There was some confusion why following yesterday's stronger than expected retail sales, which broke a 4-month series of disappointing data and which according to most economists were good enough to bring forward the Fed's first rate hike in a trader generation, bond yields tumbled.
Well, in the aftermath of the whole "double seasonal-adjustment" travesty, in which even the BEA admitted any bad economic data (read Q1 GDP) will be "massaged" enough until it becomes good under the "scientific" pretext of "residual seasonality", and following the suggestion of Dynamika Capital [5]'s Alexander Giryavets, we decided to take a look at not seasonally adjusted retail sales.
We found something interesting.
The thing about retail sales is that while they are supposed to smooth out month-to-month changes in any given data series, they should be virtually identical on a annual, year-over-year basis. After all the same "seasonal" adjustment that was applicable this May, was applicable last May, the May before it, and so on, unless of course, there was some massive, climatic or otherwise shift to the underlying reality.
To the best of our knowledge there wasn't.
And indeed, when looking at the annual change in headline retail sales data we find that, as expected, the seasonally-adjusted (blue) and unadjusted (red)retail sales series are almost identical...
[6]
... but not quite.
If one zooms in on the most recent data, one finds something surprising: a substantial rebound in SA retail sales, which according to the Dept. of Commerce rose 2.7% while unadjusted retail sales rose by just 1.0%, the worst montly print in over two years and hardly inspiring confidence that the economy is strong enough for the Fed to engage in a rate hike.
[7]
To isolate the problem we decided to look at only the annual (YoY) change in May data. The chart below shows the surprising finding: while every year for the past 4 years the Unadjusted May data was equal to or stronger than the Adjusted retail sales print, in May of 2015 this was reversed, and quite substantially.
[8]
To show just how much of an outlier May 2015 was compared to May in prior years, here is the seasonal "adjustment ratio" for the month of May for every year from 2008 to 2015, by which we define the ratio of "seasonally adjusted" to "unadjusted" retail sales. Spotting the outlier should be easy enough.
[9]
So, our question: while we know that the US Department of Truth, err. Commerce, will soon adjust GDP data for all weak quarters higher just so the narrative of a rebounding economy isn't lost when one looks at the actual fact, has this "residual seasonality" adjustment already been applied to retail sales? Because we fail to understand just why the seasonal adjustment to May retail data should be as profound as shown above.
Source: (Double?) [10]Seasonally-Adjusted [10]and Unadjusted [11]retail sales
