Maybe it’s merely a drunkard’s walk temporarily wandering on the straight and narrow; randomness in the guise of an orderly pattern. That’s probably what Leonard Mlodinow would suggest. He wrote the book called The Drunkards Walk, and he says “we habitually underestimate the effects of randomness.” But, then again, maybe it’s NOT random. Besides, what are the odds of a statistical report following such a steady cadence, repeatedly, for a quarter century? A long shot at best, I guess.
The yearly percentage change for the government’s report on US Retail Sales has its peak month ‘on average’ 68 months after the previous high water mark. The variation of the average gap has never been more than four months; it is a very consistent pattern. An important caveat is that this data series only goes back to the early nineties, so there are not many occurrences. But the regularity of it is compelling. The months with the peak year on year sales rates are marked on the chart below. The descending growth rates are notable, but it’s the gap that’s the thing. The most recent peak sales rate was in January 2017, so if the previous pattern is something more than randomness in disguise, then this could signal steadily weaker aggregate demand for some time.
The peak annualized rate of Retail Sales does NOT indicate a peak in the GDP; there is more noise in the economy other than the sound of the cash registers at WalMart or the computer click on the Amazon check out tab. But since the Personal Consumption makes up about seventy percent of GDP, the retailers cannot be ignored.
As I said above, the latest peak month for sales was at the beginning of this year, the others were in: March 1994; August 1999; June 2005; and June 2011. None of these months coincided with a quarterly GDP high point. One reason could be that the GDP inventory component was quite low in the quarters during the retail sales peaks and then quite high in the following quarters, suggesting that after the shoppers cleared the shelves the store restocked but shoppers didn’t come back. It is worthy of note that in the first three times the sales peaked the real annual GDP was at or near a high for that year. There was, however, nothing indicative about 2011, the last arrow on the right. And of course, it remains to be seen what happens with 2017.
Random or not? The chart below is the Fed Funds target rate. The circles indicate the periods when the Retail Sales annualized growth rate peaked. Rate hike cycles were underway in 4 of the 5 occasions.
Water rising in Texas; people rising above it
The brown spots on the US Energy Information Administration chart below are Gulf of Mexico oil rigs. The green and black lines track the route travelled by Hurricane Harvey in recent days and the estimate of where it is heading.