Hi,
I have a variable "Quintile" that splits the 1021 firms I have into quintiles of different probabilities based on a logit model. As such "Quintile" is 1 representing firms with the bottom 20% of probabilities , 2 if the firm is in quintile 2 of probability etc up until 5 with the highest 20%. Quintile 1 has 50 firms, I want to use propensity score matching to find 50 firms from each quintile 2 - 5 that are most closely related to the 50 firms in Quintile 1 based on various control variables. so far I have used the code below; this creates propensity scores for each firm that is not in Quintile 1 based on a new variable "Quintile1" (where 1 if in quintile 1 and 0 if not). How do I extract the 50 firms in each of the other quintiles? or even create a new variable that shows me which 50 firms from each quintile 2 - 5 are similar to quintile 1 firms. Many Thanks.
// Perform propensity score matching
gen Quintile1 = 1 if Quintile == 1 & Bid == 1
replace Quintile1 = 0 if Quintile1 == . & Bid == 1
tab Quintile1 //treatment variable
global Mvar Cashdummy Stockdummy Otherdummy Unknowndummy Diversifying samestate // deal control variables
reg CAR7 Quintile1 //CAR7 is outcome variable //t-test
reg CAR7 Quintile1 $Mvar //reg with controls
pscore Quintile1 $Mvar, pscore (myscore) blockid(myblock) comsup //
attnd CAR7 Quintile1 $Mvar, pscore (myscore)
//
Data looks like:
Firms | Quintile | Quintile1 |Var1 | Var2 |......
a | 1 | 1
b | 1 | 1
c | 3 | 0
d | 2 | 0
e | 3 | 0
f | 5 | 0
g | 4 | 0
etc...
I have a variable "Quintile" that splits the 1021 firms I have into quintiles of different probabilities based on a logit model. As such "Quintile" is 1 representing firms with the bottom 20% of probabilities , 2 if the firm is in quintile 2 of probability etc up until 5 with the highest 20%. Quintile 1 has 50 firms, I want to use propensity score matching to find 50 firms from each quintile 2 - 5 that are most closely related to the 50 firms in Quintile 1 based on various control variables. so far I have used the code below; this creates propensity scores for each firm that is not in Quintile 1 based on a new variable "Quintile1" (where 1 if in quintile 1 and 0 if not). How do I extract the 50 firms in each of the other quintiles? or even create a new variable that shows me which 50 firms from each quintile 2 - 5 are similar to quintile 1 firms. Many Thanks.
// Perform propensity score matching
gen Quintile1 = 1 if Quintile == 1 & Bid == 1
replace Quintile1 = 0 if Quintile1 == . & Bid == 1
tab Quintile1 //treatment variable
global Mvar Cashdummy Stockdummy Otherdummy Unknowndummy Diversifying samestate // deal control variables
reg CAR7 Quintile1 //CAR7 is outcome variable //t-test
reg CAR7 Quintile1 $Mvar //reg with controls
pscore Quintile1 $Mvar, pscore (myscore) blockid(myblock) comsup //
attnd CAR7 Quintile1 $Mvar, pscore (myscore)
//
Data looks like:
Firms | Quintile | Quintile1 |Var1 | Var2 |......
a | 1 | 1
b | 1 | 1
c | 3 | 0
d | 2 | 0
e | 3 | 0
f | 5 | 0
g | 4 | 0
etc...
