Friday, May 4, 2012

Was The Jobs Report Really Better Than Expected?

Prior to the market open today, the BLS released April jobs numbers that apparently largely disappointed the stock market with the SP500 down 1.5% as I'm writing this post. 

Below is a summary provided by First Trust.
  • Non-farm payrolls increased 115,000 in April and were up 168,000 including revisions to February/March. The consensus expected a gain of 160,000.
  • Private sector payrolls increased 130,000 in April. Revisions to February/March added 66,000, bringing the net gain to 196,000. April gains were led by retail (+29,000), professional/technical services (+28.000), employment services, including temps (+28,000), bars/restaurants (+27,000). The weakest sector was transit & ground transportation (-17,000).
  • The unemployment rate ticked down to 8.1% from 8.2% in March
  • Average weekly earnings – cash earnings, excluding benefits – were
    unchanged in April but are up 1.8% versus a year ago.
As noted by First Trust in that article, the interpretation of whether the numbers beat consensus depends on whether the economists includes revisions to the prior months. For April, this amounted to 66K jobs for the private sector.

Considering most economists forecast some number below 200K, the statistical significance of the revisions is in the range of 30-50% of the total jobs. Amazingly though very few of them appear to include or factor this number into the predictions.

First Trust has calculated that the average revision up for the last 10 months has been 40K. In essence knowing this statistical norm, a forecaster should've stated that with a expectation of 160K that the initial report would only be around 120K. Then as the numbers get revised up over the next 2 reports the results would more match his predictions.

Instead it appears that everybody wants to ignore the normal revisions, hence leaving a negative bias to the job market that isn't as warranted. Sure an ultimate outcome of 150-170K jobs for April isn't that great for the US economy, but it would've been inline with forecasts and sure beats the reported numbers.

What amazes me is that nobody whether on CNBC or various respected blogs points this out. They neither add the revisions to the April numbers or factor in the revisions to expectations. One would think the 'experts' on the jobs reports would be better at factoring in those revisions. Instead it appears that they want to pretend it doesn't exist.


Disclosure: No positions mentioned. Please review the disclaimer for more details. 



No comments: