Hi Stata Users,
I am a student who is doing the thesis project in Finance. I am using the STATA software for my analysis, but it is new to me and I have many doubts.
My goal is to use the propensity score to get the control sample, and from that to do a diff-in-diff analysis.

1)In PSM I need an "outcome variable" and I cannot understand what it can do. If the PSM is used to define, with a logit or probit model, the probability of being "treated" to obtain a control sample, what is the outcome variable for?! Without this variable, I can't use the "teffects psmatch" command.
I then tried to use the psmatch2 command which did not ask to necessarily specify the outcome variable. I wrote this command;
Code:
psmatch2 trattamento TOTALEATTIVOMIGLIAIA MARKETTOBOOKVALUE ROE ROA BETA, logit n(5)
where "trattamento" is a treatments variables who is 1 for treatments group, and 0 otherwise. The rest are independent variables. Finally nn (5) to obtain 5 matching for each group treated.
Is this command right? The result of STATA is anomalous, in fact the output for examples shows that it gives me 5 results very close to each other, although I am working with very different data. Array Now, i try with panel data and what cross section data, but i have the same results.
Code:
 trattamento    _pscore    _treated    _support    _weight    _id    _n1    _n2    _n3    _n4    _n5    _nn    _pdif1     1     .10830767    Treated    On support    1    479    344    345    343    346    347    5    .00004345 1    .11888814    Treated    On support    1    483    370    371    369    368    367    5    .00013978 1    .10962844    Treated    On support    1    480    347    346    348    349    350    5    .00044377 1    .10063743    Treated    On support    1    470    321    322    323    324    320    5    .00007291 1    .03988641    Treated    On support    1    454    40    39    41    42    43    5    .00005372 1    .15984402    Treated    On support    1    489    419    418    420    421    417    5    .00015104 1    .10108548    Treated    On support    1    471    323    324    322    325    321    5    .00002828 1    .04010314    Treated    On support    1    455    40    41    39    42    43    5    .00016301 1    .05172508    Treated    On support    1    459    93    92    94    95    96    5    6.594e-06 1    .14891554    Treated    On support    1    487    402    403    404    405    401    5    .00048198 1    .15500089    Treated    On support    1    488    415    416    414    413    412    5    .00010008 1    .09095192    Treated    On support    1    464    287    288    289    290    286    5    6.376e-06 1    .10368798    Treated    On support    1    475    329    328    330    331    327    5    .00059848 1    .13399818    Treated    On support    1    486    385    386    384    383    387    5    .00009612 1    .07748607    Treated    On support    1    462    230    229    231    228    227    5    .00017664 1    .10304725    Treated    On support    1    474    329    328    327    326    330    5    .00004224 1    .12397104    Treated    On support    1    485    376    375    377    378    374    5    .00029416 1    .10613654    Treated    On support    1    477    338    339    337    340    336    5    .00005225 1    .09863004    Treated    On support    1    469    312    313    311    314    315    5    .00021663 1    .11559417    Treated    On support    1    481    363    364    362    365    366    5    .00009804 1    .23019594    Treated    On support    1    495    448    447    449    446    445    5    .00554138 1    .16168199    Treated    On support    1    490    420    421    419    422    418    5    .00029365 1    .10530734    Treated    On support    1    476    335    336    334    337    333    5    .00006022 1    .04591739    Treated    On support    1    456    65    66    64    67    68    5    .00013421 1    .10162016    Treated    On support    1    472    325    324    323    322    326    5    .00005403 1    .09491507    Treated    On support    1    466    301    302    303    304    300    5    .00006167 1    .10829471    Treated    On support    1    478    344    345    343    346    347    5    .00005641
And Why the results of nn(5) is different from the "_id" which is the best in 1: 1? 2) If I entered a dependent variable in the panel data, what would then change from the diff-in-diff analysis?
I am sorry for the long message but I am very confused. I need your help and an explanation to understand where I'm wrong. Thank you very much.