Correlation Does Not Imply Causation

correlation does not imply causation


“His argument was presented with data so it must be correct.”

I stopped watching the news on TV years ago. I think it was the invention of the 6 box debate with everyone yelling at each other all-at-once. Or maybe it was when I watched a new reporter cut off Warren Buffett to tell the audience what he thought about the markets. It was so long ago I don’t remember. This was around the time that the news became entertainment, rather than news.

Despite my distaste for TV news, I was watching a news show yesterday on social media and I saw a story about how guns were dangerous and should be outlawed. This story was followed up by a story about how the temperature was becoming more erratic, this was due to global warming so we should all drive electric cars.

Political opinions aside (and those who know me, know that I have quite a few), these stories are just one of many, where commentators (frequently not scientists, specialists, or frankly anyone who is remotely qualified to speak on the subject) tend to elaborate how their point of view is correct because of this specific set of data. If you actually study the data in these issues or similar types of issues, it will become clear to you that correlation does not imply causation.

This phrase, “Correlation does not imply causation”, is actually well known among people who have an understanding of statistics. Unfortunately, it not as well known outside the field. Hopefully, I am able to shed some light on this concept in an entertaining way that does not involve yelling or the 6 box debate.


Correlation is the relationship or movement between two or more variables. This is typically presented as a number between 1 and -1. A relationship of 1 would be a 100% positive relationship where you could not tell one data point for the other in terms of movement. A relationship of -1 would be a 100% inverse relationship or exactly the opposite. An example of an inverse relationship would be a seesaw one side goes up and the other side goes down, or the performance of stocks and bonds, or the change in the interest rate of a bond and the price of that bond.


Causation is the relationship of one event to another event, where the second event is caused by the first. For example, there is a ball sitting on a table motionless. I pick up the ball and throw it, thus I am causing the ball to move.

Correlation vs. Causation

While most people will understand these two definitions individually, they frequently tend to confuse the two when the data is directed at an emotional or “heated” issue. For example, here are two charts, one has a positive correlation and one has an inverse or negative correlation. I have removed the chart labels to prove a point.

Positive Correlation

Positive Correlation

Inverse or Negative Correlation

Negative Correlation

The first graph has a positive 99% correlation, while the second one has a negative 93% correlation. For those of you who are not familiar with statistics or other data-focused fields of study, those are VERY high correlations.

Inverse or Negative Correlation

correlation and causation

While I didn’t have time to run the correlation numbers for the above chart before posting this, you can easily see that both data sets have a high inverse correlation. They look like a mirror image of each other.

Which charts are correlated and which ones are driven by causality?

Can you tell which charts above are based on correlation and which charts are based on causation?

It is very difficult to tell just from looking at the data sets. In fact, this should show you that data alone is not always a great predictor of causality. In fact, causality does not mean it must have a 100% correlation. For example, if a doctor prescribes antibiotics for a bacterial infection, this should cause the infection to disappear. There is a causal relationship there. However, for various reasons, it doesn’t always work.

The charts above did not have any labels on them, but the charts below do. Hopefully, when you read them you will understand the difference.

And the winner is…

Here are the charts with their axis and legend named to show what data is being compared. I can’t blame you if you chose the wrong set of charts. Causality is hard to determine just from a data set.

correlation without causation

(.992558 – or a 99.26% correlation of the data sets)- according to this, eating healthier leads to less divorce… Maybe there could be something to that.

correlation without causation

(-.933389 or a negative 93.34% correlation of the data sets) – According to this, we now know what the beekeepers do in their spare time. I apologize to all the adolescent beekeepers who are offended by the thought of smoking pot, but according to the numbers you are suspicious enough to arrest. Unless there is a scientific study being conducted at the moment which can prove the correlations in the first set of charts, I am going to float a guess that there is no causality.

correlation equals causation

This is a causation chart. In case you are not familiar with the second data set, the US Treasury Note yield is inversely correlated to the US Treasury Note price. This relationship is causal.

Next time you are watching a news story about how some new data suggests that oil prices are high because of a colder than normal winter, or that the stock market is going to crash because the sunspot cycle is at its maximum, turn your skeptical eye towards the inference and skip to the next channel and watch something else.

*The charts above were borrowed from ( & (

About Innovative Advisory Group: Innovative Advisory Group, LLC (IAG), an independent Registered Investment Advisory Firm, is bringing innovation to the wealth management industry by combining both traditional and alternative investments. IAG is unique in that they have an extensive understanding of the regulatory and financial considerations involved with self-directed IRAs and other retirement accounts. IAG advises clients on traditional investments, such as stocks, bonds, and mutual funds, as well as advising clients on alternative investments. IAG has a value-oriented approach to investing, which integrates specialized investment experience with extensive resources.

For more information, you can visit

About the author: Kirk Chisholm is a Wealth Manager and Principal at Innovative Advisory Group. His roles at IAG are co-chair of the Investment Committee and Head of the Traditional Investment Risk Management Group. His background and areas of focus are portfolio management and investment analysis in both the traditional and non-traditional investment markets. He received a BA degree in Economics from Trinity College in Hartford, CT.

Disclaimer: This article is intended solely for informational purposes only, and in no manner intended to solicit any product or service. The opinions in this article are exclusively of the author(s) and may or may not reflect all those who are employed, either directly or indirectly or affiliated with Innovative Advisory Group, LLC.

Scroll to Top