The U.S. Commerce Department Retail Sales Analysis Is A Useless Gauge Of Retail Health
Every month the Commerce Department releases its monthly retail sales reports. They typically report comparisons with the prior month. Let’s take November sales as an example: (“Overall US retail sales rose 0.2 per cent in November from the previous month”).
As usual, savvy retail watchers (including yours truly) groaned with the irrelevance of the comparison. Seriously, who cares? While there may be notes saying these numbers are “seasonally adjusted,” most retailers don’t even know what that means. Nor do they find it interesting.
I have been involved in the retail industry all my life (a long time). Not one retailer I ever worked for or with ever said “Oh dear… our sales numbers did not rise enough against the month before.” Never. Well, almost never. The ones who might look at those numbers are retailers up against the edge of their credit lines and running short of cash.
Here’s why. Retail is a cyclical business. In other words, revenue graphs tend to be curved, rather than linear. Certainly those curves have changed some over time. For example, January used to be a dead retail month. Fun fact: that’s how the National Retail Federation’s “Big Show” came to be held in January… retail executives had not much else to do while the accountants were taking care of physical inventory counting and closing the books on the prior year, and so looking at fixtures and technology, and comparing notes with each other on the prior season was a good way to spend a few days.
The advent of gift cards and improvements in technology have changed that. Consumers now shop January clearance sales to use their newly received gift cards, and counting products and closing the books happens far more quickly, thanks to technology advances. The Big Show is bigger than ever, and still held in New York in January, but it’s a tradition now, with no particular reason for the time and place beyond the tradition of it.
Similarly, July was also always dead as a doornail with consumers heading for the beach. Then along came Prime Day, and its associated call to buying action. Other retailers have followed suit, and July is becoming a revenue rich, profit-poor retail month.
It takes a year or two for these curves to become comparable. But apart from these two anomalies, demand curves are otherwise pretty consistent. And that’s why retailers look at the same month year-over-year, rather than comparing previous to current months. The operative word here is “comparable.”
Look at it this way. If someone published a news article saying H&R Block’s revenue numbers were down in May from April, you’d likely reply, “Well, doh… of course they are. Most taxes are filed prior to April 15.” So it is with retail sales.
So let’s take a look at year-over-year retail sales in November. According to the National Retail Federation, non-seasonally adjusted year over year sales improved 2.1%. Now, I’m not an economist (more about that in a minute), but I take this to mean that the 2.1% increase does not reflect the fact that Thanksgiving was a week later than it was last year, meaning that the “official” holiday season kick-off happened a week later. And the November numbers are therefore artificially depressed. The real question at hand is, how much labor and effort will it take to get to the 4.5% year over year increase retailers expect in one fewer week? Retailers seem confident, with most continuing to predict a strong season.
In fact, a thumbnail look at year over year November sales said to me “Hey…given the shortened season, and the potential impact of pre-buying on Prime Day, that’s a pretty good start.”
As I said, I’m not an economist. My knowledge of the subject consists of one semester of economics while I studied for my MBA and 10 years having a vacation home two doors down from Paul Krugman, whose path I never crossed in all those years. But I am, as I also said, a retailer. And the question retailers are asking today is not “How do we make December stronger than November?” It’s “How many markdowns do I have to take to get to the year-over-year comparison we promised our shareholders? Do we have enough inventory?”
I’m not sure when this month-over-month comparison become de rigueur for the Commerce Department. It’s definitely not a politically partisan issue. The same comparisons were around when Obama was president. And it’s not terribly different from the non-farm payroll comparisons (as if farming was the only cyclical industry). I just know that it’s not a useful comparison and actually shows nothing about the state of the industry.
So what’s my best advice to the non-retailer? Don’t make any stock picks based on these numbers. Don’t draw too many conclusions. While many say that comparable sales are no longer the best way to evaluate retailers, I promise you, they’re better than this month-over-month stuff. We’ll get smarter and more sophisticated over time, but frankly the Commerce numbers aren’t being helpful at all.