MM, Please Explain Time series regression analysis

[Mi6op]

You alwasy mention this, and the r2 factor in your rundown on the positives of a stock. I was curious if you could explain what it looks for and how to calculate it.

Thanks

[mrmarket]

Think of it this way. Take the 52 week chart of a stock. Then try to draw a straight line through the points of the data that minimizes the scatter around the straight line. You've just drawn a regression line.

The closer the points are to the line, the higher the r^2 correlation coefficient. My theory is that if you are buying stocks that are fueled by earnings and revenue growth with reasonable valuations, that the stocks that show the least scatter around this line are the ones that are most likely to continue their growth for another 15%.

If the datapoints were widely scattered around the line, and the r^2 was lower, then it would be more likely that the stock price would deviate from the straight line in the future...got it?

Microsoft Excel has an @RSQ function which calculates the correlation coefficient over a range of data (date, stock price). Good luck!

[stenzrob]

http://www.prophet.net/learn/taglossary.jsp?index=R

[Mi6op]

Makes sense, Thanks for the run down and link.