Upstart Holdings, Inc. (UPST) Q1 2021 Earnings Call Transcript

Upstart Holdings, Inc. (UPST -0.85%)
Q1 2021 Earnings Call
May 11, 2021, 4:30 p.m. ET

Contents:

• Prepared Remarks
• Questions and Answers
• Call Participants

Prepared Remarks:

Operator

Good day, and welcome to the Upstart Q1 FY 2021 earnings call. Today's conference is being recorded. At this time, I would like to turn the conference over to Mr. Jason Schmidt, head of investor relations.

Please go ahead, sir.

Jason Schmidt -- Head of Investor Relations

Good afternoon, and thank you for joining us on today's conference call to discuss Upstart's first-quarter 2021 financial results. With us on today's call are Dave Girouard, Upstart's chief executive officer; and Sanjay Datta, our chief financial officer. Before we begin, I wanted to remind you that shortly after the market close today, Upstart issued a press release announcing its first-quarter 2021 financial results and published an investor relations presentation. Both are available on our investor relations website, ir.upstart.com.

During the call, we will make forward-looking statements, such as guidance for the second quarter and full year 2021 related to our business. These statements are based on our current expectations and information available as of today and are subject to a variety of risks, uncertainties and assumptions. Actual results may differ materially as a result of various risk factors that have been described in our filings with the SEC. As a result, we caution you against placing undue reliance on these forward-looking statements.

We assume no obligation to update any forward-looking statements as a result of new information or future event, except as required by law. In addition, during today's call, unless otherwise stated, references to our results are provided as non-GAAP financial measures and are reconciled to our GAAP results, which can be found in the earnings release and supplemental tables. [Operator instructions] Now I would like to turn it over to Dave Girouard, CEO of Upstart.

Dave Girouard -- Chief Executive Officer

Good afternoon, everyone. Thank you for joining us on our quarterly earnings call covering our first-quarter 2021 results. I'm Dave Girouard, co-founder and CEO of Upstart. I want to start by thanking the entire Upstart team for another amazing quarter.

Despite the many obstacles the past year has put in front of you, you've continued to deliver. Most of you have never been in an Upstart office, yet the pace of innovation and the pursuit of our mission have never been stronger. Upstart is a leader in the application of artificial intelligence to consumer lending. And our first-quarter results continued to demonstrate our leadership in this emerging category.

Our Q1 2021 revenues were up 90%. Our profits were up by a factor of seven over the first quarter of 2020. Despite softer loan demand from consumers due to government stimulus programs, almost 170,000 loans were transacted by our bank partners in Q1, more than double the volume from just two quarters ago. Not unusual to see fast growth in a fintech company, but it's quite rare to see that growth paired with real profits.

So it's worth explaining what's driving our core personal loan business and why we're so optimistic about its future. Our growth is primarily technology and model-driven, which manifests as increasing conversion rates in our borrower funnel. In the first quarter of 2021, we launched a variety of model and technology upgrades that together unleashed the growth that you're seeing. To help you understand this dynamic, let me call out three examples.

First, we launched a new version of our core credit decisioning AI model that our bank partners use to price loans. This model is actually an ensemble or a blend of multiple machine learning models that work together to generate a more accurate risk assessment. Historically, we have achieved big wins by adding new and more sophisticated model forms to this ensemble. Other wins have come by simply upgrading one of the existing models.

But the big win we saw in the first quarter included an upgrade to the algorithm that governs how these models are blended together. This upgrade led to higher approval rates and lower interest rates, which, of course, translates into more loans. Second, we launched some improvements to our verification models that reduce the amount of friction applicants experience in applying for a loan, including those coming back to Upstart for a second loan. These improvements included more refined fraud signaling and increased automation of income verification.

This reduction in friction increased the conversion rate of our funnel, leading to more growth. It also helped us to more efficiently process the larger volume of applications that we experience, leading to higher profits. This is the essence of technology-driven growth. And third, we upgraded the models that underpin our marketing acquisition program.

That means for every dollar we spent on marketing, we brought incrementally more consumers to the platform and to our bank partners. AI has almost unlimited potential in spend targeting, life cycle marketing and even content generation. Applying AI to our customer acquisition efforts is a relatively new area of investment for us, and we expect it will be a significant source of growth in the future. Of course, this growth couldn't happen without the confidence of our bank partners and institutional investors.

We continue to add new bank partners to our platform, to see existing bank partners expand on their Upstart-powered lending programs and to add institutional investors who are happy to invest their capital into Upstart-powered bonds or high-quality loan assets with predictable yield. We've built an increasingly sophisticated supply chain of money that funds Upstart loans, from super-efficient bank depository capital, to broad and deep institutional and capital markets funding. While most consider money to be the ultimate commodity and lending to be a zero-sum game, we've built a proprietary platform that continues to accrue advantages by the day. It's the types of AI model upgrades described here, paired with the support of our bank partners, that have allowed Upstart to grow loan volumes by a factor of 20 in the last four years, while reducing acquisition costs and generating real profit.

Our optimism for the future comes from two facts: first, we continue to have a likely backlog of projects that will improve our funnel throughput and lead to more growth in the future; and second, we don't see competitors on the same AI-centric journey that Upstart is on. As we said in the past, auto lending is Upstart's next big opportunity. The auto lending industry is about six times the size of personal lending, and we believe it has at least as much mispricing and inefficiency, with millions of consumers paying far more than they should to finance a vehicle through a process that is displeasing to all. Our early explanation in this market has confirmed our hypothesis that there's a wide-open opportunity for Upstart and our bank partners to deliver a superior product to the market, meaning more accurate pricing, instant approvals and elimination of friction.

I'm happy to report that we're making very fast progress in this new and exciting category. Since January, we have expanded our auto refinance product, from a single state, to 33 states, representing more than two-thirds of the U.S. population. States have various rules and regulations that impact the process of refinancing a car loan, and we are building the best possible process for each.

At the same time, we're actively encouraging the space that have yet to adopt modern conveniences, such as digital signatures and electronic means and titling that make it easier for consumers to access the best possible loan. This is one area where the COVID pandemic has helpfully pushed states toward adopting these remote-friendly technologies sooner. We're also working to eliminate friction in the borrowing process, with the goal of delivering an auto refi product with the same quality experience and instant gratification that has made us a leading personal lending platform, only it's happening much faster this time. Given the complexities of auto lending, it's no small effort to build a refi product that is as simple and as accessible as an unsecured loan, but we're well on our way to doing that.

To date, we've activated only a couple of marketing channels to reach enough consumers to test and iterate on our auto refinance product. In fact, you can't yet find an auto loan offer on upstart.com. That will all change in the coming months as we begin to more proactively market and cross-sell our auto refinance product. I'm also happy to report that we closed our acquisition of Prodigy, a leader in automotive commerce software.

Prodigy is like Shopify for car dealerships, helping to create the modern multichannel car-buying experience that dealerships need and consumers rightfully expect in 2021. In addition to modernizing the car-buying experience, Prodigy will allow us to bring Upstart's AI-enabled auto loans to dealerships across the country, where the vast majority of loans are transacted. Despite potential distractions from the merger, the small but mighty Prodigy team increased our dealership footprint by 45% in the first quarter. Even at this early stage, almost $800 million in vehicles were sold through Prodigy in Q1 2021.

During the quarter, we became a certified digital retail provider for Subaru of America retailers. As I said last quarter, we believe Prodigy will enable Upstart to tap into one of the world's largest buy-now, pay-later market opportunities. To realize this potential, we're significantly increasing investment in Prodigy's technology and go-to-market teams. The first quarter wasn't just a win from a financial perspective.

We also made significant strides in our ongoing fair lending effort. From our perspective, if you want to be the industry leader in AI-enabled lending, you also need to be the leader in fairness, transparency and inclusiveness. And conversely, if you're not making clear and ongoing investments in fairness testing, it's questionable whether you're really doing AI. Last fall, the Consumer Financial Protection Bureau, the nation's preeminent regulator in matters of consumer protection, renewed its No-Action Letter with Upstart.

As a reminder, the purpose of so-called No-Action Letters is to reduce potential regulatory uncertainty for innovative products that may offer significant consumer benefit. While Upstart is now in our second three-year No-Action Letter with the CFPB, we know of no other lending platform that has received a No-Action Letter from the Bureau related to fair lending. As part of this renewal, we developed more sophisticated tests that raised the bar significantly in terms of how fairness is measured. Additionally, we increased our level of transparency by sharing the source code for the fairness tests we perform across every single Upstart loan applicant for all of our bank partners.

In the first quarter, we executed these new tests for the first time and recently delivered the result to CFPB. Our 2020 access to credit testing demonstrated once again that Upstart's platform improved access to credit in the form of increased approval rates and lower APRs for every traditionally underserved demographic tested. From my perspective, we've raised the bar on thoroughness and rigor of fairness testing. At the same time, when it comes to AI governance, I believe we're setting a standard for proactive cooperation between industry and government.

2020 was a transformative year for Upstart. The COVID-19 pandemic presented an opportunity for our technology and our team to shine during a very difficult period, and they did just that. We came into 2021 from a position of strength, with a business that's growing very quickly and generating significant profits at the same time. And we're now building on that strength.

We're upgrading our AI models and technology. We're adding bank partners to our platform. And we're expanding quickly into even larger market opportunity. It's clear to us that AI has unlocked a brighter future for lending, where consumers can access credit at a price that's fair and reasonable, with an experience worthy of 2021.

AI lending is a transformational opportunity to create a better product for consumers and for banks, and Upstart has become synonymous with this exciting new category. Thank you. I would like to now turn it over to Sanjay, our chief financial officer, to walk through our Q1 financial results and guidance. Sanjay?

Sanjay Datta -- Chief Financial Officer

Thank you, Dave, and thank you to everyone for joining us. Beyond some of the exciting details that Dave just walked us through, we aspire to be a company whose numbers speak for themselves. So with no further ado, we dive in. Our revenues in Q1 were $121.3 million, up 90% from the same quarter of the prior year and up 40% quarter-over-quarter.

Of that amount, revenue from fees represented $160 million or 96% of the total. Underpinning this fee-based revenue was the origination of 170,000 loans by all of the bank partners across our platform, up 102% from the same quarter of the prior year and a conversion rate of 22% on rate requests, up from 14% over the prior year. As Dave mentioned, our top-line performance this quarter benefited from continuing improvements to model accuracy, some sizable advances achieved in targeting models we use to optimize our acquisition campaigns and work we completed to further reduce friction in our borrowing process, including beyond the initial transaction with the customer. Our contribution profit, a non-GAAP metric, which we define as revenue from fees minus variable costs for borrower acquisition, verification and servicing, was $55.8 million in Q1, up 117% year-over-year and representing a 48% contribution margin, up from a margin of 38% in the year prior.

The margin improvement over the past year has been partially driven by improvements to marketing and operating efficiency, some of which were implemented in the wake of the pandemic in order to manage the business more conservatively through the turbulence. As we discussed on our last earnings call, this level of contribution margin continues to be slightly above our expected long-term trend line, and we expect it to normalize downwards moderately over the coming quarters as our marketing channels and operations hiring return to full scale. Q1 operating expenses were $105.8 million, up 67% year over year or 58% year over year when netting out the impact of stock-based compensation. Investment in engineering and R&D remains our priority, growing 171% year-over-year to $19 million in Q1.

General and administrative spend grew slower than revenue in Q1, increasing 72% year over year to $20 million, boosted in part by additional costs incurred from the closing of our Prodigy acquisition. We continue to ramp our fixed expense base following lower-than-typical investments in 2020. The other expense categories of sales and marketing and customer operations were largely driven by the variable cost increases in borrower acquisition, verification and servicing previously discussed. Our Q1 GAAP net income was $10.1 million, up from $1.5 million in the prior year despite incurring a onetime $5.3 million expense from the voluntary repayment of our PPP loan.

Adjusted EBITDA, which we adjust for stock-based compensation, came in at $21 million in Q1, up from $3.7 million in Q1 of 2020, despite being similarly impacted by the repayment of the PPP loan. Adjusted earnings per share for Q1 was $0.22 based on a diluted weighted average share count of 91.5 million. Turning our attention to the balance sheet. We ended the quarter with $336 million in restricted and unrestricted cash, up from $72 million at the end of the first quarter 2020.

This balance reflects the repayment of our $5.3 million PPP loan, but does not reflect the proceeds from the follow-on stock offering we completed on April 13, which resulted in an additional $265 million raised, net of underwriting discounts. In terms of loan assets, we carried an aggregate balance of loans, notes and residuals of $73.2 million, down from $227.5 million at the end of the same quarter in the prior year. This reflects the continuing reduction in the percentage of platform loans funded through our own balance sheet. As highlighted during our last earnings call, these loan assets represent the totality of the direct exposure we have to credit risk.

Now turning to our near-term and full year outlook. As expressed in our full year guidance communicated last quarter, we are now seeing a catch-up in business results to where we believe our technology improved to over the course of 2020. As well, we continue to see our contribution margins as slightly above their historical trend, and we expect them to mildly contract as we return to a more normal operating spend. There are two additional trends we would like to call your attention to that will be impacting the margins in our guidance.

The first is that we expect to direct incremental earnings upside toward accelerating the expansion of the Prodigy Software platform we recently acquired, consistent with our level of excitement at the magnitude of that opportunity. This will have an impact on our EBITDA margins through the duration of 2021. The second is that given the relatively sharp rise in our stock price since the beginning of the year, we anticipate that our stock-based compensation expenses for our recent grants will be rising disproportionately as we expense them over the next few quarters, which will impact our net income margins. With these points in mind, for Q2, we are expecting total revenues of $150 million to $160 million, representing a quarter-over-quarter growth rate of 28% at the midpoint, we note that year-over-year growth rates for Q2 will become somewhat meaningless as we lap the peak impact of the pandemic; contribution margin of approximately 44%; net income of $8 million to $12 million; adjusted net income of $21 million to $25 million; adjusted EBITDA of $23 million to $27 million; and a diluted weighted average share count of approximately 94.9 million shares.

As a result of the continuing strengthening in our business as well as in our underlying model technologies, for the full year of 2021, we now expect revenues of approximately $600 million, representing a growth rate of 157% year-over-year and up from the $500 million that we indicated last quarter; contribution margin of approximately 42%, up from 41% indicated last quarter; and we continue to expect an adjusted EBITDA margin of approximately 10%. That is all of the information we had to share with you today. I want to again thank you all for taking out time to join us and hear about our results. And to also reiterate Dave's gratitude to all of our hardworking teams at Upstart, whose talents and efforts are contributing to our progress, you all are making it very easy to be the financial mouthpiece of this company.

With that, Dave and I are now happy to open the call to any questions. Operator?

Questions & Answers:

Operator

Thank you. [Operator instructions] We'll take our first question from Ramsey El-Assal from Barclays. Please go ahead.

Ramsey El-Assal -- Barclays -- Analyst

Hi gentlemen. Thanks for taking my question this evening. I wanted to ask about your loan conversion rate, it came in well above our model. And I know you gave us some granularity on some of the improvements on the technical side that you've made.

How do you expect that metric to trend through the year? It feels like that's a pretty decent contributor to the full year guidance raise and maybe a higher conversion rate than our models reflected. And also, just speak to what the upward limit might look like on that rate over time or if there is one.

Sanjay Datta -- Chief Financial Officer

Sure. Yes. Hey, Ramsey, thanks for your question. This is Sanjay.

I guess what I would say is, so it's obviously the super important driver of our top line, the performance of the funnel. It is, in a sense, a beneficiary from the underlying accuracy of the models and the underlying strength of the technology. And that technology strength can manifest in one of two ways. It can show up as improvements to our conversion rates.

Or sometimes, it can show up as improvements to the top of our funnel because as a conversion funnel improves, we then expand our marketing campaigns and scale. And maybe one way to think about this is we sort of signaled that our contribution margins themselves are a little bit abnormally high right now. And that's related to what you're talking about because typically, as the conversion funnel improves, and if we can predict that accurately and anticipate that what we will do is expand our marketing campaigns, and the fact that our contribution margins are somewhat high right now, is maybe a signal that we haven't expanded our marketing campaigns appropriately. So we're still, in a sense, leaving some profitable loans on the table.

So over time, what you might expect is that the technology will hopefully continue to improve, our conversion funnel will improve, but then expansion on the sort of acquisition side will bring it back down a little bit and will convert it more into top-of-funnel traffic. So that's the dynamic. It's a bit hard to predict how improvements to our technology will show up between the conversion rate improvements and just overall sort of marketing scale. And as for upper limits, for that very reason, it's a bit hard to predict the upper limit because a lot of what's driving the improvement of our technology, as you get to very high conversion rates, will just naturally result in larger acquisition campaign size.

But as Dave said, and I think the core point that's important is, the backlog of projects we have, that sort of have a very direct line to improving the accuracy of the model, is pretty long. And that's really the underlying driver of what's driving a lot of the business results that you're seeing.

Ramsey El-Assal -- Barclays -- Analyst

OK. A really quick follow-up for me is just on the Prodigy network and your kind of projected penetration rate of the auto dealers in that network. If you could just remind us, is that something where you have to go dealer-by-dealer and kind of convert them over to your loans? Or is that something where you can kind of make a back-end change and sort of migrate everybody over to you guys rather than other competitors offering loans?

Sanjay Datta -- Chief Financial Officer

So it's a multistep process. I mean the first thing that needs to happen and the place that we're focused right now is in getting the Prodigy Software adopted by the dealers themselves. And as that sort of footprint expands, then within that transaction volume, we have the opportunity to offer our loans. Our loans still need to compete with the rest of the marketplace, obviously.

At the end of the day, you need to have the best prices and the highest approval rates in order to win. And that's an area, I think, we're relatively confident in. So I think, to us, strategically, what's important is the expansion of the footprint of the commercial software that Prodigy is developing. From that will flow opportunity for loan volume, but that would be almost a sort of a second step, if you will.

Ramsey El-Assal -- Barclays -- Analyst

Got it. Thanks for taking my questions.

Operator

Thank you. We'll move on to our next question from Pete Christiansen from Citi.

Pete Christiansen -- Citi -- Analyst

Good afternoon. Thanks for the opportunity to ask questions. I just want to back off of Ramsey's first question a little bit. I'm trying to understand the upgrades that you made to the platform this quarter and how that translated to a higher conversion, also notice that average loan size was down year-over-year.

Are you just identifying more near-prime borrowers that are, according to your models, more creditworthy? Or is it a function of just getting more creditworthy borrowers into the funnel?

Dave Girouard -- Chief Executive Officer

Thanks, Pete. This is Dave. I'll answer that question. Yes, the model accuracy does both of those things.

It gets smarter about separating good credits from bad credits. It thereby helps us avoid offering loans to those who are more likely to default. And the effect of that, which is the most basic sort of building block of our business, is that it allows us to approve more people at generally lower rates. So that fundamental dynamic is about as core to what we do and how we do it as anything.

And we don't term anybody near-prime because a lot of times, that's a legacy label applied by a FICO score that's not very predictive any longer. So in our mind, we are looking for people with a very high likelihood to repay a loan by a very specific model. And that dynamic is the core of what has driven the growth, along the reduction in friction for people who have seen a loan that they would like to get access to and have to just get through the approval process. And so those are sort of the twin pillars of how our growth happens generally.

At least at the first level is a more accurate credit model that tends to approve more people at lower rates. And secondly, a reduction in friction for those who want to get the loan. So those are the sort of the primary effect of growth, and that's exactly what happened in the first quarter.

Pete Christiansen -- Citi -- Analyst

Thank you. That's helpful. I guess we all need to change our mindsets when thinking about your model. And then just as a follow-up question, as it relates to getting Prodigy in the hands of more dealers, is there a commission or a rebate issue that you would have to compete against other lenders into that category?

Dave Girouard -- Chief Executive Officer

No. The Prodigy Software, as it exists today, is sold on a sort of subscription basis for a few thousand dollars to a dealership to sort of modernize their auto selling experience. We haven't changed that model to date. In the future, the Upstart loans will appear in there, and they will become an option to finance a vehicle for consumers.

It's not likely the only option, but an option would be Upstart-powered loans. And the money can flow in different directions on that, depending on the nature of the borrower and such. But in any case, we are extremely optimistic that it will be a source of what amounts to very low, if not even negative acquisition cost for loans, which is just an incredible opportunity. It's the nature of what point-of-sale finance generally does overall.

And in this particular very large market opportunity, the auto lending market, it's an incredible opportunity for us.

Pete Christiansen -- Citi -- Analyst

Thank you very much.

Operator

Thank you. We'll hear next from Ron Josey with JMP Securities. Please go ahead.

Ron Josey -- JMP Securities -- Analyst

Great. Thanks for taking the question. Dave, I wanted to ask two things. One is just on your bank partner loans and as it relates to the Upstart Referral Network.

We're seeing more and more partners join, and I think it's a shorter day to onboard, sort of days to onboard. So if you can just talk about the process here as more bank partners join and just how that process is going. And then secondly, when you talked about, maybe to tie onto the question that was just asked, you talked about reducing friction in terms of newer products in first quarter. And I think around new and I heard repeat users for this improving automation, reducing friction, can you talk a little bit more about just the repeat process and how that's coming along? Thanks again.

Dave Girouard -- Chief Executive Officer

Sure. And those are great questions. So Upstart Referral Network is the term we use when a borrower comes to Upstart and then is referred to one of our bank partners, and Upstart Referral Network is the term we use to market that service to the banks themselves, just for clarity. And basically, yes, I mean, if banks are heavily regulated, they certainly have a lot of diligence to do if they're going to use technologies such as ours to originate credit.

So that process, I think, will always be a significant lift. But having said that, we're doing a lot of work to streamline that process, particularly for smaller community banks as well as credit unions. There's just a very quickly growing interest from credit unions in what we're doing. So we're really trying to do what I think almost any technology company tries to do to get enterprise adoption, which is to shrink the selling process, shrink the diligence process, shrink the onboarding process.

And there's no sort of magic bullet there other than we continue to refine each part of that. And we have now, multiple times, been able to sign up a bank or a credit union in less than 90 days from our first conversation. And that, for us, is a massive step forward. It's not the standard yet.

We would like it to be in the future. And we're also just kind of shrinking the diligence steps in the onboarding steps, just the refinement of the technology and having more sort of off-the-shelf diligence materials ready. So it's just an ongoing effort, and I don't think we'll ever be done with it, frankly. But we are making nice progress.

There are, I think, 18 banks and credit unions on our platform today. And...[Technical difficulty]

Operator

I do apologize for the interruption. It looks like we had a speaker disconnect. Give us just a moment to go ahead and get Dave reconnected. One moment.

Dave Girouard -- Chief Executive Officer

OK. Sorry. No, I'm not sure where I dropped, other than to say, we're shrinking the process. The second part of the question was on sort of repeat borrowers, if you will.

Yes, that's something that we've been streamlining for some time. We are now, just in the first quarter, we began to actually market to those who had gotten loans in the past and might be eligible to get one again. So we're still a somewhat young platform compared to some others. And so this is still a modest part, and we have focused to date on making a profitable, growing a profitable business, really where we were just getting that first initial loan with the consumer.

And so this is really upside to our business and it seems to be coming together very quickly. This is really just personal loan and then a second personal loan, but of course, a bigger future is when we can cross-sell between products, a personal loan to an auto loan, or vice versa, and all that is in front of us.

Operator

We'll take our next question from Arvind Ramnani from Piper Sandler. Please go ahead.

Arvind Ramnani -- Piper Sandler -- Analyst

Congrats on a terrific quarter. I just had a couple of questions. The first question is really around the market size for personal loans. So as more loans are processed through Upstart's platform, reducing acquisition costs, pricing risks more accurately and you start to gain scale, over time, does it pressure more banks to work with Upstart because you'll have a different approach, an AI-powered approach to solving this problem? Just really trying to understand what's the market share ceiling in the personal lending space.

Dave Girouard -- Chief Executive Officer

Hi, Arvind. This is Dave. I'll answer that question. So personal loans is one of the fastest-growing categories of credit and because consumers love it.

It's very simple and has high utility and fast to access, generally low rates, etc. So it's a very fast-growing category. I think it's taking market share away from credit cards. I think it's taken market share away from HELOCs, which have a much more difficult process and take more time.

So we believe it will continue to be a very fast-growing category. We also think we're really the first with a proprietary product because it's been sort of a fragmented market for a long time, but we really think we have a proprietary advantage that can allow us to grow market share very, very quickly and probably beyond what's been seen in the past. So we're very optimistic that it's a category in its early stages in terms of how big it can be. And we'll continue to improve and grow very quickly, particularly as better products come to market from people like Upstart.

With respect to banks, and will banks feel the pressure to join Upstart or work with Upstart or somebody like us, I think that will certainly be an effect we would expect over time, particularly as AI moves into lending categories that are more and more meaningful and central to their business, and there's just more and more proof points of the better economic outcomes of using AI in lending. So personal loans, for example, is not a very central business for most banks. So they don't necessarily feel a lot of pressure by a personal lending product that's markedly better than another one. Auto lending is certainly much more central to more and more banks.

So we are getting a lot of interest because there's a lot of large banks that have significant auto businesses. And now, naturally, this is an area of interest. The idea of being able to approve more people, have lower loss rates, a more profitable portfolio of loans is obviously compelling. And as we move to other categories over time, whatever those might be, I think we're going to just see more and more interest.

Our view generally is all flavors of lending are going to be AI-enabled in the future because the economics are so much better. And it's only a matter of time until all banks will get onboard, if not with Upstart, then with building something themselves or trying to source this type of technology somewhere.

Arvind Ramnani -- Piper Sandler -- Analyst

Terrific. And then I have a couple of questions on Prodigy. Like just from a kind of the way you're powering Prodigy, are you able to apply the same AI algorithms or data for Prodigy as you're doing in the personal lending space? And the second thing is like, certainly, there's been like a massive ramp. I think you mentioned $800 million in Q1.

It's like a significant growth rate. How should we kind of think about volume, the volume you can deliver in 2021 with Prodigy?

Dave Girouard -- Chief Executive Officer

Sure. So just to quickly explain how Prodigy fits in. Prodigy is really about the point-of-sale auto lending, meaning at the dealership. There's a first step, even prior to that, that's already been in process for some time with us, and that is refinance of auto loans, which is done really through upstart.com for someone who got an auto loan elsewhere.

And generally speaking, we are taking every bit of our model that is applied to personal loans and applying it to auto loans. And what that really means is, when you underwrite an auto loan and you have an individual to consider, and that's, of course, what we've been doing in our whole personal loan business and you suddenly have an asset that is backing the loan that adds some more complexity and opportunity to the model. So it's really an extension of what we built and refined for years in the personal lending space. And it's now being applied, as I mentioned, in 33 states, still at the early stages, but in the auto refinance product.

That exact model will then become part of what we offer through Prodigy to dealerships. So you can think of it as it's being initially refined and built as a refinance product, but more or less the same model will apply for a new vehicle purchase that will happen at the dealership. And that's obviously central to our success is we don't start from scratch on a new model. We actually learn from what we have and then extend it in some way or another to make it relevant to a new market.

It's exactly what we're doing with the auto market and with Prodigy.

Arvind Ramnani -- Piper Sandler -- Analyst

Thank you very much.

Sanjay Datta -- Chief Financial Officer

So Arvind, hey, this is Sanjay. I was just going to tack on and maybe address the second part of your question, which is the fact that the Prodigy footprint is growing very quickly, which it is. But it's also a massive market, as Dave said, meaning the sort of the dealership financing market overall, and they're at the very early stages. And we think the opportunity is large, and we do believe it will grow fast.

But I guess, suffice to say, My belief is their footprint and how quickly it grows won't be the limiting factor in our ability to scale the auto lending side of Upstart. I think they're going to provide plenty of transaction volume, and it will be up to the rest of the business to build the operations and funding infrastructure to scale. But I think as and when we can scale the sort of the lending part of our auto business, the sort of commercial software footprint that Prodigy's providing us, I think, will be more than ample sort of transaction volume and size for us to grow well into the future.

Arvind Ramnani -- Piper Sandler -- Analyst

Yes. I'm certainly excited about the auto offering and looking forward to more on that as the year progresses. Thanks so much.

Operator

Thank you. [Operator instructions] We will hear next from John Hecht from Jefferies.

John Hecht -- Jefferies-- Analyst

Good afternoon guys. Thanks for taking my questions. Actually, a lot of my questions have been asked. And you did talk about getting the, call it, the sales cycle more productive with the banks and credit unions.

I'm wondering, OK, can you share with us? You mentioned 18 banks on the platform. Is there a characteristic that you're identifying of asset size or just balance sheet disposition of the banks that are, call it, migrating to you more rapidly? And same question with the credit unions.

Dave Girouard -- Chief Executive Officer

Well, sure, that's a good question. I would say, first of all, it's obvious that we tend to be in smaller community or regional banks and credit unions. There are some, I mean, in the $40 billion-ish assets or more. So they're not the smallest banks.

But our interest really is appealing to all of them, having a simple process that can work for a very small bank or a credit union that doesn't have a lot of resources. But of course, there's no reason we can't partner with and wouldn't hope to partner with the largest banks in the country over time. I think today, because it's still an early stage market, it's very helpful to have a bank where the CEO would have awareness, would be involved in this, the credit committee would be aware, the lending, the EVP of consumer lending, etc., would be involved. And so it tends to make it a little easier to be these smaller banks where you can get everybody on the same page.

There will always be a process and the process is going to be more complicated the larger the bank is. So I think that's what we're seeing in the earlier stages as maybe not unlike technology adoption in any kind of new industry as it tends to happen a little bit smaller and then move up to the large and larger enterprises over time. That's exactly how we see this playing out. Though again, at the same time, we had started off a couple of years ago saying, look, we wanted banks between $5 billion and maybe $100 billion in assets as our target.

And I think we've realized we can probably go below that and above that as the product matures. And so I think we're definitely broadening over time.

John Hecht -- Jefferies-- Analyst

OK. That's helpful. And then is there a resale cycle if the bank partners are buying unsecured loans, can they just, in a sense, hop in the queue for the auto loans? Or do you have to reengage them and go through a separate sales cycle?

Dave Girouard -- Chief Executive Officer

Well, that's a great question. We are definitely working toward sort of standard agreements that you can add another product on with a sort of simple one-page exercise. So it's a very insightful question. Like you don't want to go through a full contracting process to add a new product to what the bank might be doing with us.

So we've really moved toward kind of a master agreement style, where you can say, you started with personal loans, now you want to add auto loans. You might have started getting loans from upstart.com, now you actually want to use the technology on your own website, etc. So all these just trying to streamline that process of a bank starting with us in one place and then working with us multiple places over time, which is what we're seeing for sure.

John Hecht -- Jefferies-- Analyst

OK. And then final question and I apologize if this was addressed, I had some phone issues as well. As the last round of stimulus gets absorbed, have you seen any changes in kind of payment patterns, like the payment rate may be slowing? Or are you seeing any indications that stimulus is getting absorbed into the system at this point?

Sanjay Datta -- Chief Financial Officer

Hey, John. This is Sanjay. Yes, we haven't seen too much change in the macro outlook as of right now. Obviously, on the one hand, the economy is beginning to open up a little bit.

On the other hand, there was a fresh round of stimulus that has reduced people's demand for loans. And so I think, from the indicators, things have been pretty steady since last October, last November. And I think that's consistent with the macro metric we track most closely to, which is credit card balances since a lot of our loans are aimed at refi in those. And those balances have fallen in on the order of, I feel like, 10% since the beginning of the crisis, and they've only regained about 1% or 2% as of March.

So I think our macro sort of indicators would sort of reflect that.

John Hecht -- Jefferies-- Analyst

Great. Thanks very much guys.

Operator

Thanks. We'll take a follow-up question from Ron Josey from JMP Securities.

Ron Josey -- JMP Securities -- Analyst

Thanks for the questions. Just we're getting a lot of questions here on just the guidance. And so Sanjay, can you unpack a little bit more of the increase in the annual guidance and wondering if auto is just maybe a larger part there given, I think, Dave, you mentioned making fast progress overall, not just at Prodigy, but now in 33 states in general. So just any more insights, Sanjay, on the increase in annual guidance from the $500 million to $600 million would be great.

Thank you.

Sanjay Datta -- Chief Financial Officer

Sure, Ron. I mean maybe just to put a point on it, it still includes really nothing meaningful from the auto side. We continue to be, I guess, you might call it an incubation mode there, developing the operations, expanding the Prodigy dealer footprint. So we're still not counting on any meaningful contribution.

A lot of it boils down to the kinds of things that Dave talked about every single week. We are watching improvements to the credit models, improvements to the verification models, improvements to the acquisition-targeting models. And as those improve, as those drive our daily and weekly numbers, we propagate them through with some statistical confidence interval, if you will. And so we've had a very good quarter, obviously.

We're obviously leaving the quarter on a good front. Our guidance for the next quarter is a little bit under 30% quarter over quarter, and we're propagating that strength through to the end of the year.

Ron Josey -- JMP Securities -- Analyst

Got it. Thank you.

Jason Schmidt -- Head of Investor Relations

I think that's our last question. So I'll turn it back over to Dave.

Dave Girouard -- Chief Executive Officer

All righty. Thanks, Jason. Again, I just want to repeat a thank you to the Upstart team for an amazing quarter. We're really thrilled with our results.

And thanks to all of you for spending the time with us today. We will see you all next time.

Operator

[Operator signoff]

Duration: 54 minutes

Call participants:

Jason Schmidt -- Head of Investor Relations

Dave Girouard -- Chief Executive Officer

Sanjay Datta -- Chief Financial Officer

Ramsey El-Assal -- Barclays -- Analyst

Pete Christiansen -- Citi -- Analyst

Ron Josey -- JMP Securities -- Analyst

Arvind Ramnani -- Piper Sandler -- Analyst

John Hecht -- Jefferies-- Analyst

More UPST analysis

All earnings call transcripts