The Fat Lady Is Clearing Her Throat

|
Includes: DIA, IWM, QQQ, SPY
by: Kevin Flynn, CFA

Summary

"The fat lady sings" when employment rolls over.

Clues from monthly and weekly data.

Time is running out on the business cycle.

In a Seeking Alpha column I wrote last September, entitled "It Ain't Over Until the Fat Lady Sings," I put forth the assessment that although certain sectors of the economy were clearly struggling, the cycle wouldn't be over until the "fat lady sings," an event that would be signaled by trouble in the employment sector.

Trouble has indeed arrived, but the good news is that it is as of yet not the definitive sort, as we will see below. That gives you time to exit the equity markets in a peaceful manner. The bad news for the investor community at large is also that the trouble is not the definitive sort, making it possible for those who make a living on keeping your money to insist that there is nothing at all wrong, that the news is just an aberration and you really need to put more money into our fund, sir. Brexit, shmexit.

Let's start with what constitutes trouble. I need hardly tell you that the last jobs report was an addition of a paltry 38,000 (38K), seasonally adjusted (SA). That number was reduced by a strike at Verizon (NYSE:VZ) that lopped an estimated 35K off the count.

But after we add back the 35,000 workers and get 73,000, or even call it a rounded (adjusted) 80K, that is still showing a distinct downward slope over the last few months: 186K in March (down from 233K in February), 123K in April, 80K in May. There's no weather to blame this year.

Wall Street can call it an aberration and many do, but the data would suggest otherwise. I like to slice and dice the employment data in various ways, and one statistic I watch in particular is what I call the annual payroll replacement rate, or replacement rate for short. What I mean by this is that every January, about 2% of the payroll count disappears from the raw count (not seasonally adjusted, or NSA). It's a separation issue due to a mix of retirement, layoffs, firings, and just plain not counted as workers go from one firm to another. In a good year, that 2% is made up by the end of April, though it sometimes drags into May (in a recession, it just keeps heading south).

Using the May payrolls data, the replacement rate (RR) through the first five months - that is, the net change in the raw count from the end of December through the end of May - was a gain of +0.33%. That is distinctly on the low side, down from 0.64% the year before and from +0.65% as the average over the previous five years. It is also ominously reminiscent of 2007, which sported a RR of 0.40% through May of that year, down from +0.75% the prior year. If we add back the 35,000 Verizon jobs, the RR only increases to 0.35%. Clearly something is amiss. The only equivalent drop in the last thirty years not to be followed by recession was in 1986, and even that was followed by a rough 1987.

Employment is a lagging indicator, one that lacks the good form of behaving exactly the same way in each cycle (wouldn't that be useful!). Certainly there are similarities, as in periods when jobs decline month after month: we call those recessions. However the monthly job declines usually begin after the recession has already started. The employment slowdown isn't only in the May report, or the first five months, though - it is also starting to show up in weekly claims data.

The claims data usually signal earlier than the monthly jobs figures, and one of the more timely indicator is a slowdown in the year-on-year comparisons. So long as week "X" claims are consistently running lower than the same week a year ago, things are still good. As the business cycle ends, however, the declines will start to flatten and disappear, and the weekly numbers will instead begin to show sporadic increases - small ones, but increases nonetheless - over the prior year. Six of the last twelve weeks - or since the end of the first quarter - have show increases over the prior year (using unadjusted data). By comparison, the same period in 2007 (the first twelve weeks of the second quarter) had seven increases out of twelve.

A similar phenomenon obtained in August of 2000. In the 2007 example, a recession followed six months later (the Bureau of Economic Analysis (BEA) dates December 2007 as the begin date of the last recession), while in 2000, one followed seven months later, the BEA start date being March 2001. And while neither the stock market nor the Fed are paying much attention to the central bank's own "Labor Market Conditions Index," it has been negative every month this year and just touched a six-year low going back to May 2009. This is not good news. There is a silver lining in that the deterioration tends to be slow, has been so thus far, and usually takes two or three quarters to come apart (though you should be wary of policy blunders that can speed everything up).

It's perilous to focus too much on any one data series, but others are already in recessionary territory, including business investment, wholesale sales, retail sales, and corporate profits. The U.S. economy has been running at stall-speed while the global economy has been very near it. Now we have Brexit stage right just as the fat lady is clearing her throat on stage left. Be very careful, for I fear it won't be much longer before she sings.

Disclosure: I/we have no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours.

I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.