Can we draw a parallel between the rise (low) in oil and the rise (low) in inflation ?
When we look at the inflation graph and the oil graph, we can infer that there is a direct correlation.
Image: Oil Barrel price chart Source: Trading Economics
Image: Inflation in the United States by Trading Economics.
But can we establish some mathematical correlation between the future price of a barrel of oil and inflation?
It's one thing to infer just by looking at graphs and plotting possible trend lines . However, statistically speaking, let's check if these two variables are correlated.
To begin this journey, many assumptions and adjustments have to be made. The first difficulty to be overcome is this: inflation data is presented every 30 days, while future price data for a barrel of oil is traded every minute. So, the main question is posed in mining and pre-processing the raw data.
To overcome this difficulty; I then decided to calculate the average price of a barrel of oil that occurred within a month to simplify our investigation.
The second problem in data mining and preprocessing is also challenging. Can we treat oil prices as a lagged variable ( L ag Variable ) ? In other words, one month's oil prices are only felt in the next month, and are not reflected immediately.
As we can see, the problems in generating our possible correlation model between the independent variable , in this case, inflation , and the dependent variable , here represented by oil , are numerous.
Image3: Table of results.
What were the results of the first tests?
We see that the correlation between these variables , as organized, is less than 5%. However, our model predicts that 8% of the variation in US inflation is explained by future oil prices . However, our P-Values show us that statistically we have to discard the test, since it is not within the range of 95% statistical reliability , which would prove that our variable is indeed significant in the variable we are trying to explain.
As a result, in our first attempt to draw a correlation , in the search for creating a regression model with a single variable failed profusely.
A more technical subject today, entering into what we call Data Science and Data Analytics.
Follow our next posts in search of a mathematical model to correlate these two variables.
How to take advantage of Business Analytics or Data Science in your investment fund?