In keeping with the now constant trend of baffle with BS, since everyone was expecting a weaker advance retail sales print, just like with the BLS report, it was virtually assured that the data would prove everyone wrong. Sure enough, moments ago the Census department announced that headline retail sales rose by 0.1% in April, from a downward revised -0.5% in March, beating expectations of a second decline in a row of -0.3%. Retail sales ex autos were in line with expectations at -0.1%, on expectations of a -0.2% print, but it was the sales number ex-autos and gas which surprised the most, rising 0.6% on expectations of a +0.3% increase, up from a -0.1% decline. Of note: gasoline stations saw a dramatic 4.7% drop in sales, following a 3.2% drop in March: are all US commuters now using Back to the Future type hoverboards or just driving to work in the Tesla Model S?
A breakdown by retail sales category:
But perhaps most notable is that just like every other data point, the magic was in the seasonal adjustment, because the NSA data declined from $427.8 billion to $416.5 billion. And in an econometric model in which now more and more 'experts' agree there is a big problem with the winter to spring seasonal transition, one wonders just how credible not only the April, but the last several month's adjusted accuracy is if indeed the adjustments are not to be trusted?
The chart below shows the March to April change for the past decade on an Adjusted and Non-Adjusted basis. Perhaps there is something to the whole "seasonally adjusting" our way out of trouble in the third worst such monthly transition in the past decade.
