[Sebastian Grein]

David Radwin I also want to perform Propsensity Score Matching. My sample consists of M&A deals (deal = 1) and potential peer companies (deal = 0) from which I want to create a matched sample (M&A deals and control group) for further calculations.
Since I want to examine the influence on deal probability, deal is my treatment variable but also my dependent variable in the regression I want to perform afterwards. Therefore, I can not use the following command since t is here also y for me.

reg y x1 x2 t [fweight=_weight]

I want to ask if I'm doing anything incorrectly or how I can use the results of the PSM?

[David Radwin]

I'm afraid I do not understand what you are trying to accomplish, but I do recommend reading up on matching. You might start with the resources at the bottom of post #4: https://www.statalist.org/forums/for...36#post1682736

[Sebastian Grein]

My questions is more related to the application of PSM in Stata rather than the theoretical foundation. Therefore, I might think that the resources are not very helpful.

Regarding my sample, which part is unclear?

Description of relevant variables/covariates:

- treatment variable: deal (1;0)
- covariates: BTM and SIZE

Additionally, the control sample should match exactly on Industry (variable: sic_two) and year (variable: Year). After applying PSM I have my matched sample with treated and untreated observations. Since I want to conduct further calculations with the sample I need to consider the weights of each observation within the sample.

ssc install psmatch2, replace
logit deal i.SIC_two i.Year BTM_lag1 BV_ln_lag1
predict double pscore if e(sample)
gen double pscore2 = ind_year*100+pscore 
psmatch2 deal (pscore2), out (total_acc_freq_5) neighbor (5) caliper(0.05) common //noreplace

psmatch2 deal (pscore2), out (total_acc_freq_5) neighbor (5) common

pstest BTM_lag1 BV_ln_lag1, both

For my further calculations, I am doing a regression with the binary variable deal as the dependent variable to assess the influence on the deal likelihood of the company. Now my problem is, that deal should be my dependent variable and is also my treatment variable which is why the following command does not work (I would need to add deal as t again):

// general notation
//reg y x1 x2 t [fweight=_weight]

//applied with variables/covariates
regress deal total_acc_freq_5 BV_ln BTM LEV ROA REV R_D [fweight=_weight]