7.6 Parameter estimates

We will first use linear regression to obtain parameter estimates. In the following code we will run regression model separately for each company and save them in a large list called lm_list

lm_list <- price_merge %>% 
  filter(-260 <= event_diff & event_diff < -10 ) %>% 
  split(.$ticker) %>% 
  purrr::map(~ lm(ret ~ MktRF + SMB + HML + RMW + CMA, data = .))

Note that we could have simply extracted the coefficients from each linear regression to calculate predicted returns manually. However, we will need the prediction intervals and base R’s predict() function produces them automatically.