The stock for artificial-intelligence lending fintech Upstart Holdings (UPST -7.23%) has been on a monstrous run. Since going public at the very end of 2020, Upstart's stock price has risen to $327 and is up a whopping 643%.
Upstart uses AI and data to evaluate the true risk of a loan, and the company claims it can lower loss rates at U.S. banks by almost 75%. Currently, Upstart is in the business of originating personal loans, but it is also preparing to dive deep into the auto loan market. The company partners with banks to originate the loans and then sells those loans to institutional investors, or the originating bank might choose to retain the loan on its balance sheet.
This is clearly a strong company with great technology, superb growth, and profitability, which is rare for an up-and-coming fintech. But at this point, I think investors are getting way ahead of themselves. Here are three reasons why.
1. The (over)valuation metrics are obvious
I'm not telling anyone anything new here, but Upstart is currently trading at what seems to be an absurdly high valuation, whether you look at it on a price-to-earnings or price-to-sales basis.
Analysts on average expect Upstart to generate $1.33 in earnings per share this year, and management has told us to expect the company to do about $750 million in annual revenue for 2021. So for Upstart to trade at 50 times earnings, it needs to grow annual EPS by nearly 400%.
The good news is that analysts on average are projecting EPS to grow by nearly 30% in 2022, so it is growing the bottom line at a nice pace. And those are just estimates. Upstart could explode into the auto market and continue increasing bank partners and personal loans. But it's a lot to wager on so much uncertainty.
And while Upstart has been growing revenue a ton (roughly 60% year over year during the second quarter), management's guidance implies less growth in the third quarter. Management is guiding for revenue to be between $205 million and $215 million next quarter. At the high point, that would imply 11% year-over-year revenue growth in the third quarter, which is all well and good except when investors' expectations for growth are so massive.
2. Is there enough of a competitive edge?
Part of the reason Upstart has such lofty growth expectations is that investors really believe in the company's technology. Its goal is to one day replace the FICO score, a key part of the traditional bank underwriting process, which many believe is out of date and inaccurate. Bank partners that use Upstart's technology can set their own credit standards, but management on its most recent earnings call said one of its bank partners recently got rid of a minimum FICO score when evaluating a borrower's credit.
Management's goal is for more bank partners to follow suit and then maybe Upstart's technology becomes more and more crucial to the way banks underwrite loans across a broad spectrum of categories. But it's way too early to know if this will ultimately happen. Even if it did, wouldn't other companies eventually be able to build similar technology? JPMorgan Chase spent $10.3 billion on technology in 2020.
There's also not enough evidence to assume that Upstart's technology is superior to its competitors. Because the company sells all of its loans to third-party underwriters, it's hard to know how well the loans are holding up. There is reason to believe the results are very good so far as the company claims it can lower bank default rates by 75%. There's also the fact that the company is growing its number of bank partners. But we still haven't seen how those loans do in a higher interest rate environment.
Also, the fintech space is crowded with competitors trying to streamline the lending process. When describing its underwriting models in its 2020 annual report, Upstart said its models are informed by 1,000 variable points and 10.5 million repayment events, numbers that will grow as the company acquires more data. In its IPO registration statement, Upstart said its models are informed by 15 billion cells of data. Another popular company in the fintech space is LendingClub (LC -2.95%), which also uses technology to more effectively originate personal loans. LendingClub has been around longer than Upstart and uses models informed by more than 148 billion cells of data and more than 2,000 attributes from 74 million repayment events.
I'm not a tech expert, but looking at these numbers found in publicly available documents, it seems reasonable to think that there could be companies besides Upstart that have underwriting models using technology that is just as sophisticated or perhaps even better than Upstart's.
3. Can Upstart grow market share?
If Upstart wants to meet the growth expectations the market has put on it, the company is going to have to grow its market share and maintain that share. It said in a recent presentation that between the second quarter of 2020 and the first quarter of this year, total unsecured personal loans originated in the U.S. were $84 billion. Upstart in the last two quarters (its only ones as a public entity) has $4.5 billion in originations. That puts its share of that $84 billion figure at about 5.3%.
That's a respectable share, and I'm sure that overall number is going to grow as personal loans become more popular and if Upstart moves toward other global markets and lending products. But we also know that LendingClub is targeting $10 billion of personal loan originations for the full year, and another competitor, SoFi Technologies (SOFI -3.62%), has done $3.3 billion of originations over the last 12 months, so there's clearly some good fintech competition out there.
Another thing to consider is that most of Upstart's customers right now are credit unions and smaller banks. The company noted in its annual filing (things may have changed since then) that most borrowers find Upstart and then are referred to its bank partners. Small banks and credit unions are likely fine with being in the background because they are too small to have huge marketing budgets or to build a big brand.
But larger banks might care more about being visible in the transaction because banking is a relationship business. When a bank gives someone a personal loan or mortgage, it also hopes that person will do other banking transactions with them because the more products and services a bank sells to that customer, the more profitable the customer is to the bank.
That means larger banks could be less inclined to work with Upstart out of fear of losing the relationship with the customer. The company does offer banks the ability to use its technology on the back end and offer "Upstart-powered loans" through their own websites. But if a large bank were going to do that, it might also consider building its own technology instead. The debate that a lot of bank executive teams have internally is whether to buy or build new technology. And a lot of times (especially at the bigger banks), they choose to build, not that it's always the right decision.