Upstart - potential headwinds

Investment Analysis Clubs Saul’s Investing Discussions
1

Thanks to the great discussions in this board, I’m slowly dipping into buying growth stocks (currently own crwd and upst). I bought Upstart a few months back and added last week after earnings for a cost basis of $145.

There has been a lot of discussion on all the positives of Upstart, so I will focus on a couple of areas to watch for.

  1. Personal loan market share

As per an earlier post, the personal loan market is 84B per annum. Let’s say it’s $100B (there’ll be growth due to ease of lending from platforms like Upstart). In Q2, Upstart originated $2.8B, ~11% market share. It’s clearly very impressive and the main reason we’re all bullish.

With just a handful of bank and CU partners, I’m curious to understand how they were able to get such a big share. I would think there are 100s of lending entities, did Upstart manage to enroll a few big fish as early partners? Maybe someone with more knowledge of the industry could shed light into this.

Also, realistically they can only get to 20-30% of the personal loan market. Growing 20% sequentially will get them to this share in a year. For reference, they grew sequentially at 60% and 35% the last 2 quarters, so the market might consider even a 20% growth as a slowdown. Are these good assumptions or do I have some of my numbers wrong?

As I was researching this, another concern popped up. Apparently fintech’s account for 40% of the personal market and banks/CUs are ~50%. Fintechs likely won’t use Upstart and this makes getting 20-30% overall share even more challenging.

  1. Auto loan market opportunity

From CFO in the cc (answering the last question) - “But I think our expectation remains that the loans are clearly bigger and so there will be a higher dollar revenue per loan. And we’re expecting probably a similar dollar contribution profit per loan as we see in personal loans.”

We have been excited about the auto opportunity as the size is 5-6x personal loans. The above comment makes it clear Upstart’s fees would be lower. The CFO says they’re looking for same dollar revenue per loan which is concerning. If the average auto loan is 5-6x larger than personal loans, Upstart’s revenue from this market would be similar to personal loans.

The profile of the typical borrower is also very different. Personal loans are mainly for people with not so great credit, very high interest rates (20%!), rate spread between lenders is high and enables a platform like Upstart to bring efficiency, value and lower rates (win-win for everyone). Auto loans are dominated (in dollar terms) by people with better credit. Also, dealers have agreements with lenders for pre-approved rates - I have never needed to pay more than 2-3% so far. Upstart can still optimize rates for borrowers with bad credit and higher rates, but that market size is smaller.

Combining the above two makes me believe the overall auto opportunity (where they can actually win and their profit share of that business) would be smaller than personal loans.

Looking forward to thoughts from the experts who have looked at Upstart lot more closely.

-T2SP

34 Likes
2

With just a handful of bank and CU partners, I’m curious to understand how they were able to get such a big share. I would think there are 100s of lending entities, did Upstart manage to enroll a few big fish as early partners?

I wrote in my 10Q post: https://discussion.fool.com/upst-10q-34901669.aspx which shows that in the first 6 months of 2021, Cross River Bank and Finwise Bank together accounted for 85% of Upstart’s revenues.
Recall that these two banks are the primary lenders that sell the majority of their originated loans underwritten by Upstart to 150 institutional investors in securitized trusts.

The other 23 current bank partners account for 15% of Upstart’s remaining revenue in the first 6 months of 2021, and these banks/credit unions typically hold the loans on their own balance sheets (though Customers bank does sell some to institutions too)

Growth right now is being powered by direct traffic to Upstart.com which is mostly funded by CRB/Finwise, and not from the other bank partners.
Bank partnerships are a slow growth engine for Upstart right now, but hopefully they will reach an inflection point in several more quarters, if large enough banks keep signing up (Associated Bank, recently announced, is a big win)

As I was researching this, another concern popped up. Apparently fintech’s account for 40% of the personal market and banks/CUs are ~50%. Fintechs likely won’t use Upstart and this makes getting 20-30% overall share even more challenging.

I’m not sure I understand this concern. Can you elaborate?
Keep in mind that:
The personal loan origination market is about $92 billion annually.
The outstanding balance of personal loans in the United States was $151 billion in the first quarter 2021.
They are very different terminology/concepts. Is that where the confusion is coming from?

Many fintechs, including Upstart, mostly just underwrite and +/- originate the loans, and then let banks or institutional investors carry the balance. So why would Upstart ever have to partner with fintechs??

The profile of the typical borrower is also very different. Personal loans are mainly for people with not so great credit, very high interest rates (20%!), rate spread between lenders is high and enables a platform like Upstart to bring efficiency, value and lower rates (win-win for everyone). Auto loans are dominated (in dollar terms) by people with better credit.

Borrowers with subprime credit on average pay much higher interest rates. Borrowers with credit scores from 501 to 600 paid an average APR of 17.26% on used-vehicle loans in the first quarter 2021 per Experian. Borrowers with credit scores 781 or higher paid an average of 3.71%.
Yes, Upstart is not going to make money from lowering the already low 3% rates of super prime folks, but there is clearly massive room to capitalize on creditworthy folks with below-prime scores.

According to the New York Fed, auto loan originations were $202 billion in 2021 Q2. That could be an estimated $800 billion originated in a full year. Of that $800 billion, there is huge chunk of inefficient loans with high interest rates to target, from FICO scores 300 to 660 range:
https://images.cutimes.com/contrib/content/uploads/sites/413…

So, I believe the realistic auto loan origination addressable market for Upstart is likely 40% of the total $800, at perhaps $320 billion. This is roughly 10 times the size of the (realistic) personal loan origination market of about $30-40 billion annually. The average personal loan originated by Upstart is $10K. The average auto loan is $20-30K in size. So the total amount of “dollar contribution profit” Upstart can grab in auto origination is 3-4 times that of personal loans.

There are also those with bad credit who are currently priced out, even though they have plenty of demand to buy cars. With Upstart, the hope is that they can ‘recreate’ the supply to the non-prime market that doesn’t exist right now, by allowing banks to confidently supply the loans with fewer defaults (i.e. Upstart’s AI helps banks to stop rejecting creditworthy borrowers who just happen to have a bad FICO score.)
Per Experian, subprime (defined by them as credit scores 600 or below) accounted for just 17.75% of total loans and leases originated in the first quarter 2021. That’s a steep decline from 22.31% a year ago; so there’s that missing “5%” where Upstart can add them back to the auto market.

But. The big kicker here is the additional auto refinance market. Recall that all of the above is referring to the auto loan origination market. The auto refinance market is targeting the entire outstanding auto loan balance market which is HUGE at $1.42 trillion as of Q2 2021. If we assume 40% of existing auto loan balances are at unreasonably high interest rates, that’s an additional $568 billion in auto loans to attack. This makes a lot of sense as to why Upstart is super aggressive on the auto-refinance front right now.

I don’t think my back of napkin math is going to be very precise, but I hope I’ve made it clear that the realistically addressable auto loan market (in terms of potential profit for UPST) is indeed far larger than personal loans.

https://www.newyorkfed.org/newsevents/news/research/2021/202…

https://www.businesswire.com/news/home/20210527005217/en/The…

106 Likes
3

This2ShallPass,

My apologies. I think I poorly articulated a portion of my reply to your post. My revision is below:

As I was researching this, another concern popped up. Apparently fintech’s account for 40% of the personal market and banks/CUs are ~50%. Fintechs likely won’t use Upstart and this makes getting 20-30% overall share even more challenging.

The originating banks partnering with fintechs do not actually have to stick with one fintech company. For example, Cross River Bank deals with Upstart, Upgrade, Marlette, etc and they all compete with another. Banks don’t have to be exclusive.

And even though fintechs are already originating nearly half of new personal loans, that shouldn’t matter at all as long as there is unrelenting demand from banks/institutional investors to carry Upstart’s superior underwritten loans on their balance sheet.
Remember that creditworthy borrowers are not beholden to a particular bank or fintech brand, they just want the lowest interest rate loan and a frictionless application process. No need for UPST to partner with fintech competitors.

(I’ve written about how the “lowest cost to consumer, wins”, in a previous post: https://discussion.fool.com/how-upst-differs-from-fintech-compet… )

10 Likes
4

Thanks for the thoughtful response jonwayne235.

“Growth right now is being powered by direct traffic to Upstart.com which is mostly funded by CRB/Finwise, and not from the other bank partners.”

Yes, this makes sense. This would equate to ~10% share of the personal loan market. I guess you could also argue since these are new channels, maybe most of this is expansion of the total addressable market (TAM), which is even better news for Upstart. There are two ways Upstart could grow in the personal loan space

  1. Grow their market share. Currently at 10% and very fast growth due to their strengths. How high do you think they can go? I think it would be hard to go above 30%…
  2. Overall market / TAM grows. It’s ~$85B per year today and for my calculations I assumed 100B. This would grow relatively slowly.

To put some numbers, let’s say Upstart continues it crazy fast market share growth and reaches 30% in 2 years from now and TAM grows 10% per year. Upstart loan originations by year would be

2021 - $11.5B (from guidance. TAM=100B. Market share = 11.5%).
2022 - $22B (TAM = 110B, Market share = 20%). YoY growth = 90%
2023 - $36.3B (TAM = 121B, Market share = 30%). YoY growth = 65%
2024 - $40B (TAM = 133B, Market share = 30%). YoY growth = 10%

“And even though fintechs are already originating nearly half of new personal loans, that shouldn’t matter at all as long as there is unrelenting demand from banks/institutional investors to carry Upstart’s superior underwritten loans on their balance sheet.”

It would matter for what’s the market share they can realistically get, since this means Upstart is only playing in half of the market.

The overall personal loan market is ~50% Bank / CUs and 40% fintechs . Link: https://www.supermoney.com/studies/personal-loans-industry-s… (you have to scroll to the middle section for this data. If there was a way to attach images to posts, it would be much easier).

To reach 30% of overall personal loan market share, they would have to get 60% of all the Bank / CUs loan origination! In my opinion, they won’t be able to get much business from other fintechs as they pride in their own technology capabilities.

“This is roughly 10 times the size of the (realistic) personal loan origination market of about $30-40 billion annually. The average personal loan originated by Upstart is $10K. The average auto loan is $20-30K in size. So the total amount of “dollar contribution profit” Upstart can grab in auto origination is 3-4 times that of personal loans.”

I was going with 5-6x personal loan market I saw in other posts. $30-40B seems too low, I was assuming 85-100B.

Average personal loan of $10k is correct per the Q2 report, this was a mistake in my initial post as I assumed it was closer to $5k. ~35-40% addressable auto market is also realistic (Link: https://www.businesswire.com/news/home/20210527005217/en/The…)

Assume 800B yearly auto loan market (another post said it was $650B) and $25k average loan, $100B personal loan market and $10k average loan. If Upstart targets same revenue from each auto and personal loan as per the CFO, then yes auto market would be ~3x bigger.

5 Likes
5

The overall personal loan market is ~50% Bank / CUs and 40% fintechs . Link: https://www.supermoney.com/studies/personal-loans-industry-s…… (you have to scroll to the middle section for this data. If there was a way to attach images to posts, it would be much easier).

The banks, Upstart or any other player - none have a lock on the personal loan market at 50% or 40%. Even assuming the numbers are correct, these numbers at best simply state where the end consumer today gets their personal loans from.

If Upstart offers the best rates, more consumers will gravitate towards them. They will gain market share over others. The current split does not matter. The numbers reflect this is likely what is happening.

Neither is the size of the personal loan market constant. Personal loan market is a subset of larger personal credit markets. If a personal loan works out to be cheaper than credit cards for their scenarios, more will opt for personal loans.

2 Likes
6

This2ShallPass,

I was going with 5-6x personal loan market I saw in other posts. $30-40B seems too low, I was assuming 85-100B.

I conservatively used 30-40B as the realistic TAM, because only 42% of the personal loan market is “below prime”. But now I just came across some info that makes me wrong on that assessment and you are right. UPST should be able to attract borrowers all the way into the prime bracket based on the following data points of interest:

Average interest rates by credit score bracket
760-plus 9.30%
720-759 13.32%
680-719 17.82%
640-679 22.16%

And huge loan pricing dispersions by credit score bands equates to big opportunity for UPST, all the way up to the 740 FICO bracket:
https://d15584r18i7pqj.cloudfront.net/wp-content/uploads/202…

We should also keep in mind, if origination fees are charged, the APR is going to be much higher than the interest rates shown above.
Based on KBRA data, UPST often gives average APR of 22% to an average FICO pool of 670. So if we assume origination fee of 8%, that could mean interest rates of 16% were offered by UPST on average.
That’s a huge discount to a competitor who charges 22% interest rate and 8% fee, amounting to 28% APR to the same borrower!

Plus UPST does have some traction in the super-prime market as long as they get bank partners that don’t charge origination fees.
If you apply directly through their partner Associated Bank which I tested myself, you are not charged an origination fee versus if you went to Upstart.com and got a loan funded by Cross River Bank; this seems to be the experience based upon what some people have shared with me directly.

9 Likes
7

Some fantastic posts by both This2ShallPass and JonWayne235!

This2ShallPass: To put some numbers, let’s say Upstart continues it crazy fast market share growth and reaches 30% in 2 years from now and TAM grows 10% per year. Upstart loan originations by year would be

2021 - $11.5B (from guidance. TAM=100B. Market share = 11.5%).
2022 - $22B (TAM = 110B, Market share = 20%). YoY growth = 90%
2023 - $36.3B (TAM = 121B, Market share = 30%). YoY growth = 65%
2024 - $40B (TAM = 133B, Market share = 30%). YoY growth = 10%

I love when people “put some numbers” to the page – they’re inevitably inaccurate (mine especially!) but they give us somewhere to start. I think these numbers at least directionally agree with JonWayne235’s estimates for the Personal loan opportunity. And quite frankly if Upstart is still living on Personal alone in 2024 I think we’d all agree the thesis would be broken.

This2ShallPass: Assume 800B yearly auto loan market (another post said it was $650B) and $25k average loan, $100B personal loan market and $10k average loan. If Upstart targets same revenue from each auto and personal loan as per the CFO, then yes auto market would be ~3x bigger.

Again we can quibble with the specific numbers (is it 3x or much more?), but I think this again agrees directionally with what JonWayne235 is saying, i.e. that auto is a much bigger opportunity than personal. So maybe the numbers could look more like this (don’t ask how I got the specific Auto numbers…I just made stuff up):

2021 - $11.5B from Personal(from guidance. TAM=100B. Market share = 11.5%).
2022 - $22B from Personal (TAM = 110B, Market share = 20%) and 5B from Auto. YoY growth = 135%
2023 - $36.3B (TAM = 121B, Market share = 30%) and 20b from Auto. YoY growth = 109%
2024 - $40B (TAM = 133B, Market share = 30%) and 60b from Auto. YoY growth = 78%

Again, my numbers will be spectacularly wrong in the actual, but directionally, this is the thesis.

Plus, this is just Personal and Auto. If we really believe in what Upstart is doing, and what management is saying, they won’t stop there.

Bear

50 Likes
8

Plus, this is just Personal and Auto. If we really believe in what Upstart is doing, and what management is saying, they will not stop there. Bear

Hey Bear, they are advertising to hire a general manager of mortgage!!! Which is an enormously bigger market.

https://www.upstart.com/careers/10392/apply?gh_jid=3383019

Maybe jonwayne can answer this one: I think I remember reading that they receive an itsy bitsy fee every time a payment on a loan is made by the lender. Is that correct? If so, it would mean 20 to 30 years of recurring revenue on every mortgage they placed.

Best,

Saul

15 Likes
9

Correction: I followed that link and it was no longer active so I would assume that they have filled their general manager of mortgage position.
Saul

10 Likes
10

I think I remember reading that they receive an itsy bitsy fee every time a payment on a loan is made by the lender. Is that correct?

Yes, for unsecured loans we know they collect servicing fees.

KBRA report tells us how it is determined. It is calculated as the product of:
(i)0.80% per-annum;
(ii) average daily aggregate principal balance of the loans during the collection period;
and (iii) 1/12 (for the first collection period the actual number of days in that period over 360).

For full year 2020 they had $28.34M in net servicing fee revenue.

For first 6 months of 2021 that has already jumped to $27.434M in net servicing fee revenue.

If they somehow structure a similar fee for servicing mortgages…it would be a quite significant contributor to the top line due to the sizes of mortgage balances. But that would be very speculative at this point

26 Likes
11

I looked on Upstarts website and they are looking to fill a bunch of positions! https://www.upstart.com/careers/open-roles
What caught my eye was their openings in Operations both in Onboarding and Servicing. I am sensing their rapid growth is making them move quickly. They have alot of positions to fill! I also noticed that they are filling jobs in San Mateo, CA, Austin TX, and Columbus OH. Many jobs are remote but what was interesting alot of new jobs are being directed to their Columbus OH site that opened in January 2019. I found this on an Upstart blog:
“We’re excited to announce that Upstart will open an R&D center in Columbus, Ohio in January 2019. This second home for Upstart will include teams dedicated to software engineering, data science, and operations and is expected to house more than 100 employees by year end. We’re also pleased to announce that Grant Schneider, Upstart’s longest tenured data scientist and triple-degreed graduate of Ohio State University, will lead the Columbus office.”
I could be wrong but I sense this companies runway of growth is just getting started.
EPM

8 Likes