I was made aware last week that ZeroHedge had picked up on some comments made on Bloomberg by MacroMavens economist Stephanie Pomboy about “shenanigans” in the seasonal adjustments for the most recent Retail Sales release (May 13).
Video is here.
…looking at the numbers this morning I went in – and this is a little in the weeds on the economy – [crosstalk] okay because it’s important very quickly like to cover and that is they use something called a seasonal adjustment for all the data that hits the tape each day and the idea is they adjust for things like whether Easter fell in March or April or whether it’s Fourth of July and people are buying hotdogs to put on the barbecue or Valentine’s Day they’re buying chocolates and they adjust for all those distortions, and in this current April they used a very low bar relative to what they’ve used in prior Aprils, like a third. Normally they expect April to be down from March but they expected it to be down three times as much as it is in a normal month of April. So they did a little statistical shenanigans there that boosted the number. Had they used the same factor that they’ve used in the prior Aprils, retail sales would have printed down .9% rather than up 1.3. So it’s huge story and people will push back and say well but Easter did fall in March instead of April so maybe maybe March is where the number is, so I went back and looked at that seasonal factor and they also low-balled March so it’d be one thing if they raise the bar in March and then lowered it in April, but they didn’t do that so this is really suspicious.
The claim Ms. Pomboy is making, if true, certainly should be news-making. I’ve sifted through the release and the back-data, but can’t corroborate the claim being made; nor can other folks with whom I’ve communicated on this, including several leading Wall Street economic analysts as well as the data statisticians at Census.
The actual data released by Census on seasonal adjustment multiplier in April of 2015 was 0.998; in April of 2016 it was 0.997 — virtually unchanged. To be blunt, I have no idea what the basis of her statement is in data or in fact.
That said, I am not trying to throw shade on Ms. Pomboy, I simply cannot find any data to back up her claims anywhere I look. (And trust me, I looked!)
I’m always in learning mode, and if I can learn something here, so much the better; maybe I’m looking in the wrong spots or overlooking something. So, my question is simply this: Is anyone aware of exactly the “shenanigans” to which Ms. Pomboy is referring? If so, drop a note on Twitter to @TBPInvictus.