A Look Under The Hood Of The Jobs Data

by: Kevin Flynn, CFA

It may seem tempting to say "we told you so," because we've been saying for weeks that while the economy is growing, it isn't growing as fast as the press clippings would have you believe. The March jobs report, which we examine in more detail below, would seem to a pretty solid brick in our wall.

But we refrain from gloating for a number of reasons. The first is that it doesn't exactly cheer us that we are stuck in a trend of soft growth, especially where employment is concerned. In the second place, the ISM non-manufacturing report had a decent uptick in the number of respondents saying that they were hiring "more." It wasn't so large a number as to indicate a barn-burner, but it did suggest a reasonable chance that the February trend could hold on, and we posited thusly last week in "Seeking Alpha".

The problem with diffusion surveys such as the ISM, though, is that they don't say anything about the level of activity. If two firms plan to hire ten more people and a third plans to sack a thousand, the diffusion survey will tell you that hiring is directionally robust. Or a firm might hire a hundred workers in January, fifty in February and then five more in March, but the answer to each month's survey would only be, "more."

And last but not least, it is April, the best month of the year for the stock market. It's commonplace for part of the April rally to be borrowed in March, such that subsequent stumbles in the first part of the month are practically routine. But the market - especially a trader's market like the one we have - will always try to find a way to have the rally that it believes it's entitled to.

Recent economic data coming out of Europe has been disappointing. The recession there is deepening, and the troubles of the periphery countries are moving back onto the front pages. Spain is a train wreck in slow motion: last November, we predicted it would end up being Europe's "Lehman moment."

Despite these problems, though, we doubt that a substantial market reset will happen quite yet. Sometime in the next couple of weeks, traders will find another rallying cry. The obvious candidate is of course QE-3, or round three of quantitative easing by the Federal Reserve. The Fed cannot head off the European recession, nor can the ECB, but all it took was a remark by ECB governor Benoit Coeure that the bank might buy some bonds again to send equities flying upward again (no, it wasn't Alcoa's (AA) earnings).

Back in the U.S., the non-farm payroll number of 120,000 on Friday was roughly half of consensus and below the bottom of the entire range of estimates. That's bad enough, and had stocks down triple digits for a couple of days. The underlying details were no treat either.

Not all of the report was bad - the unemployment rate fell a tenth, and the "under-employment," or U-6 rate, fell a bigger four-tenths to 14.5%. Unfortunately, the household survey reported that the number of people working actually fell, seasonally adjusted: the unemployment rate fell because people dropped out of the labor force. Those are only first estimates, and in particular the household ones are subject to large revisions. But they aren't good news.

Temp hiring, which is a leading indicator, fell. Department store jobs fell for a second straight month, lending additional weight to the warm-weather argument that claimed jobs and sales were being pulled forward from latter months. We suspect that the number of additions were overstated, the number of departures similarly overstated, and we are now seeing the netting out.

Average weekly hours ticked down, and the weekly aggregate payroll index - which tends to track personal income - had its smallest gain in months. There is no hiring pressure nationally, though there is some tightness in select specialties and regions. We read a self-styled "bright spot" noting that hourly wages were up, albeit barely, without mentioning that average weekly wages - i.e., the paycheck - ticked down (also barely) with the drop in hours.

It isn't disaster, but it isn't "escape velocity," nor "gaining momentum" nor the other folderol being splashed around by the media in recent weeks. It's a slow economy that got a little lift from some inventory accumulation, a little rouge from the warm weather, and is now returning to its dowdier self.

You may have also read that the weekly jobless claims were the lowest since April of 2008. It was really May of 2008, but that's not much of a quibble. The observation helped the markets rebound a bit in the face of disappointing data from Europe. So if claims were so good, what's up with the jobs report?

A quick answer might be that the last couple of weeks fell after the end of the jobs report measurement period - maybe the report was a fluke? But we don't think so, for a couple of good reasons. The first would be the year-over-year change in monthly claims. It's a relatively decent leading indicator, and the data show that the decreases in claims have been slowing. The year-on-year changes (about 150k for February, 140k for March) suggest an employment market in the early stages of flattening out.

The second reason is the number of actual jobs - that is, unadjusted - and here's where we think the weather has played a role in flattering some of the data. The week before, we watched a fund manager - a creature bullish by definition - reply rather smugly that what with the economy having "created almost 800,000" jobs this year, things were obviously in good shape. If only it were true.

The manager was probably betting on a good March report that would put the quarterly total in the neighborhood of 750,000, maybe even higher with a favorable revision or two. However, that is not the way the economy works. Businesses run on quarters and years, mostly calendar ones. The shopping season of Christmas comes in the last month of the year, adding a surge in seasonal help. The effect of this is that in unadjusted terms, the January workforce is, on average, about 2% smaller than December's. In absolute terms, that comes out to about 2.5 - 3 million jobs lost every January, as contracts end, seasonal workers get let go, and workforces get trimmed.

Not all of the jobs are really lost - many people simply change jobs with an offer in hand, and the January count can't accurately capture the turnover. Excluding recession years, roughly a third of those losses are recaptured by February, and roughly another third is recaptured by March. Going back to 1980, at least, the number of actual jobs is nonetheless always smaller in March than December, by about half of the January loss. In a normal year, the progression continues until June. By then the actual headcount has (usually) grown larger than the previous December; then the mid-year axe falls, and there is another drop-off.

The seasonal factors do a good job of smoothing out these fluctuations and presenting a more accurate picture of underlying trend. Better, for example, than reporting a loss of 2.5mm or so jobs every January. However, it can be worth the time to study the changes in the raw data. In the current case of 2012, the January decrease was below average, the smallest in percentage terms since January 2006 (don't get excited, though, as 2012 was the same percentage decrease as the bubble-peak of January 2000, and a really low number is a sign of an overheated economy). The February increase was above average - that's good - and the March increase about average. The intriguing bit is that the March increase was smaller than February (and should stay that way regardless of revisions). That doesn't often happen - the last time was in 2001.

It isn't a recession indicator. However, we do think that it's another strong piece of evidence that says preceding jobs data were somewhat flattered by the warm weather. Fewer people were let go, particularly in retail sales, when warm-weather apparel sales were apparently pulled forward. Usually it isn't easy in the northern half of the country to go out and shop throughout January and February, but it was this year. Fewer work days were lost, too. So it's not quite that the trend is reversing, but that a smaller-than-usual decline is being balanced by a smaller-than-usual rebound.

There are pockets of relative strength: manufacturing has been adding jobs steadily. Unfortunately, it's a much smaller sector than it was twenty-five years ago. Health care is a steady grower, and leisure and hospitality have been additive of late as well, though one has to worry there too about a rebound effect.

So we haven't really added 800k jobs yet this year, nor the 635k BLS to-date seasonal total, nor finished recovering the January loss. Not yet anyway, not so far as can be counted. What the seasonal totals tell us so far (in case you're wondering, everything does balance out by the end of the year) is that based upon historical patterns, that would be the underlying trend rate.

Neither we nor the Labor Department, though, know yet how the warm weather has wiggled the trend, and we won't really know until the second half of the year, probably September. But the March report is a strong suggestion - we won't say it's definitive quite yet - that the wiggle was real, and lines up with the nearly-forgotten February assertions that the warm weather was giving us a tilted view of the economy. The NFIB small-business optimism index dipped unexpectedly in March, all the more surprising in light of the month's rising stock market.

But it's still April. Coming back to that Alcoa report, we got a barrage of notifications from our CNBC "breaking news" app that the company had beaten estimates. It was a curious thing, as we'd read on the weekend that the company was expected to earn two cents a share. When the results broke, the consensus was instead stated as a loss of two cents. Fast work. The next morning, Bloomberg television avowed that consensus had been for a loss of four cents a share. Faster work. Very little was made of the fact that the year-on-year earnings comparison was minus 69%.

Between Thursday's close and Friday's open, things are going to get more interesting: Google (GOOG), Wells Fargo (WFC) and JP Morgan (JPM) are all reporting, and the market cares about them a lot more than Alcoa. We're not so sure about Google, but beats by the two banks should be a complete lay-up: last week we observed that first quarter earnings growth expectations had been lowered to 5%, but now they are down to 3.3%.

While the bank earnings may not actually be up year-on-year - that would be greedy - my crystal ball tells me another season of good old "earnings beats" has been underwritten by companies and the Street. Yes, Friday should be interesting indeed, because in addition to said earnings, the really big numbers of the week are coming Thursday night, from China: first-quarter GDP, along with monthly industrial production and retail sales. It'll be interesting to see what the authorities have decided to report.

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