Fama and French 3 Factor Model Excess Stock Returns

I am currently faced with two data frames: (1) panel data with many different companies and their monthly returns (2) the Fama and French factors by month between 12/1984 and 12/2017. not all the firms will have the same date length as the FF3 due to the fact that some companies will have gone public on different dates. I need to accomplish 3 key steps with your help: 1) Create a new column (name: risky_ret) on the companies information df, based on the return for month X of company Y minus the risk free (rf) rate in month X in the FF3 df. Thus basing ourselves on some kind of conditional looping based on the date columns. 2) Regress, always based on dates, the 'risky_ret' on the corresponding values for 'mktr', 'smb' and 'hlm' - and retrieve the expected return based on the FF3 (by adding back the corresponding risk free rate) and save it a new column (name: exp_return) in the stock data df. 3) Save the residuals between the realized returns ('return' in the stock df) and the FF3 expected return in a column in the stock data df. Please find some pictures below :)