Being an amateur economist and statistician I enjoy performing regression analyses on the fundamentals of various stocks. My latest project has been creating an accurate model for James River Coal Company (JRCC).
Without being much of an expert on JRCC and the details surrounding its performance, I have put together a few key variables that I thought would be important to the company's pricing. I downloaded the year end operating statements and balance sheets to Excel, and began to test certain variables.
The variable I used to indicate the health of JRCC operations was the difference between annual revenue and the annual cost of revenue. I see a very clear pattern when this is charted out, reminding me of the recent stock trends.
The next variable that I thought would be important is the total equity at year end. This also gives obvious clues to stock price, but adds the "below the line" items that the previous variable might not include.
The next variable I used was mean coal price for the year. I gathered this information in a simple way, typing "mean coal price" and the year into Wolfram Alpha. The result came in terms of coal futures for the given year.
The final variable I used was the Arch Coal share price for the given date. I figured that since Arch Coal has a market capitalization of a little over one billion, it would be a good data point suggesting where a more liquid coal company stands in a macro point of view (including the affect of financial markets on share price).
The results to my surprise showed an R-squared of 1. I am really no expert in statistics, I just know how to do basic linear regressions in Excel, but nevertheless my coefficients and intercept were able to return the share price for each of my 5 report dates.
Since my dates only show information up until year end 2012, I decided to add Q2 2012, Q3 2012, Q4 2012, and Q1 2013 together to give me an indication of where the price should be at the end of the most recent report. After multiplying through the coefficients and adding the intercept, I get a share price of $2.81. That is about $0.89 above the price on 05/02/2013 (I assume the report was released after hours, so I take the price of the following trading day). $0.89 is quite a large difference for a stock at this share price. I assume the price was overshot because of the change in optimism about where the company is heading and perhaps the realization that company financials continue to decay (my training in economics also has indicated that markets tend to overreact).
Now for a bit more fun, I am going to use the same company fundamental numbers, and simply change the ACI share price to the current price per share (06/07/2013), as well as mean coal price between 05/01/2013 and 06/07/2013. This will give me a bit of indication of where the price should be gravitating.
The results give me a share price of $2.75, but since my last estimate was off the share price by $0.89, I will subtract that to try to correct for error. This leaves me with a current estimated price of $1.86 - $0.65 below the current share price of $2.51 (06/07/2013 11:20 AM).
I keep in mind that my Q1 forecasted share price was off by about $0.89. I consider this to be the an indication of optimism or pessimism about the company's future. Since the gap is less than $0.89 cents now (it has come in $0.24), it seems that optimism has increased, perhaps because of the debt announcements.
I estimate that since there hasn't been a lot of really solid news in the last month (other than restructured debt, which I understand will actually be more expensive to service than previous maturities (I recall a 10% senior note or something of the like)), I would say James River Coal is a bit overpriced at the moment. The upper price bounds that my model suggests is $2.75 - and after falling the past weeks below that limit, perhaps the model is correct. The lower bounds on the price should be around $1.86. So in the current environment I am a buyer below $1.86, a seller above $2.75, and a holder anywhere in between.
Of course all of this analysis is in good fun. I am not an expert, I just enjoy doing simple linear regressions in my spare time. It is quite clear that linear regressions don't always provide correct predictions, but at very least linear regressions can show us how pricing compares to historical trends.
Additional disclosure: I bought some put options expiring in June during the price bounce in the past couple weeks.