Record Retail Year - December Recorded Gains Reported as A Drop from November

We were on track for our best retail sales year prior to the Government Shutdown. Preliminary data was that Visa/Mastercard saw a spike in purchases and possibly the best December Retail Sales Month ever.

Yesterday's article on the December MARTS Retail report, "Dec. Retail Records Finally Reported (Part 1)"  examined the non-seasonally adjusted data. We saw a drop in Home and Building Goods sales as well as Gasoline Sales recorded, as expected. Furniture Sales Fell, as did Gasoline Station and Sporting Goods Sales.  Everything else was higher month to month and December to December. We continue to see the trailing year, or rolling year, growth at over 5%. There was headline news that we saw a decline in seasonally adjusted retail sales. This is very misleading.

One of these charts is unlike the others. Seasonally Adjusted month to month declines were recorded in all but two sectors.The seasonally adjusted December to December data shows that all sectors except Motor Vehicle and Auto Parts (MVP) and Building Material and Garden Equipment (BMGE) declined. Where we saw increases in all but two sectors, non-seasonally adjusted, we say declines in all but two sectors seasonally adjusted.

The December to December Growth revealed non-seasonally adjusted declines in three sectors. We saw declines in Furniture and Furnishing, Gasoline Station sales, and SGHBM. This is a good comparison because it compares the same month data. Christmas comes the same time each year. Thanksgiving is a floating holiday. Sometimes it is the final week of November, some years it is the fourth week of a five week month.  It is interesting to note that the month to month NSA chart looks very similar to the SA December to December chart.

The December to December declines were comparable for the seasonally adjusted data and the seasonally adjusted data. Declines  were reported in Furniture and Furnishing (FFS,) Gasoline (GAS,) and Sporting Goods, Hobbies, Books and Music (SGHBM) Sales. How did a gain of 34.71% in Clothing sales get seasonally adjusted to a gain of only 4.57%? How did a 29.36% increase in SGHBM become a seasonally adjusted loss of 13.02%?  How did a NSA gain of 20.4% in General Merchandise (GM) get seasonally adjusted to a gain of only 2.23%?

What the heck happened  to the month to month data? The seasonal factors skewed the data. If the seasonal factors that were used to convert the non-seasonally adjusted November MARTS data to the seasonally adjusted (SA) November MARTS were applied to the December NSA MARTS data then the SA December MARTS would have been reported at $553.639 billion dollars and not the headline $505 billion dollars.

What happened to $50 billion in retail sales? Only two sectors were reported higher than they could have been reported. The seasonal factors change category by category, month to month, and year to year. If the seasonal factor is over 1.00 then the seasonally adjusted data is reported higher SA MARTS data is reported lower than the NSA MARTS data. the the NSA data was reported. If the seasonal factor is under 1.00, say 0.90, then the SA MARTS data will be reported lower than the record NSA Values. This is what happened this month.  The seasonal factors for November were higher than those for December for all but two sectors. GAS and BMGE were the only two sectors with higher seasonal factors during December than during November.

It is critical that people acknowledge that there are seasonally adjusted data and non-seasonally adjusted data. It is important to note that we had 6.04 trillion in retail sales this year whether or not we use the SA or NSA Retail Sales data.  This is important because personal consumption expenditures is a large component of the Gross Domestic Product. The GDP data is seasonally adjusted. We grew at 5.05% NSA and 4.94% seasonally adjusted.  If you look at the NSA data there are month to month variations in the data. Some months are strong than others. May through December the SA Total retail sales were over $500 billion. The NSA data was  over $500 billion during March, May, June, July, August, October, November and December. Either way we were over $500 billion a month eight months of the year.

Look at the data. We had a record December whether or not we use the seasonally adjusted data or the non-seasonally adjusted data. We had a record year, over $6 trillion dollars, whether or not we use the NSA or SA data.  This report reveals the perils of discussing the seasonally adjusted data. Up was reported as down. This was the advance retail data. The data released this month also included the preliminary and final data for November and October.  Both October and November had upward revisions to the NSA data and downward revisions to the SA data.

This was a solid report. The reporting on the data was weak. We experience a non-seasonally adjusted world. Only the government lives in the seasonally adjusted world.

It's the economy.

 Reclaiming Common Sense