Correlation Isn’t Causation: Mind The Escape Velocity Myths

I want to break down a few myths that are currently extremely prevalent with the majority of media, analysts and other punditry such as:

  • Lower oil prices a boost to the consumer
  • The country is headed towards full employment
  • ISM Survey’s suggest stronger economy ahead

As investors, it is critically important that we separate “fact” from “fiction.” The reason I say this is by the time that headline data suggests the economy is a recession, which is where a bulk asset price declines occur, the market will have already beaten you to the punch. The chart below shows how late the NBER is at dating recession.


This is the reason that the majority of asset managers and economists never “see” the recession in advance. By not analyzing the “interconnectivity” of the data, they develop a “myopic” view of singular data points and extract false correlations.

Let me give you a very important phrase to remember: “Correlation Is NOT Causation.”

This is critically important point to understand as false correlations lead to eventually poor outcomes in portfolios.

Let me give you an example of what I mean.

The other day I was questioned about the correlation between energy-related investments and 30-year mortgage rates. It would seem odd that the 30-year mortgage rate would have a correlation with energy stocks, but after running the data over the last 15 years there is indeed a correlation.

The point here is that just because there IS a correlation, it does not mean that there is actually a relationship between the two. It is a “spurious correlation.”

For example, here are some other spurious correlations to consider:






For investors, it is important to understand not only the relative factors that affect a certain stock, sector or index, but also the relevancy of inputs into the analysis.

So, with this understanding, let’s get to “Myth Busting.”

Lower Oil Prices Are A Boost To The Consumer

The majority of analysts and economists have been ratcheting up expectations for the economy and the markets on the back of lower energy costs. The argument is that lower oil prices lead to lower gasoline prices that give consumers more money to spend. The argument seems to be entirely logical since we know that roughly 80% of households in America effectively live paycheck-to-paycheck meaning they will spend, rather than save, any extra disposable income.

The problem is that the economy is a ZERO-SUM game and gasoline prices are an excellent example of the mainstream fallacy of lower oil prices.

Example:         Gasoline Prices Fall By $1.00 Per Gallon

                        Consumer Fills Up A 16 Gallon Tank Saving $16 (+16)

                        Gas Station Revenue Falls By $16 For The Transaction (-16)

                        End Economic Result = $0

Now, the argument is that the $16 saved by the consumer will be spent elsewhere. This is the equivalent of “rearranging deck chairs on the Titanic.”

Increased consumer spending is a function of increases in INCOME, not SAVINGS.

Consumers only have a finite amount of money to spend. Let’s use another example:

Example:         Big John Has $100 To Spend Each Week On Retail Related Purchases

                        Big John Fills Up His Truck For $60 (Used To Cost $80) (+$20)

                        Big John Spends His Normal $20 Per Week On His Favorite Craft Beer

Big John Then Spends His Additional $20 Savings On Roses For His Wife (He Makes A Smart Investment)


Total Spending For The Week = $100

Now, economists quickly jump on the idea that because he spent $20 on roses, there has been an additional boost to the economy. However, this is false.

John may have spent his money differently this past week but here is the net effect on the economy.

Gasoline Station Revenue = (-$20)

Flower Show Revenue = +$20


Net Effect To Economy = $0

Graphically, we can show this by analyzing real (inflation adjusted) gasoline prices compared to retail “control purchases.” I am using “control purchases” as it removes retail gasoline sales, automobiles, and building materials from the retail sales number to focus more on what consumers are buying on a regular basis.


The vertical orange line shows peaks in gasoline prices that should correspond (according to mainstream consensus) to a subsequent increase in retail sales.

While the argument that declines in energy and gasoline prices should lead to stronger consumption sounds logical, the data suggests that this is not the case.

Furthermore, the fall in oil prices are likely a bigger drag on the economy than currently believed.

Oil and gas production makeup a hefty chunk of the “mining and manufacturing” component of the employment rolls. Since 2000, when the oil price boom gained traction, Texas has comprised more than 40% of all jobs in the country according to first quarter data from the Dallas Federal Reserve.


(Read more here)

The obvious ramification of the plunge in oil prices is that eventually the loss of revenue will lead to cuts in production, declines in capital expenditure plans (which comprises almost 1/4th of all capex expenditures in the S&P 500), freezes and/or reductions in employment, and declines in revenue and profitability. Of course, the “ripple effect”of those actions impacts all of the related and ancillary businesses from piping to coatings, trucking and transportation, restaurants and retail.

Simply put, lower oil and gasoline prices may have a bigger detraction on the economy that the “savings” provided to consumers.

Newton’s third law of motion states:

“For every action there is an equal and opposite reaction.”

In any economy, nothing works in isolation. For every dollar increase that occurs in one part of the economy, there is a dollars’ worth of reduction somewhere else.

The Country Is Headed Towards Full Employment

On Friday, the Bureau Of Labor Statistics reported that the U.S. economy added 315,000 new jobs which was well ahead of Wall Street estimates.

While the headline was certainly encouraging there are several important points that were obfuscated.

1) Skews to the seasonal adjustments caused by the government shutdown last year are now affecting the headline employment data making it appear stronger than actual. This is seen in the household data which only showed an increase of 4,000 jobs in total for the latest reported month.

2) While the BLS reported strong headline employment, it obfuscated the loss of 150,000 full-time jobs while just 77,000 temporary jobs were added for the holiday season. (Not really expecting a strong shopping season now were they.)

Here is the breakdown of the top job categories for November:

  • Professional Services excluding temp help: +63k (of which administrative assistants +12K, bookkeepers +16.4K)
  • Retail Trade: +50K
  • Leisure and Hospitality: +38K, of which waiters and bartenders +26.5K
  • Manufacturing: +28K
  • Temp Help: +22.7K


Of course, the biggest problem continues to be the number of individuals sitting outside the labor force and not counted which increased by 69,000 in November.

That last point is the most important of them all. What is either missed, or just ignored, is the rather large group of individuals that have disappeared from the fabric of the economy and, while still alive, are simply ignored by current statistical measures. Let’s do some math using data provided by the Bureau of Labor Statistics.  [Note: I am only using the population between 16-54 years of age to eliminate the argument that “baby boomers are retiring” in droves, even though more individuals than ever, over the age of 65, remain employed.]

  • Total Working Age Population (16-54 years of age): 248,657,000
  • Total Nonfarm Employees (16-54 years of age): 114,523,000
  • Percentage Of Working Age Americans Employed (Full or Part-Time): 46.05%

Just for comparative purposes here is the same calculation at the turn of the century (January 1st, 2000):

  • Total Working Age Population (16-54 years of age): 211,410,000
  • Total Nonfarm Employees (16-54 years of age): 118,602,000
  • Percentage Of Working Age Americans Employed (Full or Part-Time): 56.10%

Here is what it looks like graphically:


Of course, here is the real problem. There are currently more than 93 million individuals that are simply no longer counted as part of the labor force.


You can see the clear surge in this group of individuals that began late in the Clinton years when the definition was changed to exclude all individuals unemployed longer than 52 weeks. This created an unintended consequence following the financial crisis as large chunks of the population have remained unemployed longer than 12-months as unemployment insurance was extended to 99 weeks.

Are we approaching “full employment” with the unemployment rate sitting at 5.8%? Such a number would suggest that 94.2% of the population that wants to work has a job. However, I am sure that if you ask any of the 44% of individuals, between the ages of 16-54 sitting outside of the labor force, if they would like a “real full-time job that pays benefits” you might be surprised at the answer.

ISM Survey’s Suggest Stronger Economy Ahead

There are things going on with the financial markets currently that seem just a bit “out of balance.” For example, asset prices are rising against a backdrop of global weakness, deflationary pressures and rising valuations. More importantly, there is a rising divergence between sentiment and hard data.

For example, this past week the Institute of Supply Management released their survey data showing extremely sharp surges in the data. It was immediately reported that this bodes well for continued economic growth going forward.

The problem, however, is that as with the employment report, seasonal adjustments skews are wreaking havoc on the data which can be seen by simply looking at other MAJOR indicators of the overall domestic economy.

Markit’s US Purchasing Managers Index which saw a decline to 10-month lows of 54.7 and has, like industrial production, fallen three months in a row.


Importantly, here are the“key points” from Markit that confirms much of my analysis regarding the impact of a strong dollar/weak global economy combination on U.S. exports.

“Output growth has now fallen for three straight months, taking the pace of expansion down to its lowest since the start of the year. Unlike January, however, this time the weaker rate of growth can’t be blamed on the weather.

Export market weakness holds the key to the recent slowdown,with manufacturers reporting the largest drop in export orders for nearly one and a half years.

There’s some reassurance from manufacturers continuing to boost their payroll numbers at a robust pace, but with backlogs of work showing almost no growth, the rate of job creation looks likely to moderate in coming months unless new order inflows pick up again.

The manufacturing and service sector PMI data available so far point to GDP growth slowing to around 2.5% in the fourth quarter.”

The Chicago Fed National Activity Index (CFNAI), arguably one of the most important economic indicators available, fell to .14 from .29 last month. (Note: The Chicago Fed National Activity Index (CFNAI) is a monthly index comprised of 85 subcomponents that provide a broad measure of economic activity nationwide.)

More importantly, while the Federal Reserve and ISM surveys have been showing strong increases in recent months; the production, income and consumption and housing components of the CFNAI have declined.  The chart below shows the CFNAI index broken down into the 3-month average of supply (production, income, employment) and demand (consumption, housing, sales).


There are TWO very important things to take away from the chart above. First, supply and demand have had an extremely tight correlation prior to the financial crisis. However, since the last recession demand has underperformed supply to a significant extent which confirms the weak economic underpinnings for the majority of the country.

Secondly, while government related surveys are showing a vast improvement in economic activity, there has been little marginal improvement in the CFNAI. In fact, as shown in the chart below, the 3-month average has turned significantly lower in the second and third quarters of this year which suggests that real economic activity is slowing.


With the economic recovery now more than six years into recovery it has become a “footrace” to the finish line. With asset prices at elevated levels in anticipation of an economic revival, the failure of such a resurgence could lead to a significant disappointment for investors. With extremely cold weather threatening a large chunk of the population, there are risks to both corporate earnings expectations as well as to economic growth.

Markit CEO Survey:

“U.S. companies reported the lowest degree of confidence since the survey began in late 2009, reflecting domestic concerns and a subdued external demand environment.”

Like I said in the beginning, most of the hard data suggests a much weaker economic environment than the current sentiment does. Both sets of data cannot be right.

Pay Attention

While I am not suggesting that the markets are about to come unglued or that the economy is about to plunge into recession, it is important you pay attention to the spectrum of data as it correlates with each other.

The story of the real economy is easy to read if you are willing to look below the headlines. As I stated at the beginning, by the time the mainstream media tells you that a recession has started, it really won’t matter much in protecting what remains of your portfolio.