A Time For Caution

Includes: DIA, IWM, SPY, TZA
by: Kevin Flynn, CFA

The September jobs report was like the quiet dog of the Sherlock Holmes story - it did not bark. The question for us is whether or not this means anything at all, or is in fact according-to-Doyle significant (I beg your grace for the pun).

I think that there is something in it, though not in the vein of the sweeping pronouncements that tend to follow these reports - e.g., the Fed must cease, the Fed must keep going, it's time to flee, it's time to go all in, it's time to board the mothership. The reported number of 148,000 did fall into what has become an increasingly wide "Goldilocks" zone: Though the consensus was something in the 170K-180K range, a number below that is still taken as good enough for the semi-automatic jobs rally. It's semi-automatic because a weak number ensures more Fed bond-buying, while a number above consensus ensures the recovery is on its way. It's win-win, unless the number is an extreme outlier.

Most analysts picked up on the weakish components of the report, of which there was a goodly supply. The unemployment rate fell because the not-in-labor-force (NILF) population grew faster (+0.15%) than the civilian labor force (+0.05%), a factor that also keeps the U-6 alternative rate down (13.6%). Hourly earnings growth was miniscule at (+0.1%) for the month and +2.1% for the trailing twelve months. That's nominal - in real terms, after-tax hourly earnings growth is negative when inflation and the restored payroll tax are taken into account.

One area that caught my eye is manufacturing. The oft-ballyhooed renaissance in American manufacturing has resulted in a 0.2% increase in sector employment over the last year; that isn't much. The oil and gas extraction industry, the other savior, is up 4.8% in jobs over the last year, which is good, but 9,000 extra jobs aren't going to take the American economy to a new level.

There's always a danger of making too much of one month's data; despite the slowdown, the numbers don't look significantly different to me. Through the first 9 months of 2012, the number of jobs had increased by 0.81%, while so far this year, the growth is 0.78%. Very close, and a difference that might eventually get revised away when the entire year is revised in the annual February benchmarking.

The three-month jobs average has slowed, but more interesting to me is that the monthly year-over-year increase has been remarkably stable of late. July was up 1.66% from the previous July; August was up 1.67% and September 1.66%. I have suggested several times during the year that the close tracking may be due in part to changed estimation methods at the BLS, designed to reduce outsized benchmark revisions.

Another aspect that intrigues is that history implies that such extended periods of stability are the last stages of an expansion. The current episode hasn't had a precedent since the period from the end of 1998 to mid-2000. I'm not ready to stake a claim to any new laws of economics, but the notion that employment growth would plateau for an extended period in a business cycle's last leg does have an intuitive appeal.

Despite the reflex chat in the media about inevitable stimulus leading to inevitable rallies, my perception is that the stimulus angle is getting a little worn out in the marketplace. There's still loyalty, to be sure, but some of the magic is gone. At the same time, though Tuesday's pullback had the effect of keeping the market out of seriously overbought territory in the very short term, it is nevertheless overbought on both an intermediate and long-term basis, the latter significantly so. The Russell 2000 is sufficiently overdone that I have started nibbling at the TZA ETF (a security designed for short holding periods; not for the unfamiliar).

Add to that posture the fact that earnings season has been weak overall and fourth quarter guidance is coming down rapidly, and we are set up for a slow fade all the way to Thanksgiving. The three-day package around the turn of the month - the Fed announcement on the 30th, followed by the first-day-of-the-month trade two days later - should be good for a percent or two to the green, but otherwise I see a correction in the low-to-mid single digits ensuing over the next few weeks.

The data isn't likely to help. I had hoped that retail sales data would be available for this column, but it's been delayed until the 29th. I'm concerned about the snapshots provided by the weekly chain store reports, and suspect some rather barren figures lie in wait for us, beginning next week. Weak jobs data gets a pass courtesy of the Fed stimulus-link, but not so weak retail sales data - for one thing, the latter suggests pressure on corporate earnings. We can and will excuse part of the weakness on the Washington circus, but soft data is more of an excuse to take profits than it is to put on the buying shoes.

Though I believe a retreat is in store, I don't see anything immediately serious enough to stave off the Thanksgiving-to-New Year's rally that has practically become a stock market law. Still, be careful about jumping too eagerly into the first stages of a dip - the economy is anemic, sometimes something bad does happen, and mild pullbacks can become much bigger ones.

Disclosure: I am long TZA. 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.