was broad weakness in retail sales ex-autos and gas spending growth across metropolitan areas, with seven of the ten largest MSAs showing a monthly decline. The biggest monthly decline was in Dallas, followed by Miami and San Francisco. Both Dallas and San Francisco have experienced strong growth over the prior six months, showing a solid recent trend."
What could be causing this at the aggregate level? One explanation is far weaker than expected back to school sales:
Retail sales in August are typically boosted by back-to-school shopping. Our proxy for back-to-school sales, which includes teen retail stores, sporting goods and categories within electronic stores and school supplies, was up 2.6% yoy on a seasonally adjusted basis. But this is a slowdown from the past few years and consistent with the slower yoy trend in retail sales ex autos and gas (Chart of the month).
The late timing of the Labor Day holiday may have created a downside bias to back-to-school sales this August. Many schools start after Labor Day, which may push back-toschool sales from August to September. Indeed, the seasonal factor for our back-to-school composite seems particularly large this year given the notable adjustment in the timing of the Labor Day holiday to the 7th of September this year from the 1st last year.
Some more details:
- Sales at teen retailers declined 1.3% mom in August on a seasonally adjusted basis, based on the aggregate BAC card data. This left sales down 6.2% yoy, continuing the weak trend over the past few years.
- The weak performance of teen retailers in August is an indication of a sluggish back-to-school shopping season
- Spending at sporting goods stores inched up 0.2% mom SA. This left the trend positive with an increase of 2.0% yoy.
- Budget constrained households continue to direct spending to on-trend athletic apparel (“athleisure”), which is gaining share at the expense of casual apparel.
The weakness was most pronounced at department stores, where sales fell 0.2% M/M and tumbled 3.5% Y/Y. While this is not a new trend and indicates the shift away from bricks and mortar to online vendors, the lack of any improvement merely confirms the broad weakness in consumer spending in the past month.
Finally, as Bank of America reiterates our conclusion from earlier this week when we observed precisely this only with Gallup poll data, "the weakness in the August BAC data suggests a high risk for softness in the Census Bureau advance retail sales report given that the two measures trend closely. While we know that the retail sales figures are volatile and subject to revisions, it is hard to ignore a weak report."
Why is all of the above particularly important? Because with the August Retail Spending report due out the morning of September 15, it will be the last report on the economy the Fed will read ahead of its "most important if not ever then surely in the past decade" FOMC meeting starting on September 16, and concluding with the 2pm announcement on September 17.
Following today's plunge in consumer confidence (which as a reminder Bill Dudley warned two weeks ago he will be very closely watching ahead of the FOMC meeting) and what is set to be a big drop in the retail spending report next Tuesday, will Yellen really be "data-dependent" if she hikes just as the economy is rapidly downshifting, not to mention the US consumer is about to tap out once again?
If nothing else, the "Dow-dependent" Fed now has a very clear "data" out to delay its September rate hike by at least 3 months, even if the actual delay driver has nothing to do with the US economy whatsoever.