why is it wrong?

It's absolutely right

*your

Sorry to everyone that has asked this same question. I created this for a class I teach and didn't expect that anyone but that class would look at the videos. The class is an MBA class and it is an introduction to time series. In an introduction to time series analysis is not quite as "data driven" as "cross-sectional analysis". Here is what I mean, with cross-sectional data we have tools like R-squared, or generalized least squares to fix error disturbances, or other measures to "tweak" parameters. The "correction factors" in time series are analyst driven, that is alpha and beta, at the level of the course I'm teaching, are not estimated parameters as in cross sectional data they are simulated parameters. That is, these parameters are chose by altering the parameters and watching how the alterations impact the measure of error you are using. So, alpha and beta are not in some way being estimated at this point. (if this helps. Cross sectional data is the math of algebra to calculus. Heavily explored, yet stationary analysis. However, time series is the math of difference or differential equations. So, motion. It is possible that in a more advanced setting there may well be methods for choosing the alphas and the betas, but my own work and academic exploration has not taken me to that place yet.)

Please feel free to use it.

That's where I use it so I'm fine if you use it. As long as it is not an example of what not to do.