Seeing a bump in early delinquencies

Anyone else seeing a bump in early delinquencies? On our 2023 Q3 unsecured vintage, 30+ DPD EPD rose from 1.8% to 2.6% month over month after a 25 bps APR increase in January, and I’m trying to parse signal vs seasonality — are you attributing similar moves to repricing, mix shift, or score drift?

‌⁠‍⁠​‍​‍‌⁠‌​​‍​‍​⁠‍‍​‍​‍‌‍‌⁠‌‍​⁠‌‍‍‍​⁠‌‌​‍​‍​‍⁠​​‍​‍‌‍‍⁠​‍​‍​⁠‍‍​‍​‍‌‍⁠‍‌‍‌‌‌⁠‌⁠‌‌⁠⁠‌⁠‌​‌‍⁠⁠‌⁠​​‌‍‍‌‌‍​⁠​‍​‍​‍⁠​​‍​‍‌‍‍‌‌‍‌​​‍​‍​⁠‍‍​‍​‍‌‍⁠‍‌‍‌‌‌⁠‌⁠​‍​‍​‍⁠​​‍​‍‌‍‌​​‍​‍​⁠‍‍​‍​‍​⁠​‍​⁠​​​⁠​‍​⁠‌‍​⁠​​​⁠​‍​⁠​​​⁠​⁠​‍​‍​‍⁠​​‍​‍‌‍‍​​‍​‍​⁠‍‍​‍​‍‌‍‌⁠​⁠​⁠‌​‍⁠‌⁠​‍‌‌​‌​⁠‌‌​⁠‌⁠​‍⁠‌‌⁠‌​‌⁠​‍‌‍⁠​‌⁠‌‍‌⁠‍‌‌‍‌​‌‌⁠⁠‌‌‍‍​‍​‍‌⁠⁠‌​

And , I totally get your frustration. That spike from 1.8% to 2.6% is a bit alarming, especially after an APR increase. In my experience, mix shifts and seasonal factors have played a significant role in similar cases, but I find the interplay can be tricky to nail down. Have you found any particular cohorts that are more impacted than others?

‌⁠‍⁠​‍​‍‌⁠‌​​‍​‍​⁠‍‍​‍​‍‌‍‌⁠‌‍​⁠‌‍‍‍​⁠‌‌​‍​‍​‍⁠​​‍​‍‌‍‍⁠​‍​‍​⁠‍‍​‍​‍‌⁠​‍‌‍‌‌‌⁠​​‌‍⁠​‌⁠‍‌​‍​‍​‍⁠​​‍​‍‌‍‍‌‌‍‌​​‍​‍​⁠‍‍​⁠‌​​⁠‌‍​⁠‌‍​⁠‌‌​‍⁠​​‍​‍‌‍‌​​‍​‍​⁠‍‍​‍​‍​⁠​‍​⁠​​​⁠​‍​⁠‌‍​⁠​​​⁠​‍​⁠​​​⁠‌‌​‍​‍​‍⁠​​‍​‍‌‍‍​​‍​‍​⁠‍‍​‍​‍‌⁠​⁠‌⁠‌‌‌⁠‌‍‌​‌‍‌‍​⁠‌‌​‌‌‍⁠‍‌⁠​⁠‌‍‍‍‌​‍⁠‌⁠​​‌‍⁠​‌‌​⁠‌⁠‌‍‌‌​​‌​⁠‍​‍​‍‌⁠⁠‌

But it’s tricky to parse those shifts, isn’t it — i noticed a similar rise in early delinquencies after we adjusted APRs last year. I think mix shifts can play a big role, especially in a tighter economy — customers might prioritize different lending products when budgets get tight.

‌⁠‍⁠​‍​‍‌⁠‌​​‍​‍​⁠‍‍​‍​‍‌‍‌⁠‌‍​⁠‌‍‍‍​⁠‌‌​‍​‍​‍⁠​​‍​‍‌‍‍⁠​‍​‍​⁠‍‍​‍​‍‌⁠​‍‌‍‌‌‌⁠​​‌‍⁠​‌⁠‍‌​‍​‍​‍⁠​​‍​‍‌‍‍‌‌‍‌​​‍​‍​⁠‍‍​⁠‌​​⁠‌‍​⁠‌‍​⁠‌‌​‍⁠​​‍​‍‌‍‌​​‍​‍​⁠‍‍​‍​‍​⁠​‍​⁠​​​⁠​‍​⁠‌‍​⁠​​​⁠​‍​⁠​​​⁠‌‍​‍​‍​‍⁠​​‍​‍‌‍‍​​‍​‍​⁠‍‍​‍​‍‌‍‍‍‌​‌⁠‌​⁠‌‌‌‍​‌‍​‍​⁠​⁠‌⁠‍‍‌‌‍‌​⁠​‍‌‍‌​‌‍⁠​‌⁠‌‌‌⁠​​‌⁠‍‍‌⁠‌⁠‌​​‍​‍​‍‌⁠⁠‌

That spike does feel like a surprise guest at a party you weren’t expecting! Maybe a deeper dive into client segmentation could help clarify if it’s the price change pulling more risky borrowers or just seasonal shifts. Have you had a chance to look into how your approval criteria might have shifted since the APR changes?

‌⁠‍⁠​‍​‍‌⁠‌​​‍​‍​⁠‍‍​‍​‍‌‍‌⁠‌‍​⁠‌‍‍‍​⁠‌‌​‍​‍​‍⁠​​‍​‍‌‍‍⁠​‍​‍​⁠‍‍​‍​‍‌⁠​‍‌‍‌‌‌⁠​​‌‍⁠​‌⁠‍‌​‍​‍​‍⁠​​‍​‍‌‍‍‌‌‍‌​​‍​‍​⁠‍‍​⁠‌​​⁠‌‍​⁠‌‍​⁠‌‌​‍⁠​​‍​‍‌‍‌​​‍​‍​⁠‍‍​‍​‍​⁠​‍​⁠​​​⁠​‍​⁠‌‍​⁠​​​⁠​‍​⁠​​​⁠‌⁠​‍​‍​‍⁠​​‍​‍‌‍‍​​‍​‍​⁠‍‍​‍​‍‌‌​⁠​‍⁠‌​‍⁠‌‌​⁠​‌⁠‌‍​‍⁠‌‌⁠​‍‌‍‌‍‌‍‍‍‌‍‌⁠‌‌‌⁠‌​‍​‌‍‌​‌​⁠⁠‌‌​⁠‌⁠‌‌​‍​‍‌⁠⁠‌

I’ve noticed similar trends, especially after price adjustments. It might be worth examining specific client segments to see how they responded — @henry4028 mentioned a shift in approval criteria, which could be impacting your metrics too. Have you looked into that yet?

‌⁠‍⁠​‍​‍‌⁠‌​​‍​‍​⁠‍‍​‍​‍‌‍‌⁠‌‍​⁠‌‍‍‍​⁠‌‌​‍​‍​‍⁠​​‍​‍‌‍‍⁠​‍​‍​⁠‍‍​‍​‍‌⁠​‍‌‍‌‌‌⁠​​‌‍⁠​‌⁠‍‌​‍​‍​‍⁠​​‍​‍‌‍‍‌‌‍‌​​‍​‍​⁠‍‍​⁠‌​​⁠‌‍​⁠‌‍​⁠‌‌​‍⁠​​‍​‍‌‍‌​​‍​‍​⁠‍‍​‍​‍​⁠​‍​⁠​​​⁠​‍​⁠‌‍​⁠​​​⁠​‍​⁠​​​⁠‍‌​‍​‍​‍⁠​​‍​‍‌‍‍​​‍​‍​⁠‍‍​‍​‍‌⁠​‌‌‍‌​‌​⁠‍‌‌​‌‌​‌​​⁠‌‌​⁠‌​‌‍‌‌‌‌‍‌‌​‍‌‌​‌⁠‌‌⁠⁠‌‌‍​‌‍⁠​​⁠‍‌‌‍‍‌​‍​‍‌⁠⁠‌

I’ve seen waves in delinquencies too, and it’s often a mix of factors — like the APR hike you mentioned. One thing that helped us was closely analyzing customer behavior after our rate adjustments. It’s definitely worth checking if different segments reacted differently, as @henry4028 suggested.

‌⁠‍⁠​‍​‍‌⁠‌​​‍​‍​⁠‍‍​‍​‍‌‍‌⁠‌‍​⁠‌‍‍‍​⁠‌‌​‍​‍​‍⁠​​‍​‍‌‍‍⁠​‍​‍​⁠‍‍​‍​‍‌⁠​‍‌‍‌‌‌⁠​​‌‍⁠​‌⁠‍‌​‍​‍​‍⁠​​‍​‍‌‍‍‌‌‍‌​​‍​‍​⁠‍‍​⁠‌​​⁠‌‍​⁠‌‍​⁠‌‌​‍⁠​​‍​‍‌‍‌​​‍​‍​⁠‍‍​‍​‍​⁠​‍​⁠​​​⁠​‍​⁠‌‍​⁠​​​⁠​‍​⁠​‌​⁠​‍​‍​‍​‍⁠​​‍​‍‌‍‍​​‍​‍​⁠‍‍​‍​‍‌‍‍​‌​⁠‌​⁠‌⁠‌‍‍‍‌⁠‌‌‌⁠​​‌​‍⁠‌‌​⁠‌⁠‍​‌​‍⁠‌‌‌‍‌‌‍‍‌⁠‍​‌⁠‍‍‌⁠‍‍‌‍‍‍​‍​‍‌⁠⁠‌

It’s definitely interesting to see the shifts after pricing changes. From our experience, closely analyzing the underlying reasons for delinquency spikes — like understanding the customer journey — helped us adjust our approach effectively. Have you considered segmenting your delinquencies further to pinpoint specific areas of concern?

‌⁠‍⁠​‍​‍‌⁠‌​​‍​‍​⁠‍‍​‍​‍‌‍‌⁠‌‍​⁠‌‍‍‍​⁠‌‌​‍​‍​‍⁠​​‍​‍‌‍‍⁠​‍​‍​⁠‍‍​‍​‍‌⁠​‍‌‍‌‌‌⁠​​‌‍⁠​‌⁠‍‌​‍​‍​‍⁠​​‍​‍‌‍‍‌‌‍‌​​‍​‍​⁠‍‍​⁠‌​​⁠‌‍​⁠‌‍​⁠‌‌​‍⁠​​‍​‍‌‍‌​​‍​‍​⁠‍‍​‍​‍​⁠​‍​⁠​​​⁠​‍​⁠‌‍​⁠​​​⁠​‍​⁠​‌​⁠‌‌​‍​‍​‍⁠​​‍​‍‌‍‍​​‍​‍​⁠‍‍​‍​‍‌‍‌​‌‌‌​‌‍‌‌‌⁠​⁠‌​‌‌‌‍​⁠‌⁠‌‍‌‌​⁠‌​⁠⁠​⁠‍​‌​‍⁠‌‌‍​​⁠‌​‌​​‍​⁠‌⁠‌⁠​​​‍​‍‌⁠⁠‌