Nov 10, 2022·edited Nov 10, 2022Liked by Brad Freeman
This is a company that I was initially quite impressed with (many quarters ago). But they have continued to mess up with the execution and fiscal management....bad decision after bad decision...quarter after quarter.
As tech investors, we like investing in up and coming businesses who are trying to disrupt existing models and deliver for the consumer/customer/client. Upstart had that promise early on and who know, they might eventually get to the promised land.
Right now, they are at the bottom of the hill, a very long and tall hill, perhaps even a mountain...
As a stock investment, this has been a horrible investment for, well, everyone.
As an investment in a business though, I plan to remain invested for the longterm (years) so long as Upstart:
1) keeps signing up new partners for the platform,
2) is more accurate in rating individual borrowers than FICO, and
3) develops a viable product for all types of consumer lending.
This quarter showed progress on two of these three areas for me.
Although you note that the quality of the AI model has improved, I know I have a problem with trying to assess the quality of the underwriting. I am no expert and have no access to the data necessary to make a determination. The data given to us by Upstart as to its accuracy is self-selected. And, as we have seen, even if the data indicating Upstart's model superiority is true, that superiority is not translating to revenue increases.
And, at least for me, exacerbating the problem of thinking about quality of the AI is the fact that the model is not static - it changes. That implies to me that only a much longer time horizon over a more diverse set of economic conditions and hindsight will really show whether Upstart built a better mousetrap.
So, I have looked to the adoption by banks/credit unions as a proxy that Upstart's AI is superior. At least to an extent. Do you think that Upstart's increase in the rate of signing up lenders is a validation of the accuracy of Upstart's rating ability? Or is the increase in adoption more a validation that Upstart can provide an automated lending platform for banks and credit unions who otherwise wouldn't develop one on their own?
Hi Brad, what do you think will be the impact to Upstart with the recent falling bond yield? Will they benefit? Not sure will this help them to sell their loan at a better price off their balance sheet. Any thought?
This is a company that I was initially quite impressed with (many quarters ago). But they have continued to mess up with the execution and fiscal management....bad decision after bad decision...quarter after quarter.
As tech investors, we like investing in up and coming businesses who are trying to disrupt existing models and deliver for the consumer/customer/client. Upstart had that promise early on and who know, they might eventually get to the promised land.
Right now, they are at the bottom of the hill, a very long and tall hill, perhaps even a mountain...
Thanks again for the update and thoughts, Brad.
As a stock investment, this has been a horrible investment for, well, everyone.
As an investment in a business though, I plan to remain invested for the longterm (years) so long as Upstart:
1) keeps signing up new partners for the platform,
2) is more accurate in rating individual borrowers than FICO, and
3) develops a viable product for all types of consumer lending.
This quarter showed progress on two of these three areas for me.
Although you note that the quality of the AI model has improved, I know I have a problem with trying to assess the quality of the underwriting. I am no expert and have no access to the data necessary to make a determination. The data given to us by Upstart as to its accuracy is self-selected. And, as we have seen, even if the data indicating Upstart's model superiority is true, that superiority is not translating to revenue increases.
And, at least for me, exacerbating the problem of thinking about quality of the AI is the fact that the model is not static - it changes. That implies to me that only a much longer time horizon over a more diverse set of economic conditions and hindsight will really show whether Upstart built a better mousetrap.
So, I have looked to the adoption by banks/credit unions as a proxy that Upstart's AI is superior. At least to an extent. Do you think that Upstart's increase in the rate of signing up lenders is a validation of the accuracy of Upstart's rating ability? Or is the increase in adoption more a validation that Upstart can provide an automated lending platform for banks and credit unions who otherwise wouldn't develop one on their own?
T.
Hi Brad, what do you think will be the impact to Upstart with the recent falling bond yield? Will they benefit? Not sure will this help them to sell their loan at a better price off their balance sheet. Any thought?