Weekly Economic Vital Signs - The Inaugural Edition

|
Includes: DIA, IWM, QQQ, SPY
by: Lawrence Fuller

Summary

This is the first in a series of articles focused on analyzing the previous week's economic data releases.

Our objective is to identify and focus on what are leading indicators of economic activity in hopes of gaining insight as to whether the economy is strengthening or weakening.

This week we take an unconventional look at the January employment report.

While the headline numbers indicate a strengthening economy, digging deeper into the data reveals what may be just the opposite.

US economy The US economy is a powerful and complex machine that produces and consumes more goods and services than any other economy on earth. For this reason, it has long been considered the ultimate engine of global economic growth.

Every day reports are released about economic activity. These reports typically come on a monthly or quarterly basis, but some are as frequent as once a week. They tell us about the performance of different segments of the US economy-consumer spending, manufacturing, capital spending and housing, among others. Each is a vital sign of sorts, and collectively they tell us about the health of the overall US economy.

Some reports are more important than others. For example, retail sales is far more significant than the consumer price index. While both are reported once a month, retail sales tells us about where the economy is headed, while the CPI tells us more about where it has been.

Economic reports can typically be classified as either leading, lagging or coincident indicators of economic activity. Changes in leading indicators tell us about where the economy is headed, while lagging indicators reflect the economy's historical performance and tend not to change until after a new trend in the overall economy has already been established. Coincident indicators give us a real-time pulse on the current state of the economy. Some reports give us a perspective on all three to varying degrees.

Interpreting all of the data that is available is not an easy task, yet it is an important one if investors want to form their own opinion about the direction of the overall economy. I am not talking about forecasting recessions. Instead, I am talking about identifying changes in trends that indicate whether the economy is strengthening or weakening. Rates of economic growth are highly correlated with rates of corporate revenue and profit growth, which in turn impact stock prices. So having a handle on the direction of the overall economy is an important part of any successful investment strategy.

The first step in interpreting the data is identifying what is leading and what is lagging. This is complicated because pundits, politicians and the press regularly misinterpret the significance of one report over another. Even the experts will disagree about what a specific report means for the economy moving forward. A perfect example is the huge emphasis placed on the monthly unemployment rate, which currently stands at 5%. This is a very strong number, indicating we are near full employment. That is good news. The problem is that changes in the unemployment rate lag changes in other economic indicators.

For example, changes in the number of jobs gained or lost each month, which is more of a coincident indicator, will lead changes in the unemployment rate. The monthly payrolls report is coincident because it gives us a pulse on what is happening in the economy today.

There are several other economic indicators, including retail sales, which will have deteriorated well in advance of when the unemployment rate begins to rise again. These are the indicators that investors want to focus on, because they will help us identify whether the economy is more likely to strengthen or weaken as we move forward.

Further complicating the interpretation process is the stock market's initial reaction to the release of a report. If the market surges within seconds following the release of a consumer confidence report, is that to be interpreted as news that is good or bad on the economic front? These immediate and sizable moves are a relatively new phenomenon. The emergence, and now dominance, of computerized trading programs over the past decade is the source of this activity, because these programs all align simultaneously to buy or sell based on whether a specific number is above or below expectations. This is a silly game that has nothing to do with interpreting anything. Often times the initial move in the market reverses course just minutes later and heads in the opposite direction.

Furthermore, until very recently, after the initial computer-generated response to a report, the stock market has been reacting to economic reports on the basis that bad news for the economy is good news for the stock market. This is because bad news meant that the Federal Reserve would delay raising short-term interest rates, thereby pulling the proverbial punch bowl away from investors. That pattern was the norm throughout last year until the Fed finally ended its zero-interest-rate policy with a rate increase in December.

Ironically, the stock market itself has long been considered a leading economic indicator, because it would rise or fall, resulting in a reversal in trend, well in advance of changes in the direction of overall economic activity. Investors would collectively buy or sell stocks in anticipation of a rise or fall in corporate earnings, which would coincide with an increase or decrease in economic activity. As a result, the stock market has long been considered a discounting mechanism or leading indicator.

Many believe that the Federal Reserve's stimulus policy of quantitative easing muted the stock market's ability to serve as a leading economic indicator. Whether that was the case or not, monetary policy is no longer serving to as a tailwind to stock prices, which means that the stock market, as well as the bond markets, should now be included as one of many valuable leading economic indicators. It also means that bad news on the economic front will no longer be interpreted as good news for stock prices.

The Objective -

This article is this first in a series intended to provide a concise and unbiased analysis of the previous week's economic vital signs. I will be leaning upon the extensive expertise of my mentor, who is also a new contributor to Seeking Alpha, for insightful commentary each week. Larry Fuller is not only my mentor, but my father as well. Look him up!

Each week there can be 20 or more reports on the economic calendar. Which reports were leading indicators and which ones were lagging or coincident? Why did one report receive so much attention from the press, while another barely received mention? What was the stock market's response to the release, and why? What clues can we uncover in the more detailed reports that will help us identify an upcoming change in trend? We don't intend to analyze every report or to simply regurgitate the numbers already reported. Our goal is to address the one, or ones, that are the most relevant, and provide a unique perspective in hopes of helping investors formulate their own opinions about the direction of the overall economy and modify their investment strategy accordingly.

In reviewing all of the reports since the beginning of the year, we decided to focus on what is considered to be the most important one of all to begin this series, but not in the manner that most are used to seeing it presented. The monthly non-farm payrolls number, otherwise known as the jobs report, is a volume of information. The report includes so much information that we can identify data in it that is leading, lagging and coincident all at the same time.

In the January report we learned that hiring surged with the addition of 292,000 jobs in the final month of 2015. As with every new jobs report, the previous two months are revised with more incoming data. The gain in November's initial figure was upwardly revised from 211,000 to 252,000. That number will be revised a second and final time in next month's report. The gain in October was revised for a second and final time, up from 298,000 to 307,000, which was the biggest monthly job increase for the year. The unemployment rate stood unchanged at 5%. These numbers, on a stand-alone basis, indicate a strong economy.

The monthly payroll number is generally considered to be a coincident indicator, giving us insight into what is happening in the economy right now, while the unemployment rate is most assuredly a lagging indicator to economic activity. This is because it typically doesn't start to rise until after a recession has already begun, and it continues to rise well after a recovery has already started. The last recession began in January 2008, but the unemployment rate didn't rise substantially until May of that year. Well after the stock market bottomed and the economic recovery began in 2009, the unemployment rate continued to climb, peaking at 10.2% in October of that year.

If we dig deeper into the non-farm payrolls report, we can look at the quality of the jobs created, how many of the jobs are part-time, as well as information pertaining to average hourly earnings and the length of the workweek. These are all valuable data points that receive extensive coverage by pundits and the press, and we intend to add color to these as well next month, but for now we want to focus on a development that isn't being discussed. It has to do with a pattern that we identified while reviewing the historical payroll data on the Bureau of Labor Statistics website.

The data we are focusing on in this analysis is the average revision between the initial estimate of the number of jobs created each month and the final number. The initial payroll report can be subject to significant revisions. The BLS states on its website that the margin for error can be as great as 100,000 jobs per month. All surveys are subject to some sampling errors. This is why the BLS makes seasonal adjustments to its data depending on the month of the year.

It also makes an adjustment to the payroll number in order to account for the job creation it assumes is resulting from the formation of new businesses that are too new to be included in its survey. The statistical model it uses for this calculation is called the birth/death model. One downside to this model is that it misses turning points in economic activity. It overstates job creation by new businesses as the economy is slowing or contracting, because fewer new businesses are being created. It also understates job creation when the economy is starting a new expansion and lots of new business formation is taking place.

What I have done in the chart below is show the average revision to the initial payroll number for the 12 months of each calendar year, as well as the number of revisions that were negative, and by deduction positive. I also added the return for the S&P 500 to see if there was any relevant relationship between the data and market returns. I think the results are noteworthy.

Mean revision

Negative months

S&P 500 Total Return

2015

-10,000

6

1.40%

2014

37,000

1

13.69%

2013

21,000

4

32.39%

2012

24,000

3

16.00%

2011

28,000

3

2.11%

2010

40,000

1

15.06%

2009

12,000

4

26.46%

2008

-73,000

11

-37%

2007

5,000

6

5.49%

2006

23,000

5

15.79%

2005

31,000

3

4.91%

2004

21,000

4

10.88%

Click to enlarge

Source: Bureau of Labor Statistics

You can see that prior to the last recession in 2008, the BLS had a tendency to under-report job creation as the economy was expanding. For example, in 2005 there were only three downward revisions to the payroll number between the initial estimate and the final revision two months later. Including the nine upward revisions, the mean revision, or average, was an increase of 31,000 jobs per month for the year. As the rate of economic growth slowed in 2007, the number of negative monthly revisions increased and the size of the upward revisions began to shrink. When the economy began to contract in January 2008, the revisions were nearly all to the downside and included months when the economy lost jobs.

The recession lasted well into 2009, but by 2010 the recovery was underway, and the upward revisions were substantial (average of 40,000) and occurred nearly every month. The BLS was underestimating job creation at this point. The time period from 2008 - 2009 was obviously a major turning point.

Now focus on what is happening with the data today. We only have 10 months of data, as the final revisions for November and December have yet to be reported. Still, we have seen six negative revisions, which is more than any other year during this expansion. We are also seeing a negative number on average for each month for the first time since 2008. What are the implications?

Based on the historical pattern, it appears that these negative revisions are telling us that the economy is at another turning point. It isn't a good one. This doesn't mean that the US economy is on the verge of another recession. Yet it does lead us to believe that the rate of economic growth in the US is slowing well below what the consensus expects it to be in 2016. This has significant ramifications for corporate earnings and revenue expectations as well. We plan to do more work with these numbers and update the data in this "revision index" each month to see if it has real forecasting value. It certainly runs contrary to everything we heard about the employment situation just a few days ago. Stay tuned!

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

I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.