By "you must have the FD" I believe the other person was meaning that merely having the data file does you no good, unless you know the layout of the various data elements - example, the first 8 characters are the Customer number, the next 20 are the Customer name, the next 20 are the Customer address-1, etc..... If you don't know what the layout is, then you're not getting much benefit from just a string of data. The data record layout is contained in the COBOL FD entry (File Description, which precedes the Record description in a COBOL program).
Do you have the source module for a COBOL program which created the file, or maintains the file, or even lists the file. That source code will contain the record layout. However, remember your basic COBOL training, if the programmer used a COPYBOOK to contain the description of the record (and ALL good programmers would!), then you will need to find the source module for that COPYBOOK.
I hope I haven't lost you.
Once you've found the layout of the record, then the use of vutil to create a sequential ASCII version of the file will get you started. But again, remember in COBOL, there are many ways that data may be represented - display, binary, packed, editted, etc...... so a 20 may represent SPACE if the field is display [pic x(20)], such as the Customer Name, but it could be part of a number in a binary or packed field, such as Customer YTD Sales field [pic s9(7)v99 comp].
The use of ODBC's would be very helpful, because the XFD (Extended File Description) basically becomes a "map" of the data file, so that if you wanted to pull the file into EXCEL for example, the data element names would then automatically become your column headers. You can also filter out just the data elements you want and not have to load each entire record into your table.
I hope this helps you.