Results of Logistic Regression

[Jane Allen]

Hi there, I tried to run a logistic regression, but it seems that my results does not seem to be correct. Firstly, I realized that there were 239 failures and only 63 successes - is this because of my data set itself?

And also, I am shocked to find that it is a "." reflected below the standard error, z test and confidence intervals.

Would really appreciate it if anyone could explain this to me. Thanks so much in advance!

logistic target i.reit_id market_cap debt_equity_ratio roaa roae ebitda cash totalliabilities totaldebt totalassets cash_ta pe_ratio noi_ta dividendpayoutratio

note: 1.reit_id != 0 predicts failure perfectly
1.reit_id dropped and 36 obs not used

note: 9.reit_id != 0 predicts success perfectly
9.reit_id dropped and 1 obs not used

note: 13.reit_id omitted because of collinearity

Logistic regression Number of obs = 302
LR chi2(-1) = 309.31
Prob > chi2 = .
Log likelihood = 0 Pseudo R2 = 1.0000

-------------------------------------------------------------------------------------
target | Odds Ratio Std. Err. z P>|z| [95% Conf. Interval]
--------------------+----------------------------------------------------------------
reit_id |
1 | 1 (empty)
2 | 6.73e-26 . . . . .
3 | 1.6e-103 . . . . .
4 | 5.07e-29 . . . . .
5 | 2.4e+202 . . . . .
6 | 7.2e-150 . . . . .
7 | 1.15e-24 . . . . .
8 | 1.6e-147 . . . . .
9 | 1 (empty)
10 | . . . . . .
11 | 1.05e-21 . . . . .
12 | 6.9e-186 . . . . .
13 | 1 (omitted)
|
market_cap | 1 . . . . .
debt_equity_ratio | 3.90e-09 . . . . .
roaa | 1.90e-63 . . . . .
roae | 2.53e+23 . . . . .
ebitda | 1.001428 . . . . .
cash | .9991711 . . . . .
totalliabilities | 1.00001 . . . . .
totaldebt | 1.000889 . . . . .
totalassets | .9994988 . . . . .
cash_ta | . . . . . .
pe_ratio | .9407419 . . . . .
noi_ta | . . . . . .
dividendpayoutratio | .7330463 . . . . .
_cons | . . . . . .
-------------------------------------------------------------------------------------
Note: 239 failures and 63 successes completely determined.

[Carlo Lazzaro]

Jane:
welcome to this forum.
As far as the 302 observations are concerned, Stata warns you that

239 failures and 63 successes completely determined.

But: 239+63=302 (that is, your entire e(sample)).
I think that this problem is due to your data, not -logistic-.

[Richard Williams]

Is reit_id the unique id number for each case? If so, you are including a dummy variable for each case, so of course you will have perfect prediction. If it is the id number, then do not include it in your model.