UPST: fraud/verification AI models are 'enou

Trying to squeeze a few things into one post, to avoid clutter on the board. The title of the post is really only relevant to the last part, so please scroll down to the bottom if you’d like.

Replying to JabbokRiver (…:slight_smile:
2. Perhaps they are nabbing the best and brightest as they graduate and they never leave Columbus at all. Maybe they are partnering with the university in some way. Might they have interns from OSU working at Upstart?

I think I’ve posted this back around July or so. Columbus Machine Learners talk by UPST:

Upstart is definitely trying to encourage OSU interest in joining Upstart with upcoming grads or new grads that meet the qualifications to join their machine learning team. I recall there were a couple of “columbus machine learning” events back in the summer, sponsored by Upstart. (WHich is where the above video was taken from.)

The head of Upstart Columbus is Grant Schneider, (who is also VP of machine learning) he helped pioneer Columbus as Upstart’s HQ2, as he spent 10 years at THE Ohio State University getting finance/accounting bachelor degree plus masters in statistics and then PhD in statistics - and then cold called Upstart and got interviewed and joined the team as a very early member, in 2014, which is also right after completing his PhD!
This decision to place HQ2 at Columbus is a very calculated move, I think exactly as you postulate. Management, and especially Grant Schneider with his personal experience in Columbus, is very well aware of the massive pool of high quality intellect and people to draw from at OSU (among other great things in the city). And Columbus isn’t saturated with that many other competing tech companies like in the CA area.
[Sorry Grant… if you are reading this, I am honored that you follow me on twitter, and I know I must stalk your company too much :slight_smile:

I think I’ve read every blog post on Upstart’s website, listened to every podcast available related to UPST, youtube videos, etc etc etc…There are just so many things that I keep observing over and over, all confirming that management is hitting the spot on all fronts, and especially on hiring/choosing the right people to join their team.
This company’s executive team is really a diamond in the rough.
Yes, a lot of other companies on this board are also stellar but I do think UPST is at least a notch above even the best around here.

In the recent podcast with Dave Girouard:…
"…one of the enormous strengths of Upstart, first of all is that we still have the three founders day-to-day active as well as a lot of the exec team has been with us since the beginning.
We survived a brutal pivot. I mean, this beautiful idea of an income share agreement, so perfect on the white board, came up against the cruel reality of the world, which essentially decided it was a niche product. So going through that pivot, keeping the entire company on board to do that, I think was a huge proof point for us. I don’t think we ever were ready to quit. I think we just felt that we were a good team, we were going to figure this out and suddenly we found our way into our market that there was an opportunity for us…So a lot of challenges like that.
We decided to go from one office. We decided to go from the Bay Area only to open up Columbus, Ohio, which was a very big deal for us at the time. So just things of that nature but most importantly is really having a team that’s been around the block together a lot. No matter what hits us in the face and there’s something new that hits us in the face probably every week, it is really comforting to know you have a team that has already worked through a lot of things together."

I would like you to contrast that with UPST’s competitor, Zest AI, which was established by ex-Googlers even before UPST and also focuses on AI credit underwriting. Zest’s founder plus cofounder and early execs have all exited, marred by or disillusioned with their initial scandal/unethical lending practices…(my past post about them:…)

Oh and to answer your question about interns, JabbokRiver, yes, Upstart has interns for sure, I’ve browsed through the linkedin profiles of their employees before… I know, it’s a lot of stalking, sorry, but this is what I must do with this company as my largest overweight portfolio position. Lots of interesting tidbits you can glean from doing so.
This reminds me, I’ve also noticed on the UPST career page ( ) they’re looking for someone with “minimum requirements” of “knowledge and/or experience with credit union member growth/requirements and credit union servicing organization (CUSO) opportunities” which goes along with that idea from the Kemba webinar where they were interested in forming this.
(However, it appears UPST is “lagging” a bit behind Zest AI on this front. Zest AI has a CUSO as of last month:…

Moving onto stuff related to the title of this post:

I know I’ve written many posts about UPST’s ability to lower default rates as its key advantage, but I’ve never really focused on its automation. It’s a shame because I’ve now realized, automation is probably playing a larger role in its extreme hypergrowth than anything else!

So, the other day I went through a list of US banks by asset size. I looked at their personal lending page (didn’t check auto) and didn’t find any with a white label Upstart product (although we won’t know for sure if any are already part of Upstart’s Referral Network, but not yet announced).

I went all the way down to rank 100 (and plan to look through the rest plus credit unions later)…

Here was the take away point I got from doing so:
NOT A SINGLE bank I came across for their personal loan page had anything remotely close to an efficient application process in the same manner as Upstart’s. Not a single one.

It’s crazy how conservative these banks are. They really are still stuck in the ‘decades ago’ idea of underwriting a potential new borrower with some kind of bottle necked non-digital, manual process. The vast majority want the customer to call the bank to even initiate an application. And even if it was an online application to fill out, the back-end of things appears still manual - you have to wait for a person to review things over a few days, maybe upload more documents, and then only would know if a loan is approved.

Here was the most hilarious example I came across:
At rank 84 by asset size, Investors Bank in New Jersey has 26B in assets. Here’s their personal loan app page:
If you click on “apply now”, this is what it takes you to:

Wow! So you first have to be a customer for a year, and then SEVEN pages with six different forms on it that you must print out, fill out in writing, and mail or fax to the bank…AND provide the following documentation:

To expedite the loan process please provide the following documents with the application packet:
? Copy of 30 days’ most current paystubs;
? Copies of all W-2 forms for most recent two years;
? If Self-employed – copies of most recent two years personal and business federal income tax
? If Retired – copies of the current year social security awards letter and/or proof of pension
? If Home Owner – copy of current mortgage statement; and
? If Renting – copies of 6 months’ cancelled checks front and back;
? Borrower(s) Living with Parents – provide evidence of current address (e.g. valid Drivers’ License
or valid State Issued ID Card and cell phone bill).

How crazy is that!? This is 2021, is this really how you are asking your borrowers, who have already been your own customer for a YEAR, to apply for a loan?? Shouldn’t your bank already be able to verify its own existing customers??
This highlights why exactly the most common Upstart review on Trustpilot is about a fast/easy application process. There’s far more reviews about the ease and speed, than about how UPST offers the lowest rate, for good reason!
I think Upstart has a huge opportunity in lending even if they didn’t have any kind of AI underwriting (default lowering) prowess.

Just their fraud detection and borrower verification AI model is ENOUGH to give a huge advantage for stealing loan market share for their bank partners.
The ability to automate loan approval processes instantly at >70% rate, without the borrower needing to call the bank, talk to anyone, upload a single document is HUGE! This is what the consumer demands. This really just cannot be understated.

And here is a 2019 article I just came across about FNBO which joined Upstart’s personal loan referral network (which was a trial run in 2019, before it became “the” Referral Network earlier this 2021)…
"First National Bank of Omaha (FNBO) announced the successful launch of their all-digital personal loan powered by Upstart. In just three months, FNBO’s program has provided convenient and affordable personal loans to more than several thousand borrowers in 45 different states, totaling tens of millions in loan originations.
More than 76 percent of FNBO’s new customers were approved instantly, with a median time from application to loan acceptance of just nine minutes. FNBO’s program has already delivered an exceptional level of customer satisfaction, with a Net Promoter Score (NPS) of 83."

76% automated, nine minute median application time spent, NPS of 83.
Upstart was way ahead of the game back in 2019 and is still building its lead today. Now I’m not saying that UPST is unique in its digitization/instant approval processes, as many other fintech lenders have similar capabilities (and this is the reason why fintech has been taking share from traditional banks for the last several years), but UPST is definitely far ahead of the pack!


There is little motivation for big banks to switch from present practices. Banks are lemmings and compare themselves to peers. Why else would they all rush toward NINJA type loans in 2008, or to earlier loans to countries like Mexico and Russia? Surely any rational outside observer would be dubious about trusting people who were broke and in many cases did not even have a job to pay back loans? Or trusting Russia , the country that continued massive germ warfare production after signing a treaty banning it ?
All that matters is to not stand out - risk is defined as doing something different from peers. If they make a mistake they will look like they met the industry wide standards. If the mistake is expensive enough they will be bailed out by the government. If a potential borrower goes to another large bank he will find very similar impediments.
The “standards” are set by a smallish oligarchy of large banks. All that red tape also looks good to regulators . Who are interested in banks not in borrowers
So while the present personal loan system seems broken to a borrower or an outsider, it is fine with most lenders.

Lenders are making lots of money using the present system. And are under little pressure to change. So Upstart tech will continue to come from smaller lenders.

System wide inertia will take time to overcome. Likely Upstart and lenders using it will have to market directly to customers. Still if it is indeed the better mousetrap that I suspect it is, it will continue to gain market share. The situation is not dissimilar to the gradual encroachment of BEV on the ICE car market.



I think you are right that Upstart - and the other fintechs doing personal loans - will continue to market directly to consumers. That won’t change. And I agree that banks are glacial in change. However, once banks see that a process or product is adopted by other banks, they all rush - as you said - like lemmings - to adopt that process or product. And that happens very suddenly and en masse. So when Upstart said they expect to have hundreds of banks in the near future, they are are “banking” on that lemming effect. We are just left wondering when it will happen.

As for personal loans, I agree banks have shied away because of the risk in the past and because other providers set up shops to market personal loans and take on the risk (HFC finance, etc.) that the banks didn’t want. And you are right that banks are fine with not providing personal loans.

But I will disagree that banks are under little pressure to change. The pressure will come from seeing other banks make a little more than they are. If we look to the past with the mortgages, none of the banks could resist the securitization market that was growing and they all got caught up in the frenzy because every other bank was making money making more and more loans. Now that the “fintechs” of the world created a process that takes personal loans to everyone through the internet and makes it easier to apply and removes human contact from the process, which in and of itself likely expands the market for personal loans because people don’t have to suffer embarrassment by applying for a personal loan in person, the market is bigger. (I don’t think that market is big enough to create a crisis, we are talking at most hundreds of billions in dollars in the personal loan market versus several trillions of dollars in mortgages.)

I think what is different with Upstart is that they are providing a process for banks to easily participate in the personal loan market. Banks can now get the fees their customers would have paid to others. Banks can either continue to watch the personal loan business be conducted by others or they can opt to have Upstart do all the work, sell the loans right away, and collect big origination fees.
The banks will simply see personal loans as Cross River does, a means to make profits through the origination fees. But the difference is that these banks will get those fees from their own customers, who would have otherwise paid those fees to someone else.



I am in the process of obtaining a secured loan from Bank of America. I have been a BofA customer ever since they bought SeaFirst (Seattle First). I was a SeaFirst customer ever since they bought Rainier Bank. I am a “Platinum” customer (whatever that means). In other words, I have a very long history with this bank. My FICO is around 800 plus or minus (it moves around a few points pretty often for reasons that are quite mysterious).

Despite the fact that I am a high net worth individual, I have personal reasons I won’t go into for seeking a loan. I thought it might be enlightening if I described the process from the customer’s POV.

I made an appointment with a loan officer at my local branch. After a brief meeting with him he told me I should have no problem getting a loan. He said he would send me an secure email in order for me to send back a number of documents. I needed to fill out a multi-page loan application form and return it with two months of brokerage statements, front and back of my driver’s license, two years of tax returns, and my trust agreement.

He in turn sent that package to a different department. I don’t know exactly what this department does, but they seemed to perform a sort of clearing function. They then sent my application to a loan processor. I don’t know exactly what this person does, but she requested a few more documents after which she sent the whole package off to “underwriting”.

To the best of my knowledge that’s where the application sits now; loan officer, clearing, processor, underwriting. Granted, it’s a different type of loan, but all the same, IMO that’s an awful lot of touch for a loan.


UPST CEO today announces:

“I’m thrilled to announce the promotion of Chantal R. to Chief Marketing Officer of Upstart!

Chantal joined us four years ago as a 25 year old individual contributor and it was clear right away that she was a special talent. Chantal grew very rapidly into what can only be described as elite-level performance marketing guru. It’s a rare consumer internet company that can grow by an order of magnitude while reducing acquisition costs, because it’s a rare company that has someone like Chantal leading the effort. Having been around Silicon Valley since Chantal was a toddler, I’ve seen some amazing marketing talent, but I’m not sure I’ve ever seen the combination of brilliance, analytic horsepower, and six-sigma execution wrapped up in one person.

And just as importantly, Chantal embodies the values that make Upstart tick. She is a team player, a community-oriented executive, and a pleasure to work with. Other than her choice to be a Toronto Maple Leafs fan, there’s never a reason to doubt a decision made by Chantal.

Hiring and promoting great leaders is an exercise in extrapolation - observing the past to develop confidence in the future. And I couldn’t be more confident in Chantal’s ability to lead Upstart’s growth and the development of our brand in the years to come.”