Conditional Logit: Outcome Does Not Vary In Any Group (r2000)

Dear all,

​I want to investigate what factors contribute to the probability of a target to be taken over, so as to eventually make out-of sample predictions. For this, I have two datasets. One consists of all takeover targets from 1985-2014, and the other consists of all non-targets over the same period. The target data merely includes certain key ratios as of the year prior to the bid year (e.g., takeover bid was on July 12, 2012 -> variables are from 2011). For my research, I am planning to conduct a multivariate logistic regression. However, since other research matches targets with non-targets based on e.g., industry and year, I would like to do this as a robustness check. In order to do this, I merged the targets (master file) "many to many" using the following code:

This does seem to match the targets with non-targets based on both SIC and year (SICID is defined as: yyyy&2-digit SIC). Then to create the matched variable, I use the code:

However, when plugging this newly formed variable "match" into the conditional regression, it gives the error: "Outcome does not vary in any group r2000".

It would be great if someone knows what the problem is and how to solve the problem. Thanks in advance.

Kind regards,

Wesley