The monthly employment situation report remains the "A list" economic indicator. Pundits love it. There are so many facets. If you want to spin, opportunities abound.
The official data is the best effort of capable and talented people. They have created a methodology. When new data are available, the BLS revises the result.
The result? A happy hunting ground for critics. Since so few understand the methodology, and so many have a deep suspicion of government, the bearish spinners have the upper hand. Let's start with how I approach the problem.
As regular readers know, I have a unique viewpoint on the payroll report. I wish I could get more professional colleagues to join me.
There is an actual job change which we might call "TRUTH." There are many efforts to measure that change, including that of the BLS. While the BLS does a good job, they are hampered by their approach, including the following:
- They do not focus directly on the job change. They attempt to count every job for two different months and subtract one from another.
- They rely on a survey. This is a fine method for identifying properties of the surveyed population. It is much more troublesome when the objective is to COUNT the population.
- The survey has non-respondents who might be either firms that did not respond, or firms that are out of business.
- The survey is forced to impute business births, because there are always many business births. The BLS does it pretty well, but it is inherently difficult.
- The survey has a sampling error of more than 100K jobs.
To summarize, the BLS result is important not because it is the best measure, but because we are so interested in employment.
Meanwhile, many independent analysts have other approaches. A fair method of scoring would be to look at the final job changes, after all of the revisions, and see whether anyone did better than the BLS. Trying to guess the monthly "as reported" data is a different problem. Nonetheless, billions of dollars will trade on that information.
Our Own Estimate
Each month we ask the question, "What change in payroll employment would be consistent with other economic data from the same time period" (the middle of the prior month)?
This is not a forecast, per se, since we do not posit any causal relationship among these variables. They are all concomitant indicators of economic activity.
- We use the four-week moving average of initial unemployment claims, culminating in the week of the employment survey. This is the best direct indicator of new job losses. This has gotten worse in the last month--- 468K versus 447K. Ignore the recent weeks which are not in the survey period.
- We look at the University of Michigan sentiment survey, which we find to be more useful than the Conference Board's sentiment index. Michigan uses a panel, where some families are carried over from month to month. This is a good technique. Sentiment is influenced by employment. When people have lost jobs, or are worried about losing jobs, it shows up in sentiment. It is a good concurrent indicator. The Michigan index is now at 73.6, down a little from last month.
- We us the ISM manufacturing index, which came in at 56.5, down from last month's surge to 58.4. This is still a strongly bullish for the overall economy.
Our long-term record has been pretty good, especially when compared to the final revised data. This makes sense because our model was derived from the final data. In recent months we have been too bearish. The BLS benchmark revisions suggest that we have been much better than first thought. I am working on a comparison with the final numbers.
This month, our estimate is for a net job loss of 51,000.
It is always interesting to compare the job forecasts from different sources. We follow several because of the interesting and widely varying methods they use. A wise interpretation would be to consider all of these disparate sources of information.
- ADP has proprietary data because of its payroll management business. ADP sees losses of 20K. This estimate does not include government jobs.
- TrimTabs has another valuable approach -- tax deposits. Their forecast is for a loss of 30K jobs.
- WANTED Technologies has resumed estimating the monthly change after a protest because of the benchmark revisions. I sympathize with the problem, but I am glad to see them back in action with their unique method. They bring information from online ads, a source that others miss. They see a job gain of 5000.
- Briefing.com cites the consensus as a loss of 20K and their own forecast is a loss of 50K. I get several other private estimates with similar methods -- all in this range.
All of these sources are valuable. The 90% confidence interval on the BLS estimate, something that no mainstream media sources report, is +/- 100K or so. And that is after revisions and benchmarking. It is a survey -- a good one -- but it has an error band.
Investment and Trading Take
I have frequently recommended being short in front of the payroll number. This month there has been plenty of advance information about possible weakness. It is pretty silly. There will be discussion about weather factors. Anyone who did not work the week including the 12th was not on a payroll. Meanwhile, people ignore the large sampling error.
As a result, the data may draw less attention than usual. Meanwhile, our read on the employment growth is somewhat bearish. The jobs are not adding up as rapidly as we might hope.
A Final Thought
There are so many, measuring employment change in so many ways. This month there is a strong consensus. If the BLS number is at wide variance, it is best to put it aside and await more data and more revisions.