Upstart Holdings (UPST 2.97%) has caught Wall Street's attention in a big way. The stock price has skyrocketed over 940% since the company went public in December 2020. And the driving force behind those gains is Upstart's prospects for future growth.
In this Backstage Pass video, which aired Sept. 27, 2021, Motley Fool contributor John Bromels discusses how this fintech company brings artificial intelligence to the consumer credit industry and how disruptive its approach could be to traditional lending solutions.
John Bromels: Co-Founder and CEO, Dave Girouard is the former President of Enterprise Google and Co-Founder and a Councilman, the former Manager of Global Enterprise Customer Programs and Consumer Operations at Google. Why is that important? What does Upstart do? Upstart is an AI-driven lending platform. It uses artificial intelligence, big data, and predictive modeling to create an alternative to the FICO score to determine a person's creditworthiness.
You've all probably heard a FICO score. A bunch of credit cards will give you a FICO score. You can go to annual credit report.com every year, get your score and your credit report from the various credit bureaus. It's like a one-shot snapshot determining how likely you are to default on alone or on credit that is extended to you. The founders of Upstart said, what the FICO score does, what it looks at, it looks at your past, it looks at your credit history, it looks at do you have do you have assets, how much of your assets do you have available to you? It said, well, that's backward-looking. For people who don't have a significant credit history, or for people who maybe whose circumstances have changed and they are, looking to get a loan, the FICO score may not be a good predictor of whether they are likely to default or not.
It developed this algorithm that is AI-driven to approve these loans. What it does is you go to Upstart, you go to their apps, you fill up basically the loan application. It uses its predictive modeling and its AI to determine your creditworthiness. Then it does not do the lending itself. It goes through third-party banks and financial institutions to actually extend you the credit. It's recently entered into auto loans or with a purchase of prodigy software, and there are rumors floating about that it might be looking at the mortgage loan market as well. This is based on a job posting. The Internet will find all things for these companies. Its based on like a job hosting net referenced mortgages in the job hosting and they said, "Aha. If they are looking for somebody who has experience with mortgages, maybe they're looking at getting into the mortgage loan market."
CEO Dave Girouard has also expressed an interest in disrupting the payday lending industry, which as we know, is essentially designed. The payday lending industry argues that they need to charge these exorbitant and these very high fees because of the high risk of default. Well, Dave Girouard and Upstart says well, if our model can more accurately predict who's going to default and who isn't, perhaps we can be a better alternative for many consumers than a payday loan. That brings up the question of bias. The trick with artificial intelligence is quite often it's looking for the best efficiency, or the most mathematically perfect number or outcome.
Sometimes, however, that can introduce biases into the system, that can introduce things that we don't want to be there. Of course, the very famous example of this is when some researchers created an artificial intelligence, and basically used social media to allow it to examine how people talked. The algorithm picked up all objectionable content and started skewing it out because of the flawed input and had to be shut down and reconfigured to work that out. Upstart is really committed though to reducing bias in its algorithm because of course, you don't want unconscious bias to creep its way into your lending. Because not only is that illegal, it is also really unfair.
Upstart has actually been working with the Consumer Financial Protection Bureau, the government agency that is responsible for policing this, and they have a no-action letter, which is essentially the CFPB's way of saying, we've been monitoring this AI, and we're satisfied that it doesn't actually introduce any unlawful bias into lending process. Again, a very niche product, a very nice industry but a really good example of how AI, and using big data, and using machine learning and predictive modeling can really disrupt many different industries.
