Dear Saul, greetings from Argentina. I am very appreciative of your forum, it’s full of wisdom and crowdsourcing. I wish one day I could learn enough to retire from my day job and live from my investments as you did. I am making a daily habit surfing through it, finding key discussions to learn from them. I appreciate the ones around you taking a new position and then exiting another, as well as your knowledge base, your end of months summaries, and mid months updates when relevant.
In regards to AFRM, I am curious about the BNPL thing, as it appears in the US you cannot buy and pay in installments almost anything with your credit card as it is customary to do in Latin America? If this is the case then credit cards have decades of experience running this type of business and tons of data, and funds to develop AI (or partner with UPST 
) so AFRM is really in big trouble. Major credit cards like Visa, Mastercar and AMEX do it, some times even with Govt funding.
After studying AFRM business model, I see they rely on applying AI to Data from all the stakeholders of the buying cyle: consumer, merchant, shopping cart contents, banks and data moving along the way across time (payments, etc). So from their AI they appear to make the real profit from taking better risk than incumbents (credit cards). But credit cards have more information about consumers, as you use your credit card for so much more than BNPL. The same holds for merchants (they sell surely more through credit cards than BNPL).
Yes AFRM appears to be more of a “platform” than a BNPL, and BNPL is just the first use case. However they are so heavy into BNPL that they are tying the brand to that. And believe me, the whole concept of Buy Now Pay Later sounds so irresponsible and risky, many people will just wont’t get into it. The more afluent, the less. So I see an important risk bias here also.
A key risk I can easily anticipate a is credit crisis with all this BNPL, as happened in South Korea (which estimulated the new Squid Game series), and everyone will take it with AFRM, as well as AFRM itself, as being the lender, would have to face massive losses.
Too many things need to work out right for AFRM to really succeed. And if it becomes successful, it will become so powerful, think of it as the Google of Retail Finance, so… will the incumbents and the new entrants allow it?
UPST and all of your SaaS companies have a much easier and simpler business model, no need to risk money yet on AFRM. I must confess I have a microsized position to watch it now, as you are doing with SNOW (mine even smaller), but I will load on other better conviction companies.
I am interested in the forum thoughts about this.