I am doing my master thesis and my subject is how covid-19 has affected the stock market.
I have already calculated abnormal returns and CARS using market model and OLS regression for different sectors. But I want to do panel regression as a robustness check, using either fixed effects model or random effects model. I have already gathered variables like cash/assets ratio, book-to-market value for individual companies (using year 2019 financial statement), but I dont know how should I aggregate this variables for my formed sector index.
Example: I have four tech stocks Apple, Facebook, Amazon and Netflix. I have calculated for each individual stock abnormal returns, cumulative abnormal returns, cash/assets ratio, ROE, ROA and so on. Now I need to form an index named tech sector of these four stocks. I have used market cap as a weight for stocks and calculated that way the abnormal returns and cumulative abnormal returns, but I dont know how I combine these accounting data independent variables for these formed index. Like should I sum the ROE:s of Apple, Facebook, Amazon and Netflix and take the average?
Bottomline: I want to explain different CAR:s as dependent variables in event window with independent variables being leverage, cash/assets, ROE... how this should be done, is panel regression even necessary?
I would be so thankful if somebody could help me out!!