I'm looking to do an analysis on a bank's total loan portfolio balance given a set of independent variables:

  • real and nominal GDP growth
  • real and nominal income growth
  • unemployment rate
  • CPI rate
  • 3 mth, 5 yr & 10 yr treasury yield
  • BBB corporate yield
  • mortgage rate
  • prime rate
  • DJI
  • house price index
  • commercial real estate price index
  • market volatility index

My hope is to build a model which will allow me to "stress" the independent variables so that I can forecast the bank's total loan portfolio balance. There isn't many article/white papers regarding this topic and so I'm hoping to get some suggestions into model selection and how to best select variables from the given set of independents.

My initial thoughts is to make the data stationary followed by using an ARIMA model. I'm not necessarily set on how to best choose my independents, I figure that I should use a stepwise method but I'm hoping to get thoughts on this.

Thanks for any help!


1 Answer 1


One way to stress those kind of variable you search after copula theory to making a joint distribution of all variables and simulate some risk scenarios and estimate the expected return of a portfolio.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.