# How to model this data?

I have a dataset with the following structure:

Choice   Bank Bank_x  Customer Customer_x
0      UBS     .       1          .
1      CS      .       1          .
1      KZ      .       1          .
0      VA      .       1          .
--------------------------------------
0      UBS     .       2          .
0      CS      .       2          .
1      KZ      .       2          .
0      VA      .       2          .


Where choice is whether a customer would open an account in a given bank tomorrow (0 = no, 1 = yes). In the data there are a number of bank and customer specific variables.

My first thought was to use a Hierarchical logit model with bank choices nested within customers, so as to get parameters for each customer. However I am suddenly in doubt about whether this is the right choice, and I would like to hear any thougths from you guys before I settle on an approach.

-