Here is a great interview Dave Girouard from Upstart:
https://www.buzzsprout.com/1697077/8693359-investing-in-ai-e…
This was amazing. I can not recommend enough that everyone check this out.
Some notes that are very rough as I listened last night and this is from memory:
-
This guy really knows the development process and how the tech works. Really well. And he speaks about it in a very accessible way I think.
-
They don’t want to slowly pick up more verticals. They want to get in and be first-mover so expect expansion here
-
They still originate a tiny amount of loans (3%?) but only because they need to prove the tech works on their own dime before rolling it out to everyone else.
-
He speaks to why large lenders with tons of historical loans can’t just build a training set and make competitive machine learning models. I like his analogy of self-driving cars and comparing Tesla to Ford (I think it was). Ford has been around a long time and has sold vast amounts of cars, but Tesla is gather the exact data it needs to inform ML models whereas Ford’s data would not be geared for this until recently if at all. This is all in my own words but the gist is that the inputs to the model are what makes a model work so you have to iterate on what data is important…and it takes a LOT of data to do anything.
I doubled my holdings in UPST today. I planned to add anyway but this interview really made me go for a few extra %. This is my 4th largest now and I trimmed the largest 3 for personal reasons that say nothing about my conviction, and about 1/3 of my SHOP to fund this purchase. I haven’t shared in some time so here is what my new portfolio allocations look like:
1. Crowdstrike CRWD 19.7% -15% *(buy/sell change as a % of holding)*
2. Cloudflare NET 16.0% -20%
3. Datadog DDOG 15.8% -16%
4. Upstart UPST 13.3% 140%
5. Snowflake SNOW 11.0% 0%
6. Lightspeed LSPD 8.4% 0%
7. Shopify SHOP 5.6% -33%
8. Zscaler ZS 5.6% 0%
9. Twilio TWLO 5.4% 0%
Hi RafesUserName,
thanks for the Information. This looks like a very well designed Portfolio. I’ve also upgraded my portfolio on Upstart the last days.
“And he speaks about it in a very accessible way I think.”
The sentence stuck with me and gives me great confidence. I gave up some positions in Zoom (I still had small trial positions in Square). And I also sold Bandwidth today. I invested the money in Upstart.
Seems like Upstart is just providing technology to connect lender and consumer with the appropriate rate given the risk profile. It is asset light model (like Visa).
If the loan is in default, Upstart does not bear any risk since it is not putting up its own capital (like American Express).
Do I have this right ?
Newby here,
Thanks for posting this interview! I listen to a lot of AI podcasts, and this one gave me an inside view of Upstart’s mindset and prospects.
The part I found particularly interesting was Ceo’s discussion on moving the algorithms into new areas; in particular, he mentioned insurance. This piqued my ears as an investor in Lemonade. Sounds like Upstart is planning entering this market in the near future and if they take a similar approach to insurance as they do loans, IE using their algorithms for established players this could majorly disrupt Lemonades chances at market penetration.
The whole competitive advantage Lemonde has in insurance is there algorithms, and that big player can’t catch up. If Upstart comes in and applies their tech to me, it’s game over, Upstart wins, and Lemonade is finished.
Thoughts?
Hi Gbrown - essentially I agree and it is why I sold my entire Lemonade holding (along with dissatisfaction with their headline growth rate), whilst never a serious consideration for Root and put the whole lot and more into Upstart. If I had more capital I would add more at these levels.
Ant