01-28-2016, 07:25 PM
This wasn't an academic study. I just took the data in my Excel sheet for the stocks I follow and curve fitted them. My idea is that you should be able to see factors that affect the current price by curve fitting a line of price versus a given factor. For a factor to be relevant the datapoints should be evenly distributed about the line and as tight to the line as possible.
The best fit is using the forward PE ratio.
I then normalized the price by taking the stock's forward PE and dividing by the market's average forward PE (available from the Wall Street Journal website). Looking at this data allows me to look for more minor correlations. Doing this I found the two correlations mentioned.
I gave the result for the Beta correlation, but I hesitate to give the exact yield correlation, since I can't normalize it. The line for the data I used resulted in 3% yield at the average forward PE ratio, which matches the target I tend to use for selecting stocks. The actual curve parameters are probably dependent upon stock selection; however, I found the trend interesting.
With the Beta correlation, I get market value of a stock as (1.2 - 0.2 x Beta) x Forward Earnings x Market Average Forward PE. I expect the stock prices to vary about this calculated value, giving me a means of determining if it is over or under valued.
You can use this if you wish, but I really don't have the time to go much deeper.
The best fit is using the forward PE ratio.
I then normalized the price by taking the stock's forward PE and dividing by the market's average forward PE (available from the Wall Street Journal website). Looking at this data allows me to look for more minor correlations. Doing this I found the two correlations mentioned.
I gave the result for the Beta correlation, but I hesitate to give the exact yield correlation, since I can't normalize it. The line for the data I used resulted in 3% yield at the average forward PE ratio, which matches the target I tend to use for selecting stocks. The actual curve parameters are probably dependent upon stock selection; however, I found the trend interesting.
With the Beta correlation, I get market value of a stock as (1.2 - 0.2 x Beta) x Forward Earnings x Market Average Forward PE. I expect the stock prices to vary about this calculated value, giving me a means of determining if it is over or under valued.
You can use this if you wish, but I really don't have the time to go much deeper.