Plotting estimates from different regressions using coefplot

[Asaf Yancu]

Hi!
I am estimating an event study model using two alternative treatment variables:

gen more5 = (x > 1.05 * y)
gen more10 = (x > 1.1 * y)

Then, I define the event study and export its results using coefplot from SSC in Stata 18:

// Interaction terms
gen more5Xperiod = more5 * period
gen more10Xperiod = more10 * period
* Where period is the time variable, which ranges from 0 to 10.

label define moreXperiod_l 0 "-5" 1 "-4" 2 "-3" 3 "-2" 4 "-1" 5 "0" 6 "1" 7 "2" 8 "3" 9 "4" 10 "5"
label values more5Xperiod moreXperiod_l
label values more10Xperiod moreXperiod_l

// The event study uses two alternatives:
reg z more5 b5.period b5.more5Xperiod, robust
estimates store results5
// I am only interested in the interactions:
coefplot results5, vertical base keep(*b5.more5Xperiod*) xline(6) yline(0)

reg z more10 b5.period b5.more10Xperiod, robust
estimates store results10
coefplot results10, vertical base keep(*b5.more10Xperiod*) xline(6) yline(0)

This code exports two different plots. I think it would be more convenient to compare the results based on these two alternative treatment variables on the same graph. For example, for each period, the graph would present the coefficients obtained by more5Xperiod and more10Xperiod next to each other, but in different colors.

I am unable to share a subset of my data, but I hope my explanation is clear.


Many thanks!