MBS Day Ahead: Jobless Claims, Last Auction, and More NFP Reflections

Let's do the same thing we did yesterday and get the daily analysis out of the way in one paragraph so I can wrap up the "NFP Reflections."  Yesterday's got a bit wordy, so there's not much more to say, just some data that I put into chart format.

Today brings the only true "economic data" of the week in the form of Jobless Claims at 8:30am.  Markets aren't necessarily chomping at the bit to react to this release, but it could have some effect.  The last of the week's Treasury auctions wraps up at 1pm and any post-auction-supply rally possibilities will be evident (or not) shortly thereafter.  If we happen to be rallying, what we're hoping to see is a challenge of the prevailing uptrend intact since late May.

Moving on now to some additional data from yesterday's NFP Reflections.  The whole point of that rant was to suggest that there are so many nooks and crannies in a report like The Employment Situation that almost any case can be made for almost anything if one digs deeply enough.  The point of that point was to balance out the semi-viral cynical arguments that tend to surface after NFP data. 

To be very clear, it's NOT that the cynical takes on last week's NFP have provided phony data.  The data is quite real--directly from the Bureau of Labor Statistics.  It's just that it doesn't necessarily mean what they say it means.  Reason being: the alarming statistics are based on SEASONALLY ADJUSTED data.  Looking at the unadjusted data, things are less alarming, and as we discussed yesterday, the first 6 months of 2014 are actually much better than the first 6 months of 2013.  Here's a chart of the first 6 months (total full-time payrolls from Jan-June) by year:

This chart tells a pretty good story!  It shows payroll creation in positive territory, but leveling-off without the stampede toward labor market prosperity that the cynical media pieces seem to be arguing against.  Once again, the truth is usually somewhere in between the strongest opinions on either side of an argument.  It's probably the same here.

As for 2014 not fitting well with the seasonal adjustment model, it's pretty easy to see why that might be if we take a look at historical averages vs 2014.  Here's a chart of the 10yr average of monthly full-time payroll creation for the first 6 months of the year compared to those same 6 months in 2014.  As you can see, every month was higher than average until June. 

To the cynic's credit, this certainly supports the notion that June was different and not so great for full-time payroll gains!  (although it's worth noting that unadjusted payrolls still grew in June, just not by asmuch as they normally do).  Anyway, the big gap between the purple and red lines there at the end is the reason the adjusted numbers seem so awful for full-time payroll creation. And again, that awfulness was more than offset by the better-than-normal full-time payroll creation earlier in the year.

What sorts of underlying structural changes could be at work here that are throwing off current numbers from long term averages?  There are so many places one can look for such things.  Certainly the ACA argument has to have some validity (employers less eager to offer full-time positions due to ACA), but that loses quite a bit of credibility based on that "first 6 month" chart above.  One place I thought to look was at the youth employment picture (age 16-19).  Before I even exported the BLS data into excel, the auto-generated chart on the BLS site provided a clue.

In case you don't want to try to figure out what that means, it shows fewer and fewer 16-19 yr olds getting full-time jobs over the years.  This is unadjusted data and so the spikes you see are the people getting summer jobs after school.  The 2014 spike isn't up to snuff.  Maybe that has to do with school year changes or maybe it's part of the long-term evolution of the above trend.  Either way, 16-19yr olds didn't show up in the same numbers they normally do. 

The chart below shows this, but it too shows that full-time payroll creation was higher than average earlier in the year.  I don't want to speculate without being more familiar with necessary details, but the "back-from-school" full-time job is probably the most at-risk of being notched down to 34.9 hours due to ACA.  I don't know if that's happening.  But if it is, it would be a great explanation for the much weaker than normal full-time payroll creation in June.

That's about it for now on all this deep thinking regarding payrolls data.  Honestly, there's more of it in just this one report than I will ever care to look at--let alone collate and present.  We've barely scratched the surface here, but hopefully enough to convince you not to trust anyone!   Even me!  I know not everyone has time to do that, so I do my best to present both sides of an argument and let you decide which one you like best.  The "sky is falling" side of this NFP argument already had plenty of representation, so hopefully this balanced that out a bit.  Once again, I'm not saying the sky isn't falling, simply that it's not quite so black and white.