The eagerly awaited jobs report disappointed, but the market shrugged off the news. As I expected, there was too much noise this month, for a wide variety of reasons.
Few believe the jobs picture is good, but we need more data to get a better read. I'll look at this topic in more detail, but first let us do our regular weekly review.
Background on "Weighing the Week Ahead"
There are many good services that do a complete list of every event for the upcoming week, so that is not my mission. Instead, I try to single out what will be most important in the coming week. If I am correct, my theme for the week is what we will be watching on TV and reading in the mainstream media. It is a focus on what I think is important for my trading and client portfolios.
In most of my articles I build a careful case for each point. My purpose here is different. This weekly piece emphasizes my opinions about what is really important and how to put the news in context. I have had great success with my approach, but some will disagree. That is what makes a market!
Last Week's Data
It was an active week for eonomic data and earnings, and the picture was mixed.
Most major economic indicators are positive. Those who read the commentary from the armchair economists in the trading community have missed the rebound, so this is probably news. The economy is now in a self-sustaining rally. There are a variety of real economists writing on this topic, but you can get the flavor from John Williams, the SF Fed's head of research. Or you can read the pop econ guys who cater to what you already "know."
- Economic growth is still improving. The ECRI growth index has been stable in the 3.5 + range for four weeks. The ISM manufacturing and service indexes were both strong. The manufacturing index reading is consistent with GDP growth of 6%.
- Risk as measured by the St. Louis Fed Stress Index, edged slightly lower, remaining at a very low level. This measure tracks a lot of market data in the eighteen inputs. It is not a poll, nor opinions, nor a collection of anecdotes. We should all pay attention to some real data. The value moved to .048, almost the same as .050 last week. For more interpretation, the St. Louis Fed published a short paper with a very nice chart that helps to interpret this index. The chart does not reflect the recent continued decline in stress, but it identifies the dates for important recent events. The paper also has a longer version of the chart, illustrating past stress periods. I am not going to run the chart each week, but I strongly recommend that readers look at the paper. In the 2008 decline there was plenty of warning from this index -- no sign right now.
- Earnings News. The beat rate is still above average, although lower than the start of earnings season. More than 10% of companies are raising guidance. Best of all, companies that beat expectations are moving higher. These topics are all covered with the beautiful charts we have come to expect from The Bespoke Investment Group.
- Egypt. Despite the conflict, there seems to be progress toward a more representative government. There are also signs of that there is a creep toward normalcy, according to the overnight edition of the WSJ. Let us call this a cautious positive.
The bad news came from important sources -- employment and interest rates.
- Interest rates moved dramatically higher during the week. Abnormal Returns collected several good viewpoints on this important story in one great screencast. (Serious traders and investors all look to Abnormal Returns for advice on their daily reading, but the screencasts add another interesting dimension. Tadas adds commentary and interpretation to draw together several disparate views.)
- The Employment Report. Net job creation remains far below what is needed for a full economic recovery. On Wednesday I predicted that this would be a confusing report, and probably a weak one. I try to provide ideas about what to look for at a time when you can use the information. On this occasion the data really need a deeper look. Here are the most important conclusions.
- Hiring has improved, but layoffs are still at a high rate. I showed this in our employment preview which might be the only source that looks at both job losses and job creation.
- Weather had some effect, but even with a weather adjustment this was a poor report. The payroll survey counts anyone who got a check during the week including the 12th of the month. Most people do not understand this.
- The payroll survey came at the time of the annual "benchmark revisions." I understand that everyone hates revisions, but what is the alternative? In most months everyone analyzing job growth, including the BLS, is merely estimating the changes in employment. Once each year the BLS looks to state employment data and squares up the overall count of jobs with a source that provides hard data. We should applaud this. Dean Baker explains why we should focus on payroll jobs, rather than the employment survey and I agree. (I only wish I could get my fellow Michigan man to be more careful in distinguishing between job creation and net job creation. He reaches many more readers than I do, and he could help people if he was more careful with this part of the analysis.) The benchmark revisions were announced on a preliminary basis last October, so that part of the January report should not have been a surprise.
- The BLS method is doing a pretty good job in capturing job creation. The Birth/Death model, which adjusts for the residual after the "imputation step" was only revised down only slightly. This means that the actual data from state employment agencies showed only slightly lower job growth than the model had forecast. In addition, the B/D model is now based upon quarterly updates. Maybe the Johnny One-Notes who have never bothered to understand the BLS method will realize that they cannot just complain about the B/D model.
- Why was the unemployment rate so low, a drop to 9%? There was a huge disparity between the household survey, which showed major job gains, and the establishment survey's modest gains. Most people simply lose confidence when the data discrepancies are so large. Instead it is better to seek some understanding.
- First, and most important, there is no single explanation for the disparity in the two methods. Everyone always wants an easy and simple answer, but that does not mean we can find one.
- The confidence interval in the household survey is very large. Only 70,000 households are sampled. The 90% confidence interval is +/- 400K or so, much greater than the 100K for the payroll survey.
- The household survey was also benchmarked this month, based upon the decennial census. This coincidence made the report more confusing than usual. The adjustment to the most recent population estimates prevented a direct comparison of the December and January labor force numbers. Some analysts ignored this problem, attributing the decline in unemployment to a reduction in the labor force. In fact, the household survey shows a dramatic increase in employment, if one considers the population benchmarking.
click to enlarge
The jobs picture is not quite as bad as the payroll report made it seem, because the weather had some effect. It is not as good as the unemployment rate suggests because of the large error band on the survey. The criticism of the "denominator" for the unemployment rate has been exaggerated by many commentators since they ignore the decennial population adjustments, even though this is carefully explained in the report.
Too many things are changing at the same time. It will take another report or two for anyone to have confidence in job trends. Meanwhile, my own research shows that net job creation significantly lags what is needed to stay even with population growth and reduce unemployment.
Prof. James Hamilton, who certainly cannot be accused of a rosy view on the economy, sees the report as more positive than I do. Calculated Risk also had a good explanation of the disparity between the two methods.
The ugliest thing I saw this week was blatant investor deception. I have been vaguely aware of an ad predicting a market crash on January 31st. This ad sometimes ran on my own site. (I do not try to maximize revenue, but a little bit helps to defray costs. The amount collected is not especially meaningful).
It did make me wonder. I might be writing something offering a bullish long-term view for investors, while running an ad that predicted a crash. Is this important?
Meanwhile, those running this ad -- among the many selling fear, gold, and structured products with high commissions -- have shamelessly moved the prediction to the end of February. Wow! I guess they assume that no one will notice.
Our Own Forecast
We base our "official" weekly posture on ratings from our TCA-ETF "Felix" model, where we highlighted the Russia trade (NYSEARCA:RSX) Friday night. After a mostly bullish posture for several months, Felix has turned cautious. Last week we said it was a close call, and switched to neutral. This week it is still close, but we shifted back to bullish in the weekly Ticker Sense Blogger Sentiment Poll. The poll is now conducted on Thursday night, so it does not reflect Friday trading. Here is what we see:
- 61% of our 56 ETF's have a positive rating, down slightly from 63% last week.
- 73% of our 56 sectors are in our "penalty box," up from 69% last week, and an indication of significant short-term risk.
- Our universe has a median strength of only +7, about the same as +4 last week.
The overall picture is nearly neutral. We remained fully invested in trading accounts since there are several strong sectors, but we are watching the indicators quite carefully. This has been a very close call for several weeks.
[For more on the penalty box see this article. For more on the system ratings, you can write to etf at newarc dot com for our free report package or to be added to the (free) weekly email list. You can also write personally to me with questions or comments, and I'll do my best to answer.]
The Week Ahead
The upcoming week is less likely to be driven by data. As usual, we have the initial jobless claims data. There will also be more earnings reports and trade data on Friday.
I see three factors to watch:
- Bond trading and concern about inflation.
- Developments in Egypt and the rest of the Middle East.
- Bernanke's Wednesday testimony before the House Budget Committee.
The Bernanke testimony will be very interesting now that the GOP has control of the House.
Since I like the overall market prospects, I am looking for dips to buy in new accounts, where I am still trying to get fully invested. Nearly everyone is predicting a correction, so even those who are bullish are looking for better entry points. So far we have seen only minor chances.