Plotting the distance
Having not read anything written on Seeking Alpha in some time, it was nice to see Chuck Carnevale’s articles again using strong fundamentals and taking a look at the big picture of investing rather than the short term trading that takes place today. In the specific article that I was reading I saw that he was using (insert name of program here) to plot the stock price and that company’s earnings to compare. It got me thinking, can I do this in Equities Lab, and better yet, can I take it a step further.
Let’s make $SIRI our example company, let’s go ahead and plot the earnings. Before the crash in 2008, SIRI was incredibly volatile. In the interest of all of our sanity we will look at a chart of SIRI after 2008.
That doesn’t look to bad. The company increased by over 3000% in that timeline, and it appears that their earnings increased during that time as well. But really how correlated are the two?
Using the correlation operator found within Equities Lab, we can find how correlated any two values over any period of time. Here we are using “correlation-within” which will allow us to set a time frame – in this case, 504 days or two years.
Much like it appears in the original graph, the two values are highly correlated. As earnings goes up the stock price is sure to follow. Though we all know the old mantra, “correlation doesn’t mean causation” and that is completely true. However, in a case like this, earnings do tend to have an effect on stock price, and I may go into that in a different article to explain it.
So the values themselves are correlated. If one goes up, then the other will go up to, but is there any correlation in the percentages?
Adjusting our correlation tab is the first thing we need to do in order to ensure that we are running the proper test. We want to see the correlation between the change of close over the past year and the change of earnings over the past year. If the direction is correlated, could it be assumed that the percentages are correlated as well?
Wow, that chart is a lot different than the original one. It seems the percentages have been slightly correlated in the past, but now, and for the past couple of years there has actually been a negative correlation between the share price and the earnings of the company.
In order to explain why this is the way it is, we first need to normalize the earnings to plot more like the share price. We can do this by adding a “_%” to the end of the variable name like shown in the picture above. What this will do is it will check the value for earnings every day and plot it on a percentage basis. So, it starts from 0, just like the stock chart, and will increase or decrease over time based on the new earnings estimates.
If you aren’t looking for the line, you’ll actually miss it. Though SIRI has increased their share price by over 3000% in the past 8 years, they haven’t increased their earnings by nearly that much.
Let’s go ahead and log scale this chart to get a better look at where exactly the earnings curve is.
The earnings line seems to track the S&P 500 fairly well throughout this period. Percentage wise, it is pretty obvious that the share price didn’t track with earnings.
We get an even better look of that here. Over this time frame, earnings increased rather consistently by about the same percentage every year. On the other hand, share price – though it increased consistently – was not consistent in the level of growth it achieved every year.
The conclusion, in the case of SIRI, you cannot rely on earnings to predict how much the stock is going to increase, but you can use them as a guideline to see if the share price is going to increase. This sort of simple analysis can be done on most any stock traded here in the United States using Equities Lab, and with the Equihack language you have the ability to create studies as complex and well thought out as you like.