I don't question Z1P's bad debt. In the past I've tried to...

I don't question Z1P's bad debt. In the past I've tried to rationalize why there are differences between Z1P's bad debt and that of APT as I have a background in quantitative credit risk modelling. I've explained that it's normal for Z1P to have higher default rates due to ZipMoney being a higher risk product. With ZipMoney making up less and less of Z1P's portfolios and with the acquisition of Quadpay which doesn't offer this kind of product, the default rates should continue to improve.

The beauty of traditional "Pay-in-4" type of products is that they have almost no risk (comparatively speaking) due to the low amounts and the fast repayment period. The more time passes, and the more behavioural data Zip (and peers) collect, the "cleaner" their books become. Over time this "self-cleansing" has a positive impact on margins.

PS: I never made the assumption that 50% volume would be cannibalized. What that example demonstrated is what would happen if 5 of the next 10 customers pays a lower margin. Cannibalisation would be to assume that existing customers would pay less margins, in my example that would drop the margin from 4% to 2%, instead of 3%. But anyway, this is just for demonstration purposes, and not meant to be realistic.