OLS with clustered standard errors
Oleg Komarov (2020). okomarov/clusterreg (https://www.github.com/okomarov/clusterreg), GitHub. Retrieved .
I have a cross-sectional dataset with M&A acquirers and study performance. If I want to cluster standard errors on a SIC (Standard Industrial Classification) level, should g then be a N x 1 vector of SIC codes?