I thought the board might be interested in my recent interactions with Upstart, as an automobile dealer looking to partner with them. As an upstart investor(thank you Saul)and since they are beginning to assist banks with auto loans, I reached out to their business development team, who was kind enough to return my call. They are currently piloting their loan platform to a large dealer group (they were not able to give me their name at this time) and I have asked to be one of Upstart’s beta testers in my market
(stay tuned). The Upstart team, that is working with automobile dealerships, set up a Zoom call with me and I couldn’t be more excited by what was discussed. Upstart is clearly looking to disrupt the outdated and inefficient 30 year old auto loan approval process, and make it a win-win win for the lender, the consumer, and their dealership partners.
The focus is not necessarily just for subprime customers, but also for those individuals who may have no score or a “misleading” FICO score, as a result of limited credit. The current bank loan approval process (utilizing FICO scores) does not take into consideration people that pay rent, may have bank or brokerage accounts with a reasonable balance, individuals that may have a just graduated college, etc. Upstart’s AI platform will presumably take all things relevant into account and make lending decisions that are completely outside of the current world of automotive lending, that only relies upon FICO scores. I deal with this situation every day, and I can really see where this is going. As the folks from Upstart indicated in our Zoom call, their move away from FICO reliance should mean lower APR’s for consumers, instant approval with next day funding, and a substantial increase in loan approvals (they are hoping for up to a 25% increase). As an automobile dealer, what’s not to like about this? As a consumer, what’s not to like about this? As a lending bank, what’s not to like about this?
If interested, I will keep you apprised of my business experience with the Upstart platform as things progress. I am in the process of using one of the Upstart products that is offered as a result of the Upstart acquisition of Prodigy Software. The excitement of the Upstart management team, that I have been speaking to, is nothing short of contagious. I will be looking to add to my Upstart position.
I didn’t want to clutter the board with a short, obvious post, but in my opinion (and the 84 recs so far), this is very valuable information. Much more so than the experience of a single person. It gives us a systemic look at how they are rolling their product out to, if not all, many dealers. As well as insight into their vision for the product with automobiles and beyond.
I would very interested in further updates as they become available.
Thank you again,
Mike
P.S.: Increased UPST position with the latest downdraft.
geoglassman, Thank you for the inside info. I’m a veteran of the Car Wars lol (1987-2001, still have many ties to it), and curious if you operate a new car franchise or strictly used? Since new car dealers are “somewhat” beholden to their parent lender (Ford-Ford Motor Credit…GM-GMAC…Honda-Honda Finance, etc), would FMCC or HFCA work with UPST at the dealer level, or would you be using it with your secondary lenders? Also, I’m curious as to whether UPST would be working in the leasing application process as well? Many sub-prime candidates are shut out of leasing a new car due to outdated underwriting issues, but would still make great leasing candidates via UPST.
Dealers should be salivating over this generational change to consumer lending, as so many buyers walk out of dealerships unapproved for new/newer car loans, only to be swindled by the sub-prime sharks (no offense intended lol). Once Dealer B hears that Dealer A has a new tool in the toolbox (UPST), it doesn’t take long for them all to follow en masse.
I have many questions, and looking forward to your updates. Good selling!
Thanks for you comments. We have several new car franchises as well as used cars. You are correct that there will be some reliance on the captive lenders but they seem to be coming on board with considering alternatives to FICO scores. Leasing seems to be another issue to be addressed down the road. Upstart is already embedded with many dealerships as a result of their acquisition of Prodigy. It seems that once Upstart shows that they have a platform that makes the process easier for the customer, the lenders and the dealers (which it does) and is an improvement over the utilization of FICO scores, use of Upstart’s platform will become the norm.
The question will obviously be how long and how easy will it be for others to replicate or improve on Upstart’s AI. I think as first mover, and given its early traction, Upstart’s platform will become the better way to determine who is or is not credit worthy.
Geoglassman: Thank you so much for providing insight into the auto business side of Upstart.
Many large employers and other groups of people have credit unions that (in my area) offer pretty good rates on car purchases. Other buyers come with pre-approved bank loans and yet others take advantage of manufacturer finance deals.
Do the manufacturers or credit unions also use Upstart (or Prodigy) as a credit measure to gauge risk of buyers?
Or asked a different way: Can you give us a rough estimate of the percentage of your car buyers who are candidates for Upstart financing versus other options? If half the people in your dealerships are likely to bring money into Upstart, that would be very interesting to us as investors.
You bet, keep me informed. The break-out in the auto loan field is pretty close to a brilliant move.
I’d like to suggest an “out-there” possibility, that UPST should consider:
Applying their current model, why not look at first mortgage home owners (or those more captured), with low equity, good to better payment history, who are paying a high interest rate on long-term (ie,30 year) loans?
Look for those in need of refinancing but are held hostage to the current system of evaluating viability.
Disrupt and free! That’s a pretty good mantra. Pretty good for a lot more.
Applying their current model, why not look at first mortgage home owners (or those more captured), with low equity, good to better payment history, who are paying a high interest rate on long-term (ie,30 year) loans? Look for those in need of refinancing but are held hostage to the current system of evaluating viability.
That point was actually addressed - at least somewhat obliquely - in the FY21Q1 presentation (slide 6). Basically, their long-range plan is to go from unsecured loans ($92B) to auto loans ($626B) to mortgages and credit cards ($3.4T). https://ir.upstart.com/static-files/ac83e54b-34ba-4f83-bdbb-…
Basically, they’re eventually going to take aim at every form of consumer lending. And that still leaves room for expansion into business lending and international geographies.
Thanks for sharing and I greatly look forward to hearing more about your experience with Upstart’s auto team/product.
Do you have any opinion about OpenLending, which is a direct competitor against Upstart in the near prime auto lending space?
Could Upstart have a better value add proposition given its consumer facing Prodigy product?
Would love to hear your take. Thanks again.
Copy pasting from past post about Openlending:
“The company provides automated risk based loan pricing and underwriting for credit unions in near-prime auto lending and refinance. They try to boost approvals rates and loan numbers for their clients. They collect fees/profit share for ‘certifying’ loans; they also partner with insurance companies to facilitate auto loan default insurance.
They recently began expanding out of credit unions, to now include captive OEM lenders.
They claim to have 20 years of data analysis and underwriting experience, including through the Great Recession.
They say they have over 2 million risk profiles established from their data (they do not use AI/ML).”
“The question will obviously be how long and how easy will it be for others to replicate or improve on Upstart’s AI. I think as first mover, and given its early traction, Upstart’s platform will become the better way to determine who is or is not credit worthy.”
I think that is the single most important question about UPST that I am still trying to get a better feel for. How sticky is their solution, and what are the switching costs?
Inevitably someone will come along with a hot new AI model in one of their market categories. Can a credit union or auto dealer easily “switch” at that point like swapping out spark plugs or is it such a tightly integrated set of business processes and ecosystem that the costs of switching are high?
I remember selling mainframe and AS/400 solutions for many years and every couple years there was the inevitable conversation with a customer that revolved around “we are leaving your platform for the latest and greatest mousetrap”. I would say about 97% of those conversations a year later became, hey what’s the best price we can get on the renewal because we’ve decided to stick with the platform for at least another 3-5 years. Why? Not because they felt it was the best platform for a brand new “greenfield” project. But for their existing applications, which took many years to develop and perfect and were tightly integrated and optimized into all their core business processes the cost of switching and redeveloping that platform from scratch was usually 5-10X more expensive than just cutting a check (even in the millions) and continuing with the status quo.
What always drove an actual switch was when the business found a new market and growth opportunity that required a unique technology capability that the old systems couldn’t easily provide. Then they could justify the costs based on some massive growth opportunity, but business as usual never could justify a switch.
Anyway sorry to ramble… I’m long and strong UPST, already a triple for me and I’m in love with management and the TAM. No chance I sell unless the business and growth materially change. The only thing that I wrestle with is trying to understand how sticky their models are when the inevitable new hotshots and their “quantum AI model” (making up some nonsensical hyperbolic name they will come up with) comes into the fight. A lot of money to be made here and you can bet the competitors are lining up.
Maybe there is some better insight and experience on the board that can answer this question?
I think the product of OpenLending is very different from Upstart. OpenLending offers an insurance for the lender (the car dealer), based on a customized risk profile. This insurance allows more subprime loans to be offered by the lender.
OpenLending does not sell loans to the customers. The customers don’t even know that the lender has taken out an insurance with OpenLending.
What does it cost to use Lenders Protection™? There are no upfront costs to start Lenders Protection™. When submitting applications to Lenders Protection™ there is the additional cost of a credit pull request and the costs to insure the loan against default.
Upstart’s competitive edge is not only the AI.
They are building the network of lenders, such as banks and credit unions.
The more banks and unions Upstart have on their platform, the more they compete for borrowers.
And the only thing they can compete on is the APR.
With time, Upstart will be able to offer better and better APRs for customers and to compete with Upstart, a competitor will have not only copy AI and the data, but also the network of lenders.
It takes time and effort to earn trust of such conservative organizations as banks, so it won’t be such an easy task.
It looks like a magazine spread advertising Upstart Auto Retail. The ad says “special offer, get 6 months free”.
Just curious, did they mention pricing on your call? Is it like a subscription monthly fee to use Prodigy? Per transaction fee?
Geo - great post. Question for you. How reliant is the Finance product that you are going to be Piloting on Prodigy? Their re-fi is not reliant on Prodigy ---- so I think that has a LOT of legs quickly. But I’m curious to your take on the reliance of Prodigy rollout ---- to take advantage of their AI and lending network.
One asked how many people don’t have credit but are low credit risks?
If any of you are expats, unless you kept your house and have been making mortgage payments while away you will come home to a Credit score of 0. Basically this is a score that says none or not enough recent information to give a score. There are millions of expats coming home which need to buy cars and a place to live. It takes 3-4 years in the US paying utilities and other reportable payments to reach mid 700 score even for people with higher than average balances.