Zero Hedge
All posts from Zero Hedge
Zero Hedge in Zero Hedge,

Citi Discovers The Culprit For Poor Jobs Data...

Earlier this year, the BEA officially jumped the shark when, much to the delight of Steve Liesman, the San Francisco Fed (Janet Yellen’s former stomping ground), and anyone who enjoys a good goalseeked statistical revision, the US decided to double adjust its GDP figures.

The second revision was justified by reference to a bit of academic nonsense called “residual seasonality.” For those who might of missed this landmark moment in stupidity, here’s a brief recap: 

It's official: after seeing it work so well for years in China, the US Department of Commerce's Bureau of Economic Statistics has officially replaced all of its excel models with just one function. The following:


As Steve Liemsan hinted a few days ago, in what we thought was a very belated April fools joke, th eBEA has finally thrown in the towel on weak seasonally-adjusted US GDP data, and as a result has decided to officially proceed with a second seasonal adjustment: one which will take all the bad data, and replaced it with nice and sparkly, if totally fake and goalseeked, GDP numbers.


As Bloomberg reports, "the way some parts of U.S. gross domestic product are calculated are about to change in the wake of the debate over persistently depressed first-quarter growth. In a blog post published Friday, the Bureau of Economic Analysis listed a series of alterations it will make in seasonally adjusting data used to calculate economic growth. The changes will be implemented with the release of the initial second-quarter GDP estimate on July 30, the BEA said."


In other words, as of July 30, the Q1 GDP which will have seen its final print at -1% or worse, will be revised to roughly +1.8%, just to give the Fed the "credibility" to proceed with a September rate hike.

Well don't look now, but Citi (whose economists have been on roll lately when it comes to churning out riotous research calling for helicopter cash to save the world) is out with a new note that suggests the reason the Fed may have missed its window to hike is because they were relying on faulty jobs data which allegedly suffers from ... you guessed it... "serious residual seasonality." Here are some notable excerpts from the note:

Several times during this expansion, the Fed has made important policy decisions in the month of September. The history of payroll revisions for August and September has shown that some of those decisions might have been based on false signals, which could have biased policy toward accommodation.


We have found serious residual seasonality in payroll reports for the period from August through October. Weak August and September readings typically lead substantially stronger October reports, including sharp upward revisions to August and September data. This pattern has given rise to shifting market (and Fed) views of the labor market – from weakness to strength – each fall.


We have seen that the seasonal weakness in August payroll gains has supported accommodative policy decisions during this expansion. For instance, shortly after the extremely weak August 2011 payroll report, the Fed announced the plan to ease through Operation Twist. In 2012, after another soft August figure, the Fed announced QE3. In September 2013, the Fed unexpectedly delayed tapering QE3.


In each case, subsequent data showed that the August jobs data had been misleadingly weak – the labor market had been improving all along. 

See how that works? Where the data dependent Fed screwed up was when they decided to depend on the data which, contray to what it seemed to show, was actually "improving all along." This is the same logic Japanese PM Shinzo Abe employed overnight when he argued that Japan "has shrugged off the deflationary mindset,” even though the latest data shows that the economy just slid back into deflation. 

So what, you might ask, should the Fed have done if it wanted to avoid making a policy mistake by missing its window to hike? Here's Citi with the answer: 

If history is any guide, upcoming employment reports may reestablish that the labor market continues to improve rapidly. Although we expect another modest headline print of 180K on Friday , the low reading reflects the seasonal anomaly that September jobs prints tend to be among the lowest of the year. We see this reading as a lower bound of the true underlying path of employment.


Importantly, we would not be surprised if the September report included notable upward revisions to earlier months, based on the pattern of the past four years. 

Apparently, the Fed should just assume that the data is actually better than it is by factoring in a possible upward revision in the future, an assumption which would then allow the FOMC to proceed with rate hikes. Of course that raises the obvious question of what happens if the data is actually as bad as it looks and the Fed, on Citi's advice, ends up hiking into poor jobs data. This also begs the question of whether the Fed should take a similar approach with all the other data it looks at. That is, should the FOMC just assume that a sub-zero Q1 GDP print will eventually be double adjusted to show that the economy actually expanded? And if so, and if that gives Yellen the greenlight to hike, at what point can the data which telegraphs the economic malaise that results from that hike be trusted as a signal the Fed needs to reverse itself? Or should the Fed just assume that all the data is wrong, that everything is actually fine, and that inflation isn't actually running below target?

Who knows, and perhaps it's best to just close with the simplest assessment of Citi's latest pontificating: the bank has now blamed poor jobs data on... summer.