←back to thread

341 points shedside | 1 comments | | HN request time: 0.208s | source
Show context
shedside ◴[] No.20082086[source]
Don't know if anyone from Stripe is listening here but: it'd be amazing to be able to deploy chargeback protection selectively, as a Radar rule. So that for example we could say: charges from the US are protected (and subject to the extra fee) while charges from the UK are not.
replies(3): >>20082097 #>>20082123 #>>20082140 #
flibble ◴[] No.20082123[source]
That’s exactly what we’d like. 95% of our transactions come from trusted customers which have ~0% chance of a chargeback. We’d like to apply this check and insurance on only the transactions that we are not sure about.

Requiring it to be all or nothing makes the feature useless (not cost effective) for merchants with a low fraud rates.

replies(4): >>20082139 #>>20082212 #>>20082233 #>>20082242 #
shoo_pl ◴[] No.20082233[source]
Is it not the point? If stripe allowed you to do that - insure only dangerous transactions - then it could not be 0.4% but rather a few %.

The entire point behind insurance is to have enough volume so that the small % when they need to pay up won't affect the business.

replies(3): >>20082552 #>>20082588 #>>20085570 #
1. ben509 ◴[] No.20085570[source]
Adverse selection inflates premiums, simply because people who need insurance are the ones who buy it.

The point of insurance is to transfer* risk to the insurer. The insurer does that by identifying a group that is homogenous enough that their premiums are just slightly over the payouts.

So an insurer can improve competitiveness by selling multiple products that cover different risk groups. I imagine that for Stripe, the risk variance falls in a fairy narrow band: bounded at the low end by not being worth insuring, and bounded at the high end by merchants losing their account.

* As opposed to say retaining risk, e.g. you don't buy collision on a beater.