Good afternoon, everyone. Thank you for joining the Mizuho Technology Conference 2026. I have the pleasure of hosting Andrea Blankmeyer, Upstart's Chief Financial Officer, for a fireside chat. There's people coming in. Before we start, I'm going to read a quick disclaimer, and then we'll start with the Q&A. Thank you, Andrea, for joining. Thank you for having us. Our pleasure. Today's discussion may contain forward-looking statement that relate to future results and events, which are based on Upstart's information available as of today and are subject to risks and uncertainties. Actual results may differ materially from these forward-looking statements. The discussion may also include non-GAAP financial measures, which are not a substitute for GAAP results. Please refer to the company's filing with the SEC and the IR website for additional information, including GAAP to non-GAAP reconciliations along with other disclosures. With that done, welcome everyone. This is actually Andrea's first fireside chat as CFO. Maybe you can give us some background about yourself and introduce yourself. Sure. Thank you, and thank you so much for hosting Upstart here today. Hello, everyone. I'm Andrea Blankmeyer, Upstart's relatively new CFO. I joined about three months ago, coming on three months. My background was in consulting and private equity investing, and then leading risk and finance teams at sort of complex risk-based businesses. I was the VP of finance at SoFi, CFO of a technology marketplace, of a transportation marketplace business, and then most recently CFO of Cityblock Health, which is a value-based care healthcare business, that has a balance sheet and a business model that's like an insurance company. Very excited to be here today, at my first fireside chat as CFO and here with you. Thank you. Well, you're in a great place. Maybe we can talk specifically about some of the results and can you help us. bridge the Q1 adjusted EBITDA and the full year guide, which would differ pretty significantly, and I think a lot of people have been raising questions about it. Thank you. The difference between 13 and 21, to be very specific. It is, yeah. As you said, our full year guidance was 21% EBITDA, which represents $294 million of adjusted EBITDA. We did 13%, about $40 million of EBITDA in Q1. We have ahead of us are the profits for the year, basically. As we've spoken to, we believe those profits will be sort of more back half weighted. The main drivers of the performance that we saw in Q1 were really threefold. One, there's a seasonality dynamic to our business. That seasonality impacts not just demand, so we see sort of top line softening, but that softening kind of flows through into take rate and contribution margins for our core personal loan business. One, there's seasonality for us on an OPEX side with sort of a structural step up in some of our costs in the first part of the year that we expect to moderate as we move forward. The second sort of dynamic that sort of came to bear in our Q1 results is a mix dynamic, and that's the story that's been playing out for us over the last year plus. We've seen really phenomenal growth in our newer secured products. We have a home product and an auto product, which have grown three and a half and four times year-over-year from Q1 2025- Q1 2026, which is great. We've also seen growth from a borrower segment perspective. Over the course of the last year have seen growth on the personal loan side in our super prime customer base, which is different than sort of the core Upstart borrower that we built our business on over the course of the last decade plus. The mix dynamic is important because both of those sort of vectors of growth and mix shift for the business, from a borrower segment perspective into super prime, as well as the growth in our newer secured products, both of those had a dampening effect on take rates and contribution margin and are a driver in some of the sort of the trending we've seen, particularly from Q2 2025- Q1 2026. That's the second dynamic. Seasonality, mix, the third is investment that the company made coming into the year, to kind of shore up our headcount and our people teams, our sort of people resources to make sure we were set up to achieve in the year. All of those factors came together and Q1 EBITDA margin was 13%. Now as we look forward, we do expect to see EBITDA margin improve significantly, as we sort of drive a few things forward. One, obviously leaving Q1 seasonality, we should see some lift on both the demand and the take rate and the contribution margin across the board. Second dynamic at play is we intend to and have driven sort of continued focus on our core personal loan segment. That's our sub 720 borrower segment, near prime and subprime borrower base, where Upstart's historical bread and butter, where we have significant and high contribution margins and real pricing power and higher take rates. That's the second thing, growing in our core personal loan business. Third is going to be to continue to grow our newer products while improving the unit economics of those. In Q1, those products had a negative contribution margin. We expect that contribution margin to improve as we move through the year. Those are the three main things as we sort of look through the remainder of the year that will deliver, in combination with growth and sort of a natural operating leverage as we expect fixed expenses to grow more moderately throughout the remainder of the year, that sort of deliver in combination, our revenue guide, our EBITDA guide, and our implied contribution margin and dollar guide. Great. Maybe let's shift gears a little bit to a very big topic, which is the bank charter. Maybe you can talk about, this is something that always comes up in general for Fintechs, your former company as well, which we cover. What are the benefits for Upstart here and just in general for Fintechs for a bank charter? Yes. We filed our application for an OCC and FDIC charter and are making our way through the process. Our expectation would be, pending regulatory approval, to launch in early 2027. The rationale for becoming a bank, importantly, is not one to fundamentally change our business model vis-a-vis how we treat our balance sheet and how we think about capital. Upstart is part of our sort of core element of our business model is our capital efficiency, and it's selling the vast majority of our loans, not having them touch our balance sheet or having to stay on our balance sheet for a relatively short period of time, selling those loans through to third parties. We intend to sort of maintain that element of our business model as a bank. What it is that the bank will deliver for us are a few things. One is going to sort of streamline our origination process. Today we work through these originating banks, that sort of collect a fee and a toll from us along the way as they're doing so. Two, it's meant to streamline kind of the regulatory complexity of our business, which is kind of a surprising one to people sometimes because everyone's like, "Well, being a bank is very regulatory complex." It is. The reality today is that we have to satisfy the credit committees of the hundreds of banking partners that we work with. Through them, we have to speak through many regulators, and we're also regulated in all of the states in which we operate. We're under a couple hundred different sort of regulatory frameworks, I would say. Streamlining that into a single one, we think is going to be efficient for us. Those are the primary sort of direct objectives, that both of which should have some positive impact to the P&L. The first one is kind of a reduction of a contra revenue for us. The second one over time is an improvement on sort of our cost base. Obviously, there will be investment for us to become a bank and operate well as a bank. A sort of secondary benefit that we see would be one sort of a modest cost of capital benefit for us. Again, from accessing deposit funding. It's real, it's just a smaller piece of our business because our intent is to keep that business, that balance sheet relatively small. The final element I would say is a little bit more qualitative, but we've been sort of the leading voice of AI in underwriting since our inception and have managed quite a few conversations with regulators, and we think that as the AI dialogue continues to evolve in credit, that we will benefit from having a direct dialogue with federal regulators on that, sort of expedite our ability to move our innovation through. Okay. Well, it's very exciting. And ma ybe we can talk a little bit about something that's been very topical, which is the state of the U.S. consumer, which is very relevant for Upstart. Maybe do you see any weakness? Obviously the FICO kind of nature of the borrowers on Upstart, how does the K-shaped economy play there? Then maybe we can touch a little bit on the UMI index, which has been. Ve ry helpful for us as analysts. Where are we now versus, say, a few years ago. On the arc of history when it comes to the state of the borrower? Yeah. Very topical question. I'd say sort of stepping it back and sort of broadly, what we sort of see is that the consumer is, we're in the middle of kind of a multi-year normalization cycle for the consumer. In particular, post-COVID and post-stimulus ending, we saw sort of a real uptick, in default, sort of a degradation in savings rates, and sort of an overall pressure on the consumer- that is significantly above pre-COVID levels. You can sort of see that in our UMI index itself. That's the Upstart Macro Index, which is a measure of the macro-related drivers of default and prepayment, and repayment patterns for a mixed adjusted borrower base. What we see is, over the last year or so, the UMI has kind of been in the 1.4- 1.5 range, which means the way you sort of interpret that is the same borrower today has a 1.4x- 1.5x higher default rate than that same borrower in the pre-COVID time for the same loan. That's been sort of peaked close to 1.7 a couple of years ago and has sort of gradually been coming down over time. We're sort of seeing ourselves broadly when you really sort of step back in a period of normalization for the consumer as they kind of right themselves post-COVID and stimulus. Away from that, I'd say over the course of last year, we've seen UMI fluctuates from month to month, but it's sort of been in that 1.4- 1.5 range. Our most recent print of UMI was April, and it was 1.46, which is kind of right in the middle of that range. Up a little bit from March, but still too early to tell kind of if that's a sustained trend or it's just part of the normal month-to-month fluctuations. Okay. There's been, I guess I would imagine, scenarios in the past where it goes up and then goes back down again. It doesn't- Very much so, yeah. It can very easily kind of move in that order of magnitude up or down on a month-to-month basis. It's really kind of been relatively stable with some moderate improvement, I'd say, over the course of the last year. UMI is a seasonally adjusted measure, just for what it's worth. Things like sort of tax season and otherwise are factored in to the repayment risk of the borrower. Got it. Let's talk about something exciting, which are the new products and the trajectory there. Home equity, auto loans, what's the rationale of getting into these products and maybe how's it going in those ventures? Yeah. These products are off to a great start, I would say. Auto was really sort of the second product for Upstart after our personal loan product, and then Home and HELOC have followed. Really it's part of the progression of Upstart to becoming a multi-credit product company that can serve every customer up and down the economic spectrum. It'd be the go-to place for credit for Americans. These represent bets placed two and three for the company to expand what we built as our core asset as a company around our AI underwriting models, one, and then reducing friction and automating the operating process, bringing both of those core assets of the business into different asset classes, into the secured side. That's really the rationale. They're both doing really well, as I said before, growing three to four times on a year-over-year basis. Auto is a little bit more mature, I would say, as a product. I think we have strong product market fit, strong credit performance, good capital markets, buyer support there. At this point, we're in the rinse and repeat side when it comes to growing, and then also looking to continue to improve the unit economics. HELOC is a little bit earlier in the journey, so it's a little bit of a newer product for us, still growing very well. I think we see lots of room ahead of us on both of those products. Let's talk about the most topical thing of all, which is private credit. Is there a weakness? Where do you see that throughout in the process of securitization? I think a lot of people have that on their mind. We should definitely address it. We can definitely talk about it, address it. It's definitely been a hot topic, I'd say, over the last few months. What I'll say is that we are not seeing the broader turbulence in private credit flow through to our buyers, and their appetite for Upstart loans, and to continue to grow with us. We have, as of mid-May, signed and announced over $4 billion of committed capital partnerships, including a couple of key renewals and extensions. A couple of those with private credit names that are in the press a lot or have been in the press a lot for challenges, as well as landing a couple of new deals. Our first deal that extends over a committed 24-month time horizon. What we're really seeing is a lot of strength from the capital markets perspective, a lot of demand to buy loans, private credit, and in other segments of our capital base, which ultimately, as always, reflects the underlying credit quality of the loans that we are originating. Okay. No reason to get concerned for now. No from your end. Not that we're seeing. Maybe talk a little bit and then we'll open it up for Q&A from the audience. AI, for us at least, no one is more synonymous with AI than Upstart. I feel like you guys were doing AI before people knew what AI was. How does the current state of AI change your business? Is there anything you can do better, get more efficient? What is the edge of Upstart today when everyone has access to AI? Yeah. It's a great question. To your point, Upstart has definitely been an AI-driven business long before that was a trendy thing to talk about. The core of that are really, AI is a broad term, but it's primarily machine learning-based models that we have built and developed over the past decade plus to allow us to separate and understand the risk of primarily subprime and near-prime borrowers significantly better than traditional credit models. That's the thing that has powered our growth. That's the thing that has allowed us to continue to sustain and improve our pricing power, even as we've grown so significantly, and to deliver on our core personal loan product, high +50% contribution margins. Right? That's the core asset of the business. In today's modern world with AI, with the LLMs, what I'd say is that the ways we are using those, the core machine learning models are continuing to do what they're doing, and they're the sorts of models that Gen AI can't replicate very easily because they're the machine learning, more math number-based things. The ways that we're using Gen AI today are powerful for the business, but I wouldn't say they're a meaningful acceleration of the core, nor does Gen AI allow anyone to come in and replicate what we've done on the core. The things that we are using Gen AI for, one, our engineers are using it to code. Right? They're expanding and speeding up their process to do everything, which is fantastic. The second place where we're finding Gen AI to be very helpful is around automating tasks that have historically required human intervention on. A big part of what we do alongside our core underwriting models is continue to remove friction in the underwriting process for our borrowers and automate more and more. There's just been a bucket of things that historically have continued to require some degree of human intervention on. It's like if there's a title document that has a handwritten something about your home, right? We've historically had to have someone look at that. Now that's the sort of thing that these Gen AI models are very good at doing and automating are these processes that have historically required a human touch on. We're seeing a real acceleration on that end, from an operating process perspective. Great. Thank you. If anyone has questions for Andrea? Yeah. We're going to get you a mic. Hey. Is it open? Yeah. Hey. About the HELOC business. It is currently very small, but we think it's very exciting. Can you tell us a little bit about, are there buyers for that yet in the funding markets? Secondly, unlike maybe the auto loans, are these the same borrowers? Usually HELOC borrowers are more prime. They already have a home. What can you tell us about this new business? Sure. Yeah. What we are finding on our HELOC product, as you would expect, is that it is a more prime customer. It's the 700 average FICO, versus 670 or so on our auto, 660, 670 on our auto product. I'm sorry, what was the first part of your question? That was one. The second was about the funding, the topic on funding. The funding side. Yes. We do have our sort of core anchor institutional buyer on the HELOC side, and are increasing sort of quarter-on-quarter the proportion of loans that we're selling through to third parties. We're not really seeing a shortage of capital and supply interest on that product. Because it's a little bit more of a prime customer, it's a very good customer match into our dozens, into the hundreds of credit union and banking partners who are bringing deposit cost of capital onto the Upstart platform. You're sort of seeing more of those buyers of loans for our HELOC product. Thank you. Yeah. Well, I want to thank you for your time today. Thank you, Andrea, and it was a pleasure having you. Thank you so much. Appreciate it. Thank you. it. Thank you. It was great to be here.