Friday, October 21, 2011

Explaining the Retail Sales / Confidence Dislocation

My buddy Sami Mesrour (from Blackrock) had a write up earlier this month titled Follow What I Do, Not What I Say; Consumer Spending and Consumer Confidence that outlines the surprise rebound in retail sales (bold mine):

The US consumer is feeling down. Several indicators of confidence collapsed over the summer with the declines beginning in May as job growth slowed, and the bulk of the drop in sentiment occurring during August. It is likely that the intense focus on the country’s deficit problem and the attendant prospects of lower government spending going forward were the main drivers in the decline of consumer expectations. Forecasters are concerned over this development because of what it implies for the growth of consumer spending—historically sentiment has been a good reflection of sales in the US.
This time, however, something strange is going on: consumers are apparently saying one thing, but doing another
Very timely analysis, as this was ahead of last Friday's report that showed retail sales surprised, significantly, to the upside (September was up a 1.1%, while June and July were revised higher as well).

The year over year figure is even more impressive, up more than 8%.

Sami outlined (in detail - go to his report for more) the following four drivers:
  1. Borrowed time (ability for the consumer to once again borrow to spend)
  2. Income distribution (the rich keep getting richer and are driving spending, while individuals struggling are driving the confidence surveys lower)
  3. Transfer payments (outlined at EconomPic here)
  4. Foreclosures (squatters and those moving home with their parents are effectively not paying rent, increasing their ability to spend on goods)
I broadly agree with these points, especially #2 and #3, and I'll lay out a fifth... inflation. While inflation may be the best (and only) politically viable option to reduce the level of nominal debt in the system, there are always interesting implications.

The below chart outlines the nominal year over year change in retail sales, along with my best effort in matching the applicable inflation figures for each category (by all means imperfect). The real retail sales change is simply the difference between the two.

Building this out further with data going back to the mid 1990's, the below chart outlines the year over year change in the following consumption measures:
  • Nominal goods (very closely aligned with retail sales)
  • Nominal consumption (includes the less inflationary services sector)
  • Real consumption (i.e. less inflation)
  • Real per capita consumption
In addition, the dotted lines show the average for each category from 1996 to 2008 (i.e. pre-crisis).

What we see is that while the "spike" in retail sales (via nominal goods consumption) is higher than pre-crisis, all other measures are below their historical levels of growth.

In other words, the growth in the amount that individuals are consuming is lower, but individuals are paying more for what they are consuming. For those individuals where food and energy are a higher percent of their consumption basket (i.e. those that earn less), this is an even bigger impact.

Makes a lot more sense why individuals would be unhappy.

Source: BEA / BLS / Census


  1. It will be interesting to see whether continuing structural problems in the US labor market will at some point put a damper on consumer spending e.g. the high number of long-term unemployed, the rising ratio of unskilled-to-skilled workers and the falling labor force participation rate.

  2. Great post Jake. Always look forward to your posts and charts. Have a great weekend!

  3. Really good post. The assumption of correlation between confidence and spending is intuitive but, despite numerous attempts, I've never found a strong relationship. Consumers have their own agenda, and over- or under- spending accordingly. Their mood about the wisdom of the planchanges, but doesn't seem to affect outlay.