August Retail Sales Data Leave Lots of Room to Be Fooled by Randomness

When the July retail news was released, we commented that the July strength then in the numbers assuaged fears of a slowing consumer, but that they didn’t provide any indication of strengthening consumer demand. Obviously, today’s news is nowhere near indicating any strength, but is the weakening consumer story back in play?

The answer to that is just a maybe. As seen in the chart, underlying retail sales trends still look pretty steady. The recent softness pulls sales levels slightly below preceding trends, but not significantly so. The six-month growth rate shown there has fallen from 3.8% through April to 2.8% through August, but this fluctuation is within the range of (random) moves seen over the last two years.

That same trend is echoed by the store-type details of the sales data. August sales were especially soft at car dealers, clothing stores and online retailers. For car dealers, the swings were of a piece with the flat trend seen since late-2015. For apparel stores, the August declines merely offset an equally sharp July gain, leaving apparel sales right on the slow-growth trends seen over the past two years. The same is mostly true for online retailers, the difference from apparel being that the growth trend for online is better. For none of these three store types and for no other store type is there real evidence in the August data of new softness.

This last point is helpful in dispelling any claims that Hurricane Harvey influenced these numbers. Certainly the Texas coast was buffeted by Harvey late last month, and certainly retailing conditions in Texas in late-August were abnormal. However, it is arguable as to whether pre-Harvey fears stimulated sales or whether actual Harvey-related problems depressed them. Also, with today’s August data still preliminary and hurricane-affected stores likely slow to report, it is questionable whether actual hurricane effects—whatever they were—had any impact on today’s sales data. And finally, if there is indeed no real “news” in today’s data, as concluded above, why even try to spin them for hurricane effects?

The bottom line is that not much is going on, to date, in retail sales trends, one way or the other. Maybe some actual hurricane effects will show up in subsequent data, but for now, the real story in the economy is that stronger capital spending recently has given some life to factory-sector activity. Our focus will be on whether this development will be sustained.