Point-in-time versus Through-the-cycle

Point-in-time versus Through-the-cycle

Point-in-time versus Through-the-cycle, two key design elements ( philosophies) of credit rating systems are compared and contrasted. So called hybrid systems may lie somewhere between the two designs.

Aspect	Point-in-time	Through-the-cycle
Who is applying which system	Financial firms	Rating agencies, Financial firms
What are the primary uses	Pricing, Portfolio Management, Stress Testing, IFRS 9 Reporting	Underwriting, Regulatory Capital
Which predictive variables are included	Both static / slowly varying and dynamic	Only static / slowly varying
What happens to the forecast PD in a downturn	Increases	Remains the same
What happens to the rating in a downturn	Deteriorates	Remains the same
What happens to the rank ordering of risks during downturn	Might change	Remains the same
How does it impact estimated risk capital for banks	Volatile	Stable
Can rating cut-off be used to set business segmentation criteria or limits	No	Yes
How is the PD correlated in time with credit spreads	Some positive correlation	Little correlation
Is there a unique approach to obtain this design	No	No
How can estimates be validated	Comparison with realizations over short periods	Comparison with realizations over long periods

---