Irrespective of how robust the PD model is, it requires an appropriate calibration to ensure that the PDs are being estimated accurately, both Point-in-Time (PIT) and Through-the-Cycle (TTC).
- Calibration allows to account for sampling biases
- Enables the incorporation of an appropriate rating philosophy within the IRB model output
- Model scores which are mapped to PDs can be linked to external ratings
We have a robust calibration framework orchestrated by a team of global experts, which involves the following:
- Calculation of an initial PD
- Forward-looking, fully point-in-time PD for a 12 month horizon.
- Long-run forward-looking, fully through-the-cycle PD for a 12 month horizon.
- Customer Score: A Fair-Isaac style score for sanctioning, cross-portfolio comparisons, limit-setting and automated decision making.