Following the "blockbuster" jobs report from Friday, bond yields soared across the curve as experts everywhere decided that this time it's different, that this is it: with soaring October jobs, not to mention the biggest annual jump in wages, the Fed was out of excuses.
We would like to make three points.
First, the October jobs reports was good, certainly at the headline level, but there is one more payrolls release from the BLS before the December 16 FOMC decision; its impact will be far greater on the FOMC's decision especially if it now shows a sharp decline in jobs (before Friday, five of the past six payrolls reports had missed expectations).
Second, as we showed yesterday, [6]while the Establishment survey reported an unexpected jump in jobs in October jobs, which rose by 271K, far above the highest estimate even putting the very Fed to shame after St. Louis Fed's Bullard explicitly talked down market expectations the night before while trying to convince the market [7]that slowing jobs is not bad for the economy, a deeper look revealed that the Household survey suggested a far less optimistic picture, with 378,000 of the increase in employed falling in the 55 and over age group, while males 25-54 saw their jobs decline by 119,000.
[8]
But while this suggests that the headline jobs number may have gotten far ahead of itself and upon reflection will be revised well lower, it does not explain how the October report also showed the biggest jump in average hourly earnings, which rose by 2.5% in October from a year ago, the highest annual increase since the crisis.
[9]
Which bring us to the third, and most important point. Where did the wage growth come from?
And therein lies the rub, because after the very disappointing job wage growth in September, October's jump in wages suggests the economy may just be strong enough to weather a rate hike.
There is one problem with this.
To find where this increase in wage growth came from, we broke down the weekly payrolls increase (Y/Y %) by the 10 constituent industries that make up the private payrolls (excluding government). What we found was surprising.
Instead of some broad rebound in October wage growth (red bar in the chart below), and certainly not due to payrolls for mining and logging workers, which is plunging, the entire jump in October average hourly wages can be attributed to just one industry - the one highlighted with a yellow in the chart below: construction.
[10]
What this chart shows is that of the 120.7 million private payrolls in October, 4 industries (Mining and Logging; Trade, Transportation and Utilities; Financial Activity; and Leisure and Hospitality) employing 51.3 million workers saw a drop in the annual wage growth rate, which means they dragged down the average hourly earnings, one industry, Manufacturing (which employs 12.3 million), saw no change in its weekly payrolls, and another five industries, which employ 57 million, saw a modest increase in October weekly wages.
And yet nobody was as instrumental to the October wage jump as the Construction industry. That said, nobody was as instrumental to the miss in September wages as the Construction industry because not only did weekly hours for Construction workers drop, so did weekly payrolls. All of this was offset, and then some, was offset in October.
So what is going on here?
The impact of Construction jobs on weekly payroll growth is seen best in the chart below which shows that if one excludes the contribution from Construction, annual wage growth is not improving at all, and in fact is roughly where it has been for the past year, nowhere near close to the blockbuster print revealed on Friday.
[11]
Furthermore, as the next chart below shows, the annual growth in weekly
payrolls for Construction workers appears to have finally topped out, and after four
years of increases, 2015 has been the first year in which the pace of
growth declined - clearly weekly payroll growth for Construction workers
is slowing in the second half of 2015. And yet, October weekly payrolls soared, and
at 8.1%, increased at the same pace as last year,
offsetting the recent drop.
[12]
Is this sustainable?
The chart above also shows that in the New Normal, not only has there never been such a sharp drop in September Construction weekly payrolls, there has also never been as sharp a jump in October weekly payrolls.
So what happens in November? That, dear readers, is the 25 basis point question - if Construction payrolls rise as the same pace as October, a December rate hike is assured. If however, this industry which employs just 5.3% of the US private workforce, "reverts to the slowing mean" in November, and after a torrid October we have a disappointing November (see red question mark above), then all buts are suddenly off again.
Which is why the fate of the December rate hike is suddenly in the hands of some 6.4 million construction workers. If their wages jump in November, a rate hike it is. If their wages suddenly disappoint as the winter is about to unfold, then watch as the yield on the 2Y plunges in a microsecond when the next payrolls report is released as the December rate hike is once again put on indefinite hiatus.
[13]