On Friday the Department of Commerce will report August retail sales, a material report which all else equal, may influence whether the Fed proceeds with its plans to unveil balance sheet tapering in its upcoming FOMC meeting. However, as we discussed last week, the report, together with virtually all other high frequency economic reports, will be materially distorted by the destructive aftermath of hurricane Harvey (Irma's impact will be felt in the September retail report).
While Goldman recently showed the historical impact of hurricanes and other natural disasters on virtually every economic data series...
... of particular interest in the coming days will be the biggest driver behind the US economy, namely retail spending, and specifically whether the recent natural disasters led to a sharp - and potentially sustained - slump. According to internal Bank of America credit and debit card spending data released as usual just days ahead of the official government report, there does appears to be a substantial adverse impact. The question is how much of this is secular, and how much is a continuation of recent weakness in retail spending. Further complicating matters is a seasonal quirk, with the August spending report coming at the peak "back to school" spending period, coupled with the recent Amazon Prime Day which led to further distortions in retail spending patterns.
As BofA's Michelle Meyer calculates, retail sales ex-autos, as measured by BAC aggregated credit and debit card data, declined 0.1% mom seasonally adjusted in August, leaving the 3-month moving average tracking flat for the month. Consumers shifted spending to gasoline stations, which were up strongly in the month, owing in part to Hurricane Harvey.
After controlling for the increase in gasoline spending, retail sales ex-autos and gasoline declined 0.4%: one of the sharpest declines YTD, and a confirmation of the continuing divergence between BofA (blue line) which has hugged the flatline in recent months, and official government data, which while week, has demonstrated modest Y/Y growth.
According to Bank of America, there are three key factors influencing the data this month:
- Hurricane Harvey;
- the pull-forward of retail spending into July by Amazon Prime Day; and
- back-to-school shopping.
In an attempt to isolate the influence of Hurricane Harvey which made Texas landfall on August 25, BofA first examined daily spending in Texas which shows that spending picked up in the days heading into the hurricane but remained depressed through the event and in the days after, as one would expect.
We estimate that the net reduction of spending in Texas sliced 0.1-0.2pp from the monthly growth rate of total retail sales ex-autos in August. We then dug deeper and looked at the impact by the type of spend which reveals that necessary items (food and gasoline) increased in the month while more discretionary items declined (Chart 2). We also measured spending by major region in Texas (MSAs) which shows significant decline in Houston but continued growth in regions not hit by Harvey (Chart 3)
The charts below provide further evidence that Harvey caused a net drag to spending in the areas hit directly. In contrast, there was trend-like growth in MSAs in Texas which were not directly impacted by Harvey.
However, it wasn't just Harvey explaining the sharp drop in ex-gasoline sales. In addition to the adverse reginal impact from Harvey, August retail sales were also likely held back by the strong success of Amazon Prime-day in July. BofA data shows that Prime Day pulled forward activity from August into July.
Finally, and perhaps most concerning, the third indication that retail sales are set to disappoint, BofA writes that while it did not find much of a story for the back-to-school season, its proxy for back-to-school sales showed growth of just 2.4% yoy, down more than 50% the 5.4% yoy pace last year.
This is a problem because according to the National Retail Federation's annual survey, families were projected to spend approximately $29.5bn on back-to-school items which would translate to an 8% yoy increase from the prior year's spending plans. Unfortunately, those spending plans have not translated to actual spending as expectations have once again overshot spending patterns as they did in 2011 and 2012 but were below in 2013-2015.
Finally, broken down by category, BofA finds that on a % mom basis, consumer spending declined in most categories in August with only food and beverage, gasoline stations and cruise showing an increase. As noted above, spending on food and beverage and gasoline stations likely saw a boost due to Hurricane Harvey as households stocked up on essentials.
BofA's Bottom line: the weakness in August retail sales, already expected to come in at near stall-speed levels, is likely exaggerated by the hurricane and July prime-day.
The good news is that while Hurricane Irma may depress spending in September, retail sales typically bounce back after a natural disaster, suggesting upside into 4Q. Unless, of course, it forces an even greater decline in spending, as the following charts showing the secular decline in retail sales indicate.