Into the Home Stretch


“The end is where we start from.” – T.S. Eliot

Two weeks after Thanksgiving Friday, the S&P has moved a total of one half of one percent. At least it was up. The Apple-less Dow Jones index did a little better, moving up a little over one percent in the same period. All of the improvement came in the last two days, as both indices fell 2% from Monday morning, when the ISM manufacturing survey hit (see below), to Wednesday morning before turning to rally back.

Absent a surprise move in cliff negotiations, it will be a lot more difficult to pull off the same calm next week. A good-news, bad-news development from Friday morning was that the latest fiscal cliff press conference to discredit the other side during trading hours, in this case one involving Speaker Boehner, affected the markets for about ten minutes before prices recovered, in fact just as the conference finished (it must be said as well that the Speaker did not take a lower road in his remarks, which surely helped).

The downside of the market’s short-lived reaction to the latest proclamation that the other fella has doomed us all is that politicians may feel the need to raise the rhetorical ante in consequence as we get closer to the deadline. I’ll be surprised if the rhetoric doesn’t get dialed up for the Sunday morning political talk shows, and then continues that way through the rest of the week. The tight trading range of the last two weeks is ripe for testing.

As the media fills up with ever stronger assertions that a deal will or will not get done, it certainly isn’t a comfortable time. In some ways, it reminds me of the uncertainty heading into the last weekend of Lehman Brothers. Most market pros felt at the time that some sort of conveyance would be done to either keep the firm on life support, or dispose of the body in a controlled fashion, rather than simply pushing it off the roof to see what would happen. The surprise of the actual outcome gained strength from the belief that Lehman’s fate could be decided by a small group of government officials and corporate executives who could find a pragmatic solution without needing to resort to a lot of politicking. It didn’t work out that way, of course, but it should have.

There are big differences. One is that in September 2008, the economy was already in recession (though many were in denial about it), unemployment was rising, housing was crashing, and a Lehman salvage job would probably not have rallied the markets much, if at all. Since Lehman’s equity would surely have been wiped out, it would hardly have made investors feel better about buying into the rest of the sector. Today, any half-decent cliff deal or “mini-bargain,” even one that puts off many of the most contentious issues (all that is realistically possible with three weeks left in the year), would rally the markets, possibly quite sharply if sequestration is avoided.

Another difference is that there is little maneuvering room for the Fed this time. After Lehman’s crash, the Fed rescued AIG and moved to cut interest rates. The Dow fell 800 points in three days, then got nearly all of it back by the end of the week in the wake of the relief efforts (clearly the repercussions weren’t immediately obvious). This time around, the surprise factor will be less (we already went through this in August of 2011), but the heavily-extended Fed won’t be able to do much about stopping the bleeding.

The effects on the market could be quite transitory anyway, but every thinking person in the business is wondering if the Dow will have to fall another 800 points first so that politicians feel they have enough protective cover to act “for the good of the country” and hope their vote won’t be used against them. Many Republicans who voted for TARP in 2008 found their vote held against them two years later by the Tea Party movement. It’s a tricky calculation. If there is no deal by January 1st and massive layoff notices follow, though, it will be a lot less transitory effect than everyone had wished.

My unscientific impression is that sentiment is growing on the Street for the possibility of a cliff-jump of a few days duration, allowing for a “rescue” political script instead of the “sell-out” version. That leaves many wondering why the VIX “fear” index, which measures the volatility of S&P 500 index options, is so low. Why aren’t more managers hedging against a meltdown?

There are a couple of good reasons. One is that the next options expiration period is in two weeks. Since the prospect of both sides coming out in favor of total impasse with ten days still to go in the year (and just before Christmas) is more than a little remote, there isn’t much point in buying December options to hedge against the cliff.

The January option decision is trickier. To begin with, options decay with time, and short-dated options decay the fastest. December 26th is probably the real time to start worrying, and there’s a lot of premium in the 19 days until we get there. Why buy early?

Then there’s performance. The indices have been nearly impossible for fund managers to keep up with this year. If you were at all prudent and kept some cash against the possibility of a Greek default, or stupidly invested based on earnings rather than surprise central bank announcements, you are lagging the index along with about 80% of your peers. Buying a boatload of options and then watching them lose 75% of their value the second a deal gets announced isn’t going to help the situation, so most managers who can buy options but don’t do so regularly are going to wait until the last inning to buy some. The truth about the VIX is that it always starts to rise after the bad news has hit anyway.

There’s also the peculiarity of this year’s calendar: The last trading day of the year, December 31st, is a Monday. The notion of a last-weekend deal to avoid sequestration is more than a little plausible, which would translate into a big rally on the last day of the year and the possibility of a lot of option holders trying to dump them all at once. So despite the low VIX number, investors may not be quite as complacent as it seems.

Is there a first-mover advantage to special dividends? Last week I criticized the Costco (COST) $7 special payout as being more of a good deal for management (I was hardly alone in this, as most of the major business media covered the story in varying tones of skepticism), and sure enough the stock fell over $7 on its ex-dividend date. However, if you bought the stock in a non-taxable account in the days immediately following the announcement in the $100-$103 range (even $104 for a non-taxable account), you’d still have come out ahead, assuming you bought in enough quantity to overcome transaction costs.

Special dividends continue to proliferate. It seems to me that most companies have been unable to duplicate Costco’s feat of managing to keep a gain if you bought right away, making the taxable event worthwhile. I wonder what PT Barnum would say.

The Economic Beat

Nothing has changed here.” That was Moodys.com economist Mark Zandi’s initial conclusion about the November jobs report, and it looks about right. The seasonally-adjusted total of 146,000 non-farm payroll additions handily beat expectations for something under 100,000, and was in line with the roughly 150K average of recent months.

Things to like about the report included an unemployment rate of 7.7%, the lowest since the crisis (7.8% in January 2009 – the last pre-crisis number of 6.1% is still far off), and of course lower than most of Europe, whose rates are headed in the opposite direction. The U-6 alternative rate also fell two-tenths to 14.4% seasonally adjusted, 13.9% unadjusted, both numbers being the lowest since December 2008.

The ADP payrolls figure two days earlier was 118,000 new jobs, a little below consensus of 125,000. But the same Mark Zandi related on Wednesday that he and ADP had worked together to provide an estimate of a loss of 86,000 jobs from Hurricane Sandy. The Labor Department, though, found that survey response rates were “within normal ranges” and that Sandy “did not substantively impact” the data, even though construction jobs fell by 20K (and October construction spending increased).

It’s something of a guess as to who is closer to the truth. Zandi also immediately concluded that that this month’s data would be revised, and indeed the previous two months were revised for a net loss of 49K jobs, ending a string of positive revisions. The loss was mostly from the government sector, which reported a surprise additional 38,000 jobs lost in October. It’s unusual to see a revision that large on the government side.

Given the ensemble, it seems reasonable to conclude that BLS may have overestimated the reported number and underestimated the Sandy effect. The state data two weeks from now may give a clue, but for now it still translates into no change in the trend. It’s also possible that retail hiring for the early Thanksgiving offset Sandy losses, implying an even better number in the absence of the storm. Temp hiring increased for a second month in a row, and the weekly payroll index increased to 108.7, suggesting that personal income might have picked up around 0.3% as well in November.

That the job market isn’t robust, I hardly need tell you, and there were things to complain about in the report as well. Average weekly hours were unchanged. The number of additional people not in the labor force (2.4mm) over the last year is rising much faster than the number of non-farm employees (1.3mm). Presumably they are not all working on farms somewhere. The participation rate fell two-tenths to 63.6%, and the employment-population ratio fell a tenth to 58.7%, still two-tenths better than a year ago. The year-on-year growth increase in average weekly earnings is a meager 1.7%. The household survey reported a seasonally adjusted decline of 122,000 jobs.

Weekly jobs claims also support the view that the employment trend is unchanged. The number has fallen sharply in the weeks since the Sandy spike, and yesterday’s seasonally adjusted rate of 370K is back below the last pre-Sandy figure of 372K. Though the former figure will probably be bumped up another 2K-3K next week, that would still leave it about even. The year-on-year claims declines are also very nearly back to trend

If you want to put a bullish spin on the economy, you could point to the steady employment growth and the ISM non-manufacturing survey result of 54.7, not only better than expectations but – as I pointed out a couple days earlier in Seeking Alpha – including a business activity rate of 61.2, the highest November reading since 2004. Construction spending reportedly rose by 1.4% in October. Though it’s a number prone to big revisions, it seems reasonably certain that a good-sized increase occurred. November car sales rose strongly, partly as a rebound from Sandy (though apart from Ford (F), the Japanese did better than the Americans at capturing the increase).

A bearish posture would note that employment is a lagging indicator. That is so, but the trend usually isn’t. The unemployment rate increased sharply at the end of 2007 and steadily rose all the way into the crisis. A bearish view would also cite the ISM manufacturing survey, which turned in a surprise below-50 reading of 49.5 on Monday (earlier estimates had been over 52, 50 being neutral). Manufacturing is a much smaller part of the economy, unfortunately, making up just under 9% of employees, but it is the more cyclically sensitive sector.

A study of the underlying data supports the view that the manufacturing number was not a distortion, that Sandy wasn’t really to blame, and that the fiscal cliff is much more the problem. That would seem to be borne out by the data for October factory orders, whose increase of 0.8% was better than the expected decline of (-0.1%). As we get closer to the deadline with no resolution, people who sign off on spending are simply putting down their pens.

The pause-thesis is further borne out by the consumer sentiment index, which revealed a steep plunge in future expectations, pulling the headline number down to 74.5 from the previous 82.7. The current conditions component, though, had a much more mild decline to a still quite elevated 89.9. The ECRI Weekly Leading Index rose to 126.8 from 126.3.

Besides the expected increase in gamesmanship next week, we will also see the November retail sales report on Thursday. While car sales definitely improved, the ex-auto ex-gas number is tougher to handicap, both because of the effects of Sandy, and from the noticeable spreading out of retail promotions in the wake of Thanksgiving. The industrial production number on Friday will face the cross-currents of production rebounding from Sandy and recoiling from the cliff.

The Fed’s monetary policy meeting is next week, and the FOMC will announce its latest assessment on Wednesday at 12:30, with forecasts to follow at 2:30. While these things always rattle the tape for an hour or day, I don’t expect the markets to see anything shocking enough to react for long. Some additional Treasury bond buying to offset Operation Twist is more or less expected, and there isn’t much reason to suppose a dramatic change is in the offing – or that the additional buying will have much effect. What we really need is first the cliff resolution, then a European resolution. The latter’s central bank, the ECB, cut its economic forecasts for 2012 and 2013 on Thursday.

From the wonk point of view, the most interesting data may come Tuesday in the form of the monthly trade report and wholesales sales and inventories. A trade decline may provide some further ammunition for foes of jumping off the cliff, but one must measure the reality of the numbers versus Schopenhauer’s old admonition that “against stupidity, the gods themselves contend in vain.” The EU has a council meeting on Thursday and Friday. Keep your fingers crossed.

share this blog post
Facebook Twitter Plusone Linkedin Digg Delicious Reddit Stumbleupon Tumblr Posterous