Estimating Jensen's Alpha

This short video walks through the concept of Jensen's Alpha, used in analyzing a portfolio manager's returns. I then do a simple regression to estimate Jensen's Alpha for a live series of portfolio returns.

[@xiaochenwang6729]

Hi, wonderful video teaching, i just really want to make sure the intercept from regression is Jensen's alphas right?

[@ricthomas6436]

Yes. My apologies as I should have made that more clear in the video. Thank you.

[@xiaochenwang6729]

@@ricthomas6436 awesome video

[@marvinhess9597]

Hey there, great explanation 👍🏻
In 7:08 you said jensen's alpha is 0,1667% per month and about 2% per year. So if I use daily historical data over a period of 1 year, does that mean my calculated jensen-alpha is the excess return over the benchmark per day?
And if I want alpha for that 1 year, do I need to multiply it by the number of data within that 1 year period or just 360 days?

[@ricthomas6436]

Hi - yes, if you use daily data then you are estimating a daily Jensen's Alpha (risk-adjusted excess return). However, if you want to annualize that number, you would use (roughly) 252 as the multiplier, which is the number of trading days during the year.

[@marvinhess9597]

@@ricthomas6436 thank you!

[@augustusg857]

I dont get how you can compare the capm which uses beta to the alpha on a regression model which doesn't use beta. Stock's return minus risk free rate vs stocks return with beta. How? In the regression model your alpha is when the market gives no returns and the beta is at zero too. So why compare the alpha on the model to capm?????