Retail Sales and Payrolls

Includes: AGG, DIA, QQQ, SPY
by: Erik Dellith

July’s retail sales figures are the strongest in months. Of course, people are more likely to go out and spend if they have jobs. With that in mind, note that July’s BLS non-farm payrolls were also the strongest in months. In fact, if we look at the last four months, the two statistics seem to track relatively well.

As shown in the table below, the pace of job creation eased considerably in May, and the rate of increase in retail sales also eased. Job creation slipped further in June, and retail sales also grew only marginally that month. July, however, saw a large increase in hiring and a nice jump in retail sales.

Retail Sales (ex-autos) and Non-Farm Payrolls
Month Retail Sales % change Non-Farm Payroll Change
April 0.35% 217
May 0.28% 53
June 0.2% 46
July 0.48% 117

While the latest numbers will likely go a long way toward reassuring investors that the US is not going to slip into another recession, it also begs the question about the dynamics of future retail sales. To get a general idea of what to expect down the road, I take a quick look at the longer-term relationship between retail sales and non-agricultural payrolls.

I recently downloaded the retail sales figures from the U.S. Commerce Department. I calculated the one-month percent change on seasonally adjusted total retail sales, ex-motor vehicles. The data set begins in 1992. Although a longer time series would generally be better, going back this far still enables us to examine the relationship between retail sales and the labor market during at least part of the first of the jobless recoveries.

I then take the correlation between the percent change in retail sales and the change in the BLS non-farm payrolls. This is where results become somewhat surprising.

Consumer spending depends on more than just whether or not people have jobs today; it also depends on whether or not people expect to have their jobs tomorrow, and the day after that, and so on. One way to gauge this expectation is to look at the rate of job creation. After all, if people see that the economy is adding a lot of jobs, then they will likely be more willing to believe that they themselves will be able to get jobs.

On a theoretical level, while jobs added today might boost retail sales today, jobs added last month, and the month before that, matter as well. Based on such reasoning, it is possible that the rate of job creation last month matters more for retail sales this month, than job creation this month does. Figure, someone has to get a couple of paychecks and some sense of job stability before they really go out and start spending, right?

Yet, given the information in the table above, it would seem like the strongest correlation between retail sales ex-autos and non-farm payrolls would be with concurrent data. Indeed, looking at the data going back to 1992, this is what I find.

Looking At Correlations

I lag the non-farm payroll figures up to six months and see which lag has the highest correlation with the one-month percent change in retail sales. As indicated in the table below, the correlations drop off quickly.

Lag Period Correlation
0 0.33
1 0.19
2 0.21
3 0.12
4 0.15
5 0.17
6 0.14

Thus, it seems that people are willing to take steps to immediately sate pent-up demand. Once they get that job, they go out and spend. Then the correlation drops. After two months’ time, the correlation falls considerably again.

The key take-away from this is that the bulk of the benefit to new retail sales from new job creation takes place in the first three months, with the largest part of that in the same month that people first get their jobs.

Looking Down The Road

So far this year, the economy has added, on average, 133,000 jobs per month, according to data from the BLS. The economy is still down more than 6.7 million positions since the start of the Great Recession. With the current average pace of hiring this year, the economy will need about 50 more months just to make up for the positions that had been lost, not including new entrants to the labor market.

Playing the averages for a moment, such a sluggish pace of hiring also points to average monthly retail sales growth (ex-autos) of a bit over 0.6%. Given the current slow-growth path of the U.S. economy, this seems like a reasonable expectation for much of the next year (absent some new catalyst for growth or a complete tanking in Europe). Investors focused on the retail sector should watch for above-trend job creation to provide a bit of a lift to retail sales.

Disclosure: I have no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours.