Probability of Default Model (PD Model)

  • Applications
  • Risk Management

The Prescio Probability of Default Model (PD Model) is a JAVA-based application designed to aid in the validation of Probability of Default (PD) models.

The PD Model includes key elements of a bank’s credit risk framework under the Basel regulatory guidelines. The application relies on statistical methods such as the Hosmer-Lemeshow Test, Binomial Test, and Traffic Light Approach to evaluate the goodness-of-fit of a given PD model.

Statistical methods used to calculate the probability of default include:

  • Hosmer Lemeshow Test: This model performs Hosmer Lemeshow test of PD model by using Monte Carlo simulation.
  • Manual HL Model
  • Dynamic HL Model
  • Traffic Light Model: Traffic Light approach to PD validation
  • Binomial Test: This model assumes independence of default events between all credits within the chosen rate grade.