Portfolio Beta Weighting with Python

[@damneddude8299]

Thanks for all the videos, its like a goldmine of knowledge!! Thanks again a lot!!

[@carlosarrieta1145]

Amazing video! I did saw this stuff at college but never with practical examples.

[@John-xi2im]

awesome tutorial! Learnt a lot of new stuff ,thanks a lot mate!

[@victorl.mercado5838]

Regarding beta weighting for options, I prefer to beta weight my gammas and neutralize the unbalanced beta weighted delta by going long or short enough shares of the underlying to balance the portfolio delta. That way, the deltas are more likely to remain balanced as the market moves.

[@luisluiscunha]

I really enjoyed this: you explain things very well, thank you.

[@kaiwang2924]

Good time with Coefficient, Matrix and Linear Regression.

[@Cleusimacedo]

adorei❤ 21:04 21:08 21:13

[@gian_piano]

amazing as always!

[@Karemovv1998]

I tried to make ^SPX market the column 0 but it didn't work because yahoo finance makes the list in alphabet order Thank you for the video, it was informative.

[@TheAwesomeTigi]

This Video is amazing, I like your channel

[@slad1984]

Thanks for the video but in beta formula don’t you should divide covariance of market and stock to variance of market instead of covariance of market to itself

[@GG-pv2dn]

What if the inverse doesnt exist?

[@EMSxJIZL]

Assuming you're talking about step 4c, and you have not found an answer to your question already. Given this approach uses the Least Squares analytical solution to estimate the beta coefficients, it is fair to assume that all Least Squares assumptions have been checked. For your question, the linear independence assumption implies X is of full column rank. Given we know (assume) X is of full column rank and that X^T.X is square, it rules out the possibility of it being singular and, therefore, X^T.X is always invertible.

[@victorl.mercado5838]

Love the video. I wish I came across a similar video when I struggled to figure it out a few years ago. I use the pandas covariance and variance methods to calculate my betas, but it's effectively the same as your Step 4a example. I modified the code to conform to your format. See code below: def rolling_beta(df): m = df.iloc[:, 0] beta = [] for ind, col in enumerate(df): if ind > 0: # stock returns are indexed by ind s = df.iloc[:, ind] # Calculate covariance matrix between stock and market covariance = s.cov(m) # Calculate market variance variance = m.var() beta.append(covariance/variance) return pd.Series(beta, df.columns[1:], name='Beta') beta = rolling_beta(log_returns)