Artificial intelligence (AI)-powered lending platform Upstart (UPST 2.76%) touts a potential $4 trillion market opportunity. Considering it "only" saw $1.1 billion in loans transacted on its platform last quarter, this company should see exponential growth in its future. But Upstart isn't the first company that's attempted to disrupt and take share from a massive market before, and its business has taken a big step backward this year.
However, patient investors should also see reasons to be optimistic, particularly if one thing can change for Upstart from here.
A look at Upstart's business
Before we get into the thing that Upstart needs to unlock its potential, let's talk about what Upstart does. At its core, Upstart provides what it says is a better way to evaluate credit risk than the traditional tools, simultaneously giving lenders a way to make safer, more profitable loans, and to improve access to credit for millions of people. This is the ultimate potential "win-win" sort of business: Disrupting an industry that hasn't changed how it does things for many years, while providing something for all stakeholders that improves outcomes.
How does Upstart do this? When the company was founded 11 years ago, CEO David Girouard believed that "...technology and data science could improve access to affordable credit." The current system -- primarily the FICO score -- has been the standard since the late 1980s, determining who gets credit, and how much they pay. And while it has been tweaked over the years and isn't the only variable a lender uses in making a credit decision, it's at the heart of a system that Upstart believes "...fail(s) to accurately identify and quantify risk."
Upstart's answer? Discrete AI models that "...incorporate more than 1,500 variables and benefit from a rapidly growing training dataset that currently contains more than 44.4 million repayment events."
Is that all just buzzwords, or does it work? According to the company, it's better at more accurately identifying credit risk. The slide below breaks down the default rates of loans based on Upstart's "risk grades" versus the borrower's FICO score:
Image source: Upstart.
If this sort of result holds true, it means that lenders should have a better way to make loan determinations. It also means a lot of people who may not have a high FICO score could have better access to lower-cost debt. And that should mean big profits in Upstart's future.
If it's better, why is Upstart's revenue falling?
This is maybe the most important question to understand. As a starting point, Upstart isn't really a lender. It does underwrite and hold some loans, but the vast majority are transacted by or sold to its lending partners. Upstart makes money on fees, while the lender partners take on the loan risk and earn interest income.
The slide below shows where Upstart's claim of a $4 trillion opportunity comes from:
Image source: Upstart presentation.
The company's data sources include credit reporting giant TransUnion and the U.S. Small Business Administration, so we know it's a "real" market. However, at this stage almost all of its business has come from personal loans. This is both the smallest addressable market, and one that's viewed by lenders as being one of the most risky.
That second part is a big reason why Upstart's business results have cratered this year. Unlike a home or auto loan or many business loans where there is collateral, personal loans are typically unsecured. As a result, default rates are generally much higher, and defaults tend to rise sooner and faster during an economic downturn. They also have higher interest rates, because they are higher-risk. And interest rates have skyrocketed over the past year and a half.
When you combine all of those things, most lenders have pulled back from personal lending in general, and in particular on Upstart's platform. So even as the company has seen partner lenders increase 41% over the past year, revenue has fallen off sharply.
The one thing Upstart needs to return to exponential growth: Lender confidence
Why exactly are lenders partnering with Upstart, yet simultaneously reticent to use it? In a word, they lack confidence that Upstart's loans will continue to perform better during an economic downturn. Upstart hasn't operated through a "normal" recession yet; there was so much financial stimulus during the pandemic recession, default rates didn't spike in general.
Does that mean we need to go through a recession before lenders flock to Upstart's platform en masse? Maybe; that would certainly answer the biggest concern. If Upstart's loans continue to outperform their FICO counterparts during and after a recession -- Upstart says they already do better in a healthy economy -- it would likely result in huge growth in loan volume and partners.
But what if we don't actually get a recession in the near term? Chances are, lenders will jump right back on the dance floor, the pullback in lending would resolve itself, and Upstart's solid loan performance in good times would lead plenty to take on the risk and start writing more loans again.
Whatever happens next with the economy, Upstart isn't standing still. It now has over 60 auto dealerships offering Upstart car loans, a move into a market that is 5 times larger than personal loans, and it recently started a pilot program for home equity lines of credit, or HELOCs, a small move into the massive home loan market. These moves should also result in growing confidence for lenders, if borrower credit quality holds up according to Upstart's models.
So what's an investor to do?
At this point, investing in the company is a bit of a leap of faith. Either you believe that its process is better and will hold up, or you don't (or don't yet). In other words, it's also about confidence. If the company has indeed cracked the credit-risk code, it's likely to take a much bigger piece of the lending pie.
With a market cap of $2.4 billion at recent prices, the potential returns could be enormous from here. But if Upstart's quality doesn't hold up in a downturn, lenders won't have much reason to stick around. Neither would investors.
