How would you optimize a slow-running SAS query when working with a very large credit portfolio dataset?
