Amihud and Goyenko Methodology (2013), R-Squared Predictive Power of Fund Returns

Dear all,

Hello I am trying to perform Amihud and Goyenko's methodology on R for a study but have little R experience. How would I run looping regressions for 20yr time period in which I have to store r-squared values derived from an CAPM regression of 800 fund returns against FTSE returns? I know it is not specific question with code lines but it would help A LOT if someone had a general idea of methdological procedures. Thanks

I think you would need to understand how to do the appropriate regression, on any given subset.
Then learn how to iterate over subsets to repeat the analysis.
Do you think you understand your requirements well enough to perform 1 of the regressions ?
If so, parameterise it and use a purrr function to iterate over the parameters.