Default Rate Attribution: Key to Portfolio Assessment

People in credit & lending risk teams often mention that their default/delinquency rates have been dropping (or increasing).

My thought/response usually is ‘What are the real drivers?’

If it’s a fast growing portfolio, many times growth itself masks the risk as seen purely from overall portfolio metrics. Collection efforts are also lost in it.

There are several well known ways to explain the nuances using:

• ➡️ Recent Vintage/Cohort curves being steeper/flatter.
• ➡️ Roll/Cure rates increasing or decreasing
• ➡️ A simple metric I like is the default/delinquency rate of only matured portfolio (say 6/9 months old)

But overall portfolio default/delinquency rates always remain the top metric which everyone becomes habitual of using while reporting, citing & judging. It just can’t be avoided.

My favourite method has been quoting the overall default rate, but with an added monthly/quarterly differential + attribution, which makes a lot of difference in perception. e.g.:

We closed the month at X-0.10% delinquency rate vs X% last month. It came down by 0.50% due to new originations vs 0.30% reduction last month. Shows we’re accelerating originations. It came down by 0.40% due to rollbacks/reinstatements to current vs 0.20% reduction last month. Shows we’re curing/collecting better. It increased 0.80% due to roll-forwards/deterioration to new delinquencies vs 0.60% increase last month. Shows either vintages with high originations are maturing now or their default curves are growing steeper than usual.

Note how an improved month based on overall default rate, turns out to be not so straightforward. Making this differential important for any growing portfolio.

I know it’s a very basic metric/topic but I’d be curious to know how other risk folks report & judge portfolio risk through metrics/kri.

#DefaultRateDifferential #DefaultRateAttribution #CILayer #CreditIntelligenceLayer #Risk #CreditRisk #PortfolioRisk www.cilayer.ai

---

Originally posted on LinkedIn.