Upstart Call Notes

I was a little disappointed by the numbers, and I’m trying to view the drop in stock price as a buying opportunity.

I listened to the call, and I was impressed. The prepared remarks were… a bit robotic, but Dave and Sanjay were very on-point during the Q&A.

Overall - personal loans are reaccelerating, and Upstart is onboarding banks and credit unions faster and faster. Auto loans are a very inefficient market, and they are going after refinance and origination. They are adding a new short-term loan through banks - small amounts paid back over a few months. Home mortgages was also discussed with a 2023 horizon. Some time and effort was spent in Q3 bolstering defenses to identify fraudulent loan applications.

Long version:

Personal loans via banks and credit unions are re-accelerating. They’ve grown from 10 to 31 partners between Q3 2020 and Q3 2021, and they are onboarding new banks quickly - the last partner was up and running in less than 50 days.

They are expanding this offering to prime and super-prime lenders, with attractive rates and no origination fees. Basically, expanding the range of potential borrowers.

Auto lending is six times the size of personal loans and it’s highly inefficient. Americans are paying too much for car loans, and Dave referenced the Consumer Reports report on auto-lending. They’ve made progress in simplifying care loans, and they are on track to grow this like they grow personal loans.

There are two parts to auto loans (new to me) - refinance, and origination loans. Refinance has acquisition costs, but auto lending and purchase loans done at the dealership have better economics for Upstart, and they will be meaningful next year (but Sanjay declined to offer specifics).

Two new product areas - small loans (a few hundred dollars) over a few months, under the umbrella of a bank, with capped interest rates. Banks are under pressure from The Fed to service lower and modest income families and this is one way they can do it - banks and credit unions are very interested. Upstart will be bringing this to market by the end of 2022.

Dave also talked briefly about installment loans - BNPL? Did anyone catch more about this - it was over pretty quickly.

Home mortgage market - it’s the largest consumer lending category. Large opportunity to improve lives of Americans. The 2008 financial crisis made it hard for Americans go get affordable mortgages - there were 1 million fewer mortgages in 2015 than there were in 2001. The home owner are credit worthy and should have access to a mortgage. It’s a longer horizon on this financial product.

Q&A:

Private investor:
Q: Guidance - raise is lower - 60 million from 150 million.
A: Triple digit revenue growth.
Follow-up: adjacent businesses in lending - mortgages, small $$ loans - what will the financial impact be of these over the next 2-3 years.
A: Move to inefficient markets, able to offer a better product that is profitable for Upstart. Not going to low-margin areas. Generally, strong margins. Nature of each product is different, but will drive top and bottom line growth.
A: Auto is in flight right now - refi - similar to personal loans (but higher costs - fee and aquaisiton costs). Contribution margin - funnel is and will improve over time). And auto-retail lending and purchase - those economics to look better - lower acquisition costs
A: Business and lending - nothing to report on those fronts yet.
Another follow-up - 2022, 2023 - what will the auto contribution to be.
A: Auto will be meaningful next year, but no specifics - too early

Barclays:
Q: Guidance - 4th qtr estimates is more than the q3 beat - he sees lower performance in q3 but building momentum in Q4.
A: Sanjay: what’s driving performance is improvement of models over time. It happens in fits and starts. Q3 was focused on fraud mitigation.
Follow-up: more color on fraud issue - contained? Ongoing, more impact next quarter?
A: Dave: fraud constantly, since the beginning, but Q3 saw more organized fraud, but no material financial cost, but resources had to be used to defend. In a better position. Cost of doing business.

Goldman Sachs:
Q: First, small dollar loan product - how does it fit into broader strategy. Will it help with personal loans?
Second, average loan sizes per quarter, how it will go into q4 and 2022.
Dave: Small dollar question: a win across the board. For consumers, not well served with affordable product that’s not exploitive. Upstart can do it right, better, short term loan via through bank - more affordable products.
Second, banks have pressure from Regulators to serve low and moderate income Americans. Banks struggling to do this. By doing this, Upstart profits, but also gets data for models - gets unserved people into mainstream financial system, helps the model. Later on, they could get a car loan, a mortgage, etc.
Sanjay: loan sizes - two things going on. Optical versus real. Optical is about portfolio mix - FICO guardrail removed, so borrower is receiving smaller $$ sizes as they are riskier, reducing loan size. But really, loan size has been coming down since end of last spring - call it suppressed loan demand, due to stimulus in economy. Normalization in economy w/regards to savings, credit balances, so a stabilization of loan sizes, followed by larger loans.

Jeffry’s
Q: First, the funnel. What channels and sources of new customers.
A: Dave: Good recurring customer momentum. Fast growth across all channels. Direct mail very productive.
Q: What banks and credit unions are coming on?
A: Dave: high level - lots of banks originating banks for their own balance sheets, other selling loans. Rapidly growing interest from bank partners. Heave in deposits, light in loans (good for Upstart). Improved ability to get banks on Upstart quickly - less than 50 days. Demand is there, and financial returns to banks have been “excessive” - higher than they would have expected. Default rates low. Banks and CU partners added at faster clip:
A: Sanjay: loans from banks to capital markets - short term dynamic is loss rates are unnaturally low right now. Lots of capital chasing yield, so demand for loans but demand will normalize. Sanjay believes that digital loans have been accepted, even in mainstream of finance. How does Upstart look in this market? Lots of credibility in ability to underwrite risk. When Upstart is past distortion in market, still lots of demand.

Nat, BoA:
Q: Platform for other loan types - personal and auto loans sold in ABS - well funded. Homes and mortgages sold into GSEs - how would the system work with Fanny Mae deciding who gets a mortgage.
A: Dave: mortgages funded in lots of ways, some qualifying, some not. GSE rules can change over time. Lots of ways this market could go. If poeple dont have access to a reasonable mortage at all, an opportunity. Funding may be from bank balance sheets, etc, and Upstart will be looking. Very confident better underwriting someone who previousy wouldn’t qualify, very large # of people not well served, and Upstart.

48 Likes

And please point out any mistakes I made - I was typing quickly as they were talking.

1 Like

I had very high hopes for this Q and am pretty bummed out. I was hoping we might even get a $400M revenue number this quarter, but I listened to most of the call and I just felt like the overall tone was quite positive. Dave Girouard just sounds very relaxed, like a guy who is building a long-term business and doesn’t feel he has to maintain the breakneck level of growth UPST was showing in recent quarters. I am tempted to buy more, but I already have a very outsized allocation. It was probably close to 40% at todays close as I kept adding today as it kept dropping.

5 Likes

Here’s what I really like hearing (emphasis mine):

“Second, banks have pressure from Regulators to serve low and moderate income Americans. Banks struggling to do this. By doing this, Upstart profits, but also gets data for models - gets unserved people into mainstream financial system, helps the model. Later on, they could get a car loan, a mortgage, etc.”

$UPST’s competitors will need DATA in order to feed their model…data that $UPST is grabbing before anyone else can get it! WHO ELSE is gathering data, and using it to create/improve/iterate AI models, at the rate that $UPST is?

This reminds me of $NET: their free CDN product provides $NET with DATA, data that their competitors don’t have, and can’t get, because $NET is hogging it for themselves! DATA that enables them to discover new ways to make money.

25 Likes

$UPST’s competitors will need DATA in order to feed their model…data that $UPST is grabbing before anyone else can get it! WHO ELSE is gathering data, and using it to create/improve/iterate AI models, at the rate that $UPST is?

While UPST has a head start in leveraging vast volumes of data to build AI models, many lenders and banks have a lot of their own data from years of lending, so perhaps the quantity of data is not the ultimate differentiator. It seems to me where UPST has a big advantage is in their experience developing the models, market testing them in the real world, and refining them over time based on measurable outcomes. That all requires a lot of data, I’m sure, but likely the key differentiator is that it requires a lot of trial and error until you get a proven, reliable, repeatable outcome that lenders can depend on consistently to minimize risk.

Once the models are market tested and proven to perform, you can automate them and resell them with minimal additional investment.

6 Likes