Art Cashin On Unadjusted Payroll Seasonal Adjustments
We (and Charles Biderman) have previously discussed the seasonal adjustments to NFP data, which while potentially credible in a releveraging context, is far less meaningful when used on apples to apples basis for months in which there is material wholesale deleveraging and record warm weather. Yet the rub lies precisely in the seasonal adjustment, which for January and February has "added" nearly 4 million jobs based on nothing but historical regression patterns, and the "beats" represented less than 5% of the total addition, implying even a modest miscalculcation would have had a huge impact on market, and political, interpretation of the data (as explained here). Today, it is the turn of Art Cashin, quoting Lakshman Achuthan, to provide his take on "unadjusted seasonal adjustments."
From UBS Financial Services:
Seasonal Adjustment Is Unadjusted? - Lakshman Achuthan, co-founder of ECRI was interviewed yet again by Bloomberg’s Tom Keene. Achuthan is almost in campaign mode as he pounds the table reiterating the ECRI posture that we’re headed back into a recession.
His tone seems to indicate his frustration with what he sees as clearly misleading data. Here’s a bit from the new interview:
Most economic data is seasonally adjusted. This is a good thing because there are seasonal patterns during the course of the year. But the sheer size of the recession we went through had an unintended impact on the way those algorithms run. When the economy fell off a cliff in the Great Recession it was like no other recession we have experienced, so it wasn’t easily compared. The systems received data in Q4 and Q1 expecting it to be particularly weak on a seasonal basis. Therefore, they adjusted upwards and that was not intended. There is an easy, un-confusing, fifth-graders-can-do-it, way around this, which is to look at the year-over-year growth rate which shows something quite ominous. When we look at our forward-looking indicators both sets surged initially coming out of the recession. Then they rolled over. They popped up briefly again about a year ago and now they have turned down again. The Weekly Leading Indicator is now at its worst readings since July 2009. These leading indicators have hardly been swayed from their recessionary trajectory. So it brings the bigger question, can unprecedented global monetary policy repeal the business cycle? And these pictures say no.
That re-sparked curiosity about last week’s citation of the King Report’s assertion that without seasonal adjustment, payrolls have actually fallen 1.8 million jobs. We’re still trying to run down that stunner.
Just don't tell Ben about the bolded, underlined bit above. It would make him sad that he can't single handedly control the worlds of finance, economics, and statistical mean reversion.