Earlier Wednesday I posted my monthly update on Retail Sales. Those of us who routinely track this series know that the Advance Estimate, which is what I reported on yesterday, will be followed by a second estimate next month and a third estimate the month after. How big are those revisions? Big enough to warrant skepticism about the Advance Estimate?
Here is a visualization of the change from the first to third estimates from January 2007 through December 2012, the most recent month for which we have data points.
As we see, revisions abound, and they move in either direction. For a better sense of the magnitude of the revisions, the next chart shows the percent change from the first (advance estimate) to third (second revision).
During this timeframe there were 33 upward revisions and 39 downward revisions. The absolute mean (average) revision was 0.38%, which breaks down as 0.30% for the upward adjustments and -0.46% for the downward adjustments.
Are the sizes of these numbers significant? Consider: Over the same timeframe, the month-over-month absolute mean change of the latest revised retail sales series is 0.83%. A more dramatic way of thinking about this is as follows: Since 2007 the absolute second revision divided by the absolute mean advance estimate is a whopping 46%.
The message is clear: Don't take the initial retail sales data too seriously, which is the same message we saw for last Friday's jobs data.
Note: Actually there's yet more to the revisions than I've touched on in this discussion. We get annual revisions of the series as well. But for the purposes of this commentary, Im interested specifically in the month-over-month changes.