Retail Loan Loss Forecasting model

  • Featured
  • Risk Management

Prescio applied fundamental research to further develop the retail loan loss forecasting model for a major U.S. bank. The model uses a Markov chain approach to predict the movement of loans between states of delinquency and forecast the losses. The intent of this research was to determine a method to incorporate changes in macroeconomic factors in the form of impulses within the mathematical constraints of matrices. A methodology was applied using Lagrange Multipliers to incorporate multivariable impulse and subsequent counter impulse within the rows of a matrix without violating mathematical constraints.