“When the shooting starts, climb on your camel and head for the mountaintops.” – Bedouin Proverb
As we begin the final month of the third quarter, we also enter one of the more volatile times of year for the stock market. September and October are quirky months, characterized by euphoric tops, sharp corrections and the occasional crash. It’s a seasonal effect, partly induced by companies getting closer to owning up on the year, partly induced by a general trading awareness of the seasonal effect, and all of it exaggerated in recent years by the dominance of computerized trading.
The asset markets are indeed jumpy now, ready to bolt in either direction. I need hardly remind you that it’s been a sentiment-driven market this year, with the bulk of the gains over the last year coming from the rally that ended in the spring, fed by central bank easing, the debt-battle deferral and relentless momentum trading. The S&P is up six of the last seven days, yet remains below Taper Day (May 22nd), when a somewhat alarmed central bank reminded markets that it would not buy $1 trillion of bonds every year until the end of time.
Now we are confronting two highly sensitized situations at a highly sensitive time of year, with no sure direction. The poles of opinion around what the Fed might do seemed to harden after the August jobs report, with both sides more convinced than ever that the Fed was inevitably on track to announce in less than two weeks’ time (the 18th) that it would either begin or defer a proposed program to slow the pace of bond purchases.
A great deal of division remains over the efficacy of the QE program. That it has driven up asset prices is a matter of near-universal agreement, while its effect on the economy has been hotly debated and will be for decades to come. It does seem fair to say that the current consensus amongst economists has gradually come around to the thought that whatever economic effect it does have is waning, a notion fueled by studies from the Fed’s own regional banks.
Yet the import of the September meeting has not waned. In recent days, political leaders around the world have been practically beseeching our central bank to go slow, and our own asset markets have been plainly rattled by the idea. The money river is caught up in doubt. Whatever the FOMC (monetary policy committee) thinks today could be altered before the meeting by the situation in the Middle East, the August retail sales report, or Obama letting Bernanke know who he favors as successor. Guessing policy decisions is brutal, and I have given up trying to make any living from it.
If the Fed postpones tapering, if Syria stands down and offers apologies, markets could soar for a month. Different outcomes could lead to a prolonged sell-off. The main actors are aware of this, adding to the stress of their decision-making (and the odds that they get it wrong, if you ask me).
Looking a bit further ahead isn’t of much comfort. The current bull market is now 54 months old and the recovery has clearly matured. The catalyst for the economy’s purported acceleration next year remains something of a mystery to me. The recovery in home and auto sales has been gratifying, but the growth rate for both is going to be sharply lower next year. In the interim, the economy will have to negotiate the reality that stock prices are increasingly vulnerable to a pullback. The debt battle remains ahead of us and could get uglier if Obama seems wounded politically by Syria. It is very difficult to find cheap asset prices for anything.
The one genuine potential catalyst I see for the global economy comes from the chance of Europe finally confronting its debt situation and engaging in a genuine restructuring. That would involve “burden-sharing,” as the Germans call it, and it doesn’t seem any more likely now than before that the EU will achieve the grand bargain without the kind of serious crisis that forces hard choices.
Still, there are two more official weeks of summer left and the weather has been fine; one can always hope for the best.
The Economic Beat
The doubts I expressed in last week’s column about the August jobs report were partly substantiated by an eccentric mix of positives and negatives. It was essentially weak enough to reawaken hope in the markets that the Fed would maintain its current policy for at least another meeting – possibly until December, when the next press conference is scheduled. Even NPR (public radio) reported it as “soft.”
The good news was that the initial estimate was a gain of 169,000 jobs, near the low end of the consensus range but still within it, and well above the misgivings I had last week for a number closer to 120,000. The unemployment rate fell to 7.3%, average weekly hours ticked up a tenth and the payroll index was up 0.6% on the month, which is a plus for personal income. The estimated count of officially unemployed fell by 198,000.
As far as the drawbacks go, there were plenty, beginning with the downward revisions to June and July. Many veteran traders take their cue more from the direction of revisions than the initial estimate, and the one to July was a very substantial (-58,000), taking the first revised count all the way down to 102,000. Given the sharp upward arc of the stock market rally in the first three weeks of July, a move partly based on the notion of a Goldilocks economy, one has to wonder where the markets would be today if 102K had hit the tape two months ago – and what the Fed might have chosen to say instead in the aftermath.
I’ll return to the revision implications momentarily, but before I do it should also be acknowledged that the drop in the unemployment rate was entirely due to people dropping out of the labor force, as the participation rate fell to 63.2% of the civilian work force. That’s the lowest it’s been since 1978. The number of people working declined by 115K, according to the household survey that calculates the unemployment rate.
There are often wide discrepancies between the estimates that come from the establishment payroll survey, which produces the headline print of new jobs that you see (after seasonal adjustments), and the household survey. Over time, the two converge, and it would be rash to say one survey is right and the other is wrong on the heels of any month’s report.
Still, there are always plausible inferences to be drawn. For example, the number of full-time workers in the household survey was provisionally flat, as the survey’s net decline neatly lined up with the decline in part-time workers. The number of “job leavers” declined by nearly 10%, and the two together fit in with the gradual progress being made in the weekly initial jobless claims survey. The latter also fits the lack of progress in the JOLTS (labor turnover) survey, which has indicated no headway at all in the readiness of workers to change jobs or quit.
The fluctuation in manufacturing jobs – a decline of 16K in July followed by an increase of 14K in August – can be readily explained by the auto industry’s annual shutdown, as observers were quick to note. The auto sector reported a loss of 10.4K workers in July and a gain of 18K in August, and while both numbers are apt to be revised, both the direction and the gap are likely to remain largely intact. The downside is that it may well be that the monthly increase in payrolls and therefore income was a one-off due to the auto restart, and September will return to the frustratingly slow pace of 2 to 2.5% annual growth in nominal terms, leaving real disposable income growth somewhere south of the 1% line (and vulnerable to shocks).
The year’s jobs data is still preliminary, but does point to some worrisome trends. Manufacturing employment started out 2013 with year-on-year growth rates of 1%, but in the last three months it’s collapsed to 0.07%, which may as well be zero. It’s another data series that has slowed quite a bit in the last few months. The three-month moving average for business investment spending, for example, as represented by the cleverly short name of “new orders for non-defense capital goods excluding aircraft,” has fallen to (-2.45%) in unadjusted terms, the lowest since August of last year (the data are unadjusted), when fears of the debt and budget battle first slowed investment.
As usual, the bulk of jobs added were at the bottom of the pay scale – retail; health care & social assistance; leisure & hospitality (which suffered a sizable downward revision to July). Collectively they made up two-thirds of the additions in the private sector. At 0.54%, the twelve-month growth in the labor force is little better than half the rate of the population (1%), with the difference being discarded into the NILF (not in labor force) category growing at 1.7%. That’s where nearly all of the improvement in the official rate is coming from – Gallup’s employment poll, which is admittedly only a poll, came up with a higher unemployment rate for last month (8.7%) than August 2012 (8.1%). And yet the shibboleth persists that US growth is firing up global growth.
On a final note, the ADP (176K) and Bureau of Labor Services (BLS) reports were closely in synch this month, though the former’s revision to July was barely noticeable. The BLS year-on-year growth rate dropped from a year-best 1.7% in July to 1.66%, very much in-line with July 2012′s rate of 1.67%. Since the benchmark revision in February, job growth this year has run remarkably similar to 2012, with the first 8 months of 2012 showing an unadjusted increase of 0.35% compared to this year’s 0.31%. It remains to be seen whether August will get a big revision upward, as many are hoping for. It isn’t unusual, but would be counter to the trend of the downward revisions of the last two months. At any rate, the FOMC won’t have it 12 days from now.
The ISM or purchasing manager surveys, by contrast, had largely positive news. The August result for manufacturing was 55.7, about the same as July’s 55.4, but the non-manufacturing survey moved significantly higher to 58.6 from 55.0, with the business activity index hitting a recovery high of 62.2. The latter result is cheering, though it ought to be said that the non-manufacturing survey is no leading indicator.
The manufacturing survey is much more economically sensitive, but the market reaction was somewhat restrained by the subdued comments of the responders. July’s virtually identical result, which included spikes in production and new orders, failed to translate into output strength for either industrial production (unchanged), new orders for durable goods (negative) or new orders for business capital goods (-4%, according to Thursday’s factory orders report).
The best positive was that construction spending for June was revised upward from (-0.6%) to unchanged, while July reported an above-consensus gain of 0.6%. Even so, the pace of year-on-year growth continues to fall. Imports zigged up more than exports in July, but trade remains soft, with both categories very nearly unchanged from the year-ago period.
Next week has some interesting tidbits for data wonks such as myself, beginning with the JOLTS survey report on Tuesday and wholesale trade data for July on Wednesday. Import-export prices come out Thursday, followed by producer-prices on Friday, but the market mover next week, Syria aside, will be August retail sales on Friday. One weekly retail survey showed a soft August, another was strong; auto sales were good, but they seem to be cutting into other purchases; same-store sales reports were uninspiring, but the sample size is quite small.
There could be some market-moving news out of China over the next few days. The country reports August imports and exports over the weekend, followed by industrial production and retail sales on Monday night. It’s remarkable that so large of a country can report all of this information so quickly after the end of the month. Some might say that it’s even more remarkable that they bother waiting a week to tell us what the numbers are, but that might be construed as cynical.