Pagaya Stock Uses AI to Power Asset-Backed Securities
About three months ago, we revisited a fintech company called Upstart (UPST) after its stock price went soaring. In our previous article on the company, which uses artificial intelligence to analyze loan applications, we decided there was way too much risk and volatility to jump onto the IPO, despite the impressive growth story. What goes up, must come down in the wacky world of tech-stock hype: Upstart has dropped nearly -40% since our last article though it’s still up +445% in its first year of trading. That happened despite a monster quarter that beat analysts’ expectations (for what that’s worth). Yet those same analysts punished the company for less-than-spectacular guidance for the fourth quarter. This is what happens when institutional investors pay attention to arbitrary price targets pulled out of the asses of people who are trying to make MD before they turn 40.
So it goes, as Mr. Kurt Vonnegut would say. About the same time that our article came out in September, another company that uses machine learning and big data to disrupt the credit and
debt loan industries announced it would go public by – wait for it – merging with a special purpose acquisition company (SPAC). Does Pagaya stock represent a restart on an Upstart-like opportunity for retail investors?
About Pagaya Stock
Founded in 2016, the Israeli company is co-headquartered in New Yawk City (where a cream cheese shortage threatens to bring the economy to its knees). Pagaya has raised more than $221 million over the last five years, including from a number of investors that would naturally be interested in making bigger profits by loaning more people money in deals that even your corner store loan shark might balk at doing. Some of the more familiar names include Citibank and supplemental insurance company Aflac, along with other banks in Thailand and Israel, as well as financial services companies. The SPAC de jure is EJF Acquisition Corp (EJFA), whose pre-merger stock price got a mere 2% bump when the deal was announced and is still trading below the $10 default share price for SPACs.
The deal will dump nearly $500 million into Pagaya’s bank account and earn it a valuation of just over $9 billion, assuming no one pulls their money out sooner. That’s quite a jump from June 2020 when the company last raised money from a $102 million Series D that valued it at $2 billion. The valuation may not entirely be without merit, as the company just announced third-quarter revenue of $137 million, which would give it annualized revenues of $548 million. Using our simple valuation ratio (market cap/annualized revenues), Pagaya comes in at about 16 – well below our threshold of 40. The word “profitability” was also tossed around during the investor presentation but if profits exist, they haven’t been spelled out – a typical SPAC obfuscation.
How Pagaya Makes Money
Speaking of slightly confusing and opaque: Let’s talk about how Pagaya makes money. As we alluded to earlier, the company uses AI to analyze consumer data using information not typically applied to traditional credit scores. We’ve referred to this before as “the new credit score.” The idea is that by using tons of alternative data that old school credit agencies don’t consider, algorithms can unleash more capital by better quantifying risk versus reward. Lots of companies are using alternative data not just for determining credit risk but for other types of financial services and investing, not to mention keeping China’s citizens in line.
In the case of Pagaya, its platform uses more than 16 million training data points and has already evaluated more than 17 million applications in markets ranging from personal loans and automotive to credit card and now real estate. Based on the second-quarter results, network volume – the amount of loans being processed by the platform – reached $4.7 billion annualized. Pagaya partners with financial institutions like SoFi – another SPAC deal alumnus – to source their deal flow.
Pagaya earns fees when its “network volume is acquired by institutional investors.” In effect, the company is selling asset-backed securities (ABS), which pools together these different types of consumer debt and sells them to investors. It’s sort of like an index fund based on the desires, dreams, and dumb ideas of the average person. In theory, the algorithms can predict the likelihood of each loan being repaid, presumably with an interest rate and terms that reflect the risk. No doubt the platform also calculates a percentage that will default, which is baked into the bundled loans that investors buy. All of this is 100% automated – versus about 70% with Upstart, which does not bundle loans in this way – and Pagaya never itself issues a loan which means they don’t have credit exposure to the loans they’re generating for institutions. It also earns fees on assets that are originated and serviced by its partners.
In March, the company announced what it called the biggest consumer loan ABS – the $900 million Pagaya AI Debt (PAID) Selection Trust. Over the last two years (as of March 2021), Pagaya has completed 10 of these PAID trusts – all fully using its proprietary AI.
Should You Buy Pagaya Stock?
Let that sink in: Algorithms are entirely in charge of billions of dollars in debt. You might remember a little economic meltdown circa 2008 that involved hedge funds and banks creating a bunch of crappy mortgage-backed securities that left Wall Street hung out to dry. This time we’re asking AI to do it better. From an emerging tech point of view, it sounds really cool. From a risk-averse point of view, it kind of scares us a little bit. In its own long list of potential risks, Pagaya notes: “Our A.I. technology may not operate properly or as we expect it to, which could cause us to inaccurately verify transactions.” Oopsie.
Aside from the existential dread that this company evokes, it’s also just not a business model we would want any part of, even though it is capital light. Pagaya only makes money as long as consumers are buying stuff and firms are offering credit to people. If that tightens up – or even if institutional demand for consumer credit-related loans dries up – Pagaya doesn’t have much of a business. Plus, what incentive does Pagaya have to make sure these are quality loans? It’s already pushed at least 10 of these ABS bundles out the door in just a two-year period. There’s no history to how well these perform over time. This fine print in the SPAC deck pretty much sums it up.
Our A.I. technology has not yet been extensively tested during down-cycle economic conditions. If our A.I. technology does not accurately reflect a borrower’s credit risk in such economic conditions, the performance of loans may be worse than anticipated.Credit: Pagaya SPAC Deck
Another red flag: Shortly after the proposed transaction was announced, two U.S. law firms reportedly got nosey and wrote to EJF Acquisition Corp. shareholders that the deal would allegedly dilute their shares and leave them with less than 6% of the company. That could potentially derail the SPAC merger, though we never invest in a SPAC until the deal is done, and then rarely even after that. The accusation pretty well reflects remarks this week by Gary Gensler, the SEC Chairman, when he said, “[R]etail investors may not be getting adequate information about how their shares can be diluted throughout the various stages of a SPAC.” The SEC may finally be making a move to bring a little more transparency to these SPAC transactions.
There’s a good chance that IF Pagaya stock goes live on the Nasdaq next year that it will initially take off based on the current growth trajectory. We won’t be going along for the ride, as the business model just seems too risky to us. It becomes a double jeopardy situation when you factor in the high-risk assets that institutional investors are acquiring based on the advice of an untested AI system. Did Wargames teach us nothing?
Tech investing is extremely risky. Minimize your risk with our stock research, investment tools, and portfolios, and find out which tech stocks you should avoid. Become a Nanalyze Premium member and find out today!