Pinning down the January effect on U.S. jobs figures

February 7, 2014

With Wall Street grappling to hold on to its record highs, a lot is riding on good news from the U.S. economy, no matter how high the Federal Reserve has set the bar for backing off its clear plan to end its monetary stimulus program this year.

After two huge upsets in a row on the important U.S. economic data releases since Christmas — December non-farm payrolls and the January ISM manufacturing report, forecasters are lining up again for an improvement in hiring.

The latest consensus from Reuters Polls is for a rebound to 185,000 after net hiring collapsed to just 74,000 the month before.

The new figure is surprisingly similar to the last consensus of 196,000 – giving the impression forecasters are hoping the December figures were a blip and are sticking to their usual tendency since the Great Recession began to predict  that the economy will soon revert back to the trend it was at before the collapse of Lehman Brothers in September 2008.

Only a handful of economists have dared to revise down their forecast after a disappointment on the ADP payrolls data from Wednesday. That suggests most have learned the lesson that disappointment or surprise on ADP – it was much stronger than expected in December – gives no clear steer one way or the other on what will happen with the official jobs data that come out a few days later.

Joseph LaVorgna at Deutsche Bank is one of the few economists who is still looking to ADP for cues.

Our analysis has found the ADP report to be the single best predictor—albeit a flawed one among a handful of less than perfect indicators—of monthly changes in nonfarm payrolls. Hence, we are carrying the relatively small -25k forecast miss on ADP over to our estimate of January nonfarm payrolls, which is now +175k versus +200k previously.

Never mind that almost all the other analysis concludes otherwise.

Deutsche Bank and others have warned clients that the January data are likely to be affected by weather patterns across the United States in January. Extremely cold weather was one of the explanations offered for the biggest growth slowdown in factory new orders since 1980.

But the problem is that there was unseasonably cold weather in December as well. And, as everyone knows, it gets cold in most of the United States around this time of year.

Taking a look at the most optimistic and pessimistic forecasters in the Reuters Polls sample may reveal more.

First Trust Advisors, a Chicago-based portfolio investment firm that has consistently had an upbeat view on the U.S. economy over most of the last decade, surprisingly had the most pessimistic forecast for the December payroll numbers, at 120,000. They turned out to be the most accurate. This time around they are again almost at the bottom of the range, calling for a 130,000 job gain.

Societe Generale, on the other hand, is looking for a blowout month for hiring, first forecasting 290,000 jobs, revising this estimate down to 270,000 after the ADP report. They were one of the least accurate forecasters last time, having called for 225,000.

Weather-related and fundamental factors support our expectation for a marked pickup in job creation last month.

But unlike the cold weather that was used to explain away a dismal report on manufacturing, Brian Jones at SocGen appears to be betting on higher temperatures.

Figures collected by the National Oceanic and Atmospheric Administration (NOAA) revealed that at 32.9°F, temperatures nationwide were 2.1°F warmer than usual over the month-to-date ended January 18 – a fairly dramatic improvement from the mean 27.8°F (5.6°F below normal) suffered over the first two weeks of December.

Our statistical work revealed that the 7.7°F positive swing in the month-to-survey-period temperature anomaly – the biggest witnessed over the past six years – could add roughly 60,000 jobs to the January report.

SocGen also appears to be betting that the December payrolls disappointment was a one-off:

Traditional labour market barometers also suggest that December’s surprisingly low print was a fluke.

That may be, but the weather in the United States in winter is rarely a fluke – it’s usually just cold.

Which brings us to another set of conventional wisdom about January that is widely held although not well-explained.

Many on Wall Street say that how Wall Street trades during the month of January sets the tone for the rest of the year. It’s a convenient theory to hold, along with the “Santa Claus Rally” as a way to get people to believe that stocks will only go in one direction.

When they don’t, the theory becomes a lot more difficult to, well, sell.

Brian Wesbury, chief economist at FT Advisors, whatever his view on how cold it was in Chicago this January, is having nothing of the “January effect” theory. But not for the reason you might think.

Lots of people believe that what the stock market does in January is a signal for what it will do all year long. I believe this whole theory is bogus. It’s not true; it’s a mistake to believe in it. And let’s just think about this logically. There’s 12 months in a year, and yes, January is the first month. But why would one month signal what the next 11 months are going to do – it just doesn’t make any sense.

Stocks are cheap right now. Profit margins are very wide. Over 70 percent of companies have beaten their earnings estimate for the first quarter of 2014. We expect earnings up, we expect the economy to grow. So even though the market was down in January do not look at this as a signal of what will happen in the rest of this year.

With a stock market view that bullish and a non-farm payrolls forecast that modest, conventional wisdom might call for caution on the January data.

No comments so far

We welcome comments that advance the story through relevant opinion, anecdotes, links and data. If you see a comment that you believe is irrelevant or inappropriate, you can flag it to our editors by using the report abuse links. Views expressed in the comments do not represent those of Reuters. For more information on our comment policy, see