Upstart Holdings, Inc.

Good day, and welcome to the Upstart Q3 FY 2021 earnings call. Today's conference is being recorded. At this time, I would like to turn the conference over to Jason Smith, Vice President of Investor Relations. Please go ahead, sir.

Good afternoon, and thank you for joining us on today's conference call to discuss Upstart's third quarter 2021 financial results. With us on today's call are Dave Gerrard, Upstart's Chief Executive Officer, and Sanjay Datta, our Chief Financial Officer. Before we begin, I want to remind you that shortly after the market closed today, Upstart issued a press release announcing its third 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 fourth quarter related to our business and our plans to expand our platform in the future. 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 events, 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. To ensure that we have as many analyst questions as possible during the call, we request that you limit yourself to one initial question and one follow-up. Later this quarter, Upstart will be participating in Citi's 2021 FinTech Conference on November 16th and 17th. Now I'd like to turn it over to Dave Girard, CEO of Upstart.

Good afternoon, everyone. Thank you for joining us on our quarterly earnings call covering our third quarter 2021 results. I'm Dave Girard, co-founder and CEO of Upstart. Today marks our fourth earnings report as a publicly traded company. and I continue to be amazed and delighted with the progress the Upstart team has made. This is an entirely unique time in our history, and executing as a newly public company in this environment is not without challenges. Four out of five Upstarters joined our company during the pandemic, and many have neither seen the inside of an Upstart office nor met their new colleagues in person. Yet collectively, they continue to knock down walls and make amazing things happen. Artificial intelligence is perhaps the most transformational technology the world has yet to see. And Upstart is at the forefront of applying AI to the multi-trillion dollar financial services industry. So the scale of the opportunity is not lost on us. Q3 was another strong quarter of triple-digit growth and profits. And we'll get to that shortly. But it's worth pausing for a moment to reflect on what our team has accomplished in the years since we went public. In the six years leading up to our IPO, 620,000 loans were originated on the Upstart platform. A year later, our bank and credit union partners have originated more than 1.5 million Upstart-powered loans, totaling more than $16 billion in originations. While there were 80,893 loans originated on our platform in Q3 2020, which was the quarter prior to our IPO, we facilitated 362,780 Upstart-powered loans in Q3 2021. That's a growth rate of 348%. As you could read in Upstart's S1 filing, our AI platform had experienced 9 million repayment events and was trained on 15 billion cells of training data as of a year ago. Today, our platform has processed 17 million repayment events and is trained on 28 billion cells of training data. A year ago, we had 10 bank and credit union partners on our platform. Today, we have 31 partners, and we're adding them faster than ever. We're also making progress in how rapidly we can onboard these partners. In fact, our most recent bank partner went live on the platform in less than 50 days. A year ago, a handful of auto loans had been refinanced in a single state. Today, more than 4,000 upstart powered auto loans have been originated in 47 different states. We also became a digital-first company this year. In Q3 of 2019, 100% of our new hires were located in either California or Ohio. In Q3 2021, we hired new upstarters in 25 different states, as well as the District of Columbia. Now we'd like to shift and talk about the products on our platform today, personal lending and auto lending, and the progress we're making with each. Then I'd like to touch on a few new product areas that are important to our future and that we're beginning to invest in today. Personal loans continue to drive the growth and profitable economics of Upstart. Post-pandemic, the demand for these simple installment products has reaccelerated. I like to refer to personal loans as the duct tape of credit, a fast and simple solution that consumers love for their usefulness, affordability, and accessibility. And banks are beginning to realize that offering instant, all-digital personal loans makes sense. unless they want their customers to find them elsewhere. Last quarter, I told you that, for the first time, an Upstart bank partner decided to eliminate any and all FICO requirements for their borrowers. Today, I can happily report that of those who were approved for Upstart-powered loans because of this change, 59% were Black, Hispanic, or of low to moderate income. And even better, I'm also pleased to tell you that four upstart bank partners have now dropped their FICO requirement. In just a few months, what was perhaps a canary in the coal mine has now become a trend. We applaud any and all lenders that eliminate credit score requirements. These trailblazers are building businesses that are more inclusive, more equitable, and yet more profitable. Their commitments represent a small step to the future but provide a major boost to the financial inclusion that our country desperately needs. While this is just the beginning of a trend, we anticipate a day when lenders will struggle to explain why they rely on a three-digit number invented 30 years ago to make a credit decision. While a lot of our energy goes towards serving quote-unquote future prime consumers currently left out of the financial system, we're also launching initiatives to better serve those with no shortage of credit options. This makes sense because borrowers who use Upstart often see rapid improvements in their credit score and become suddenly very interesting to the entire consumer finance industry. Our bank partners want to serve these customers over their lifetime rather than seeing them defect to competitors who constantly pursue them. And if it's important for our bank partners, then it's important for Upstart. It turns out it's important for Upstart for other reasons as well. Only by serving consumers across the entire credit spectrum can we reach our full potential. Even if your credit score begins with a 7 or an 8, you are more than your credit score. By working with our bank partners, we're beginning to deliver instant loan offers with no origination fees and with rates that are as attractive as any on the market. Full-spectrum competitiveness unlocks broader branding opportunities for Upstart, such as television. These are marketing channels that might not have made economic sense to us previously. Our efforts in auto lending continue to make progress as well. As a reminder, the auto lending market is at least six times the size of personal loan market. In our view, it's at least as inefficient. Millions of Americans pay too much for their car loans every month, and I'm sure each would prefer to spend that money on something else, or even better, to save it for a rainy day. If you don't believe me, just Google Consumer Reports auto loans, and you'll see an interesting article published just a few weeks ago. On the auto refinance front, we continue to make fast progress to eliminate the time and effort required to refinance a car loan. While the complexities of liens and titles, as well as a bewildering array of state-by-state processes, fees, and regulations conspired to keep Americans trapped in their mispriced car loans, we're on track to repeat the funnel gains we experienced over the years in our personal loan product. In their auto insurance commercials, GEICO delivers a message that resonates with me. 15 minutes can save you 15%. While we can't use GEICO's tagline, my goal is that Upstart and our bank partners will emulate that value proposition to consumers who are paying too much on their car loans. We're also making rapid progress on our auto purchase product. In the third quarter, we rebranded the company and the product formerly known as Prodigy to Upstart Auto Retail. But our progress in auto retail went far beyond rebranding. In fact, we've now tripled the number of dealers on our platform compared to a year ago. In a Q3, we added an average of more than one rooftop a day. I'm also excited to tell you about a major new milestone for our company. Just last week, the first upstart powered loan was originated through our auto retail software with one of our longstanding dealership partners right here in the Bay Area. Our early experience is showing us how powerful it can be to offer instant decisioning, high approval rates, a broad selection of terms and payments, all integrated into a digital process. This is a big deal, not just for Upstart, but for the entire auto industry. It represents the first step toward leveraging AI to rewire and revamp the entire car buying process. And finally, we hear a lot of questions about where Upstart plans to go next. We've been clear with our lending partners and with the investor community that we intend to expand our AI platform beyond the personal and auto lending categories. And today we want to share more about those plans. First, we're working toward a small dollar loan product designed to help consumers with unexpected and immediate cash needs. Think a few hundred dollars repaid in just a few months. But importantly, we're building a bank-ready product at bank-friendly APRs. always operating within the 36% rate cap prescribed to nationally chartered banks and to those who serve U.S. military service members. In short, with better technology, superior risk models, and a dramatic reduction in the cost of origination, we hope to welcome millions of Americans into the mainstream financial system who would otherwise be left with far less attractive options. Upstart stands to benefit from this small-dollar product as well, it can significantly accelerate the pace with which we can bring the underbanked into the financial system, and it can likewise accelerate the pace of learning by Upstart's AI models. We aim to bring millions of marginalized consumers onto our platform in the coming years so that our lending partners can serve them with a host of affordable financial products over time. Our bank partners rightly feel pressured to better serve low to moderate income Americans, and we want to help them do that right. The interest in this small-dollar product from our bank and credit union partners is off the charts, and we hope to bring it to market before the end of 2022. Second, we believe there's an unmet need to provide fast, easy access to affordable installment loans to business owners across the country. Every small business is different, and they operate across a crazy-wide spectrum of industries. So there are significant challenges to delivering a compelling loan product that is useful to business owners, yet is reliable, performant, and efficient enough for lenders. This challenge is tailor-made for Upstart. While there's no shortage of credit options to business owners, we aim to deliver the zero-latency, affordable credit solution that modern businesses require. This is another product in high demand from our bank and credit union partners, and we hope to bring it to market during 2022 as well. And finally, I would be remiss if I didn't mention what the late great sportscaster Keith Jackson would call the granddaddy of them all, the home mortgage market. It's by far the largest consumer lending category. And to upstart, it represents a proportionally large opportunity to improve the financial lives of Americans. We're all familiar with the financial crisis of 2008 to 2010, triggered by irresponsible mortgage lending, among other things. Well, one sad outcome of that crisis was a seemingly permanent reduction in access to affordable mortgages to the average American. In fact, a study by the Urban Institute found that if you compare the number of mortgages originated to Americans with FICO scores less than 700 in 2001, six or seven years before the crisis, with the number originated to that same group in 2015, you'll find that there were more than a million fewer mortgages in 2015. This is what we call the missing million. And from where we stand, it's crystal clear that a huge fraction of these million would-be homeowners are more than creditworthy and deserve access to an affordable mortgage. This is an opportunity that we're excited about and will begin to invest in significantly throughout 2022. While this initiative has a longer time horizon on it, we felt it's important to share our intention right now. Before I turn it over to Sanjay, I want to say thank you to the entire Upstart team for once again making me proud to be part of this company. Our team understands the impact that Upstart has already had in the world, and the magnitude of the opportunity in front of us is that much larger. But without the amazing talent and dedication of each team member at Upstart, that incredible opportunity would go unrealized. Thank you. And now I'd like to turn it over to Sanjay, our Chief Financial Officer, to walk through our Q3 financial results and guidance. Sanjay?

Thank you, Dave, and thanks to everyone for joining today. I hope all are well. I will take us for a quick spin through the numbers. Starting at the top of the P&L, revenues came in this quarter at $228 million, up 250% from the third quarter of last year, and up 18% sequentially from last quarter. Of that total revenue, $210 million, or 92%, came in the form of revenue from fees. The volume of transactions across our platform this quarter was approximately 363,000 loans, up 348% year-over-year, and representing approximately 315,000 new borrowers. As Dave alluded to, we are methodically expanding our footprint to serve more of the credit spectrum. both towards more traditionally prime borrowers, as the proliferation of bank deposit funding on our platform enables us to be more rate competitive in segments where we have historically had limited presence, as well as deeper into the millions of Americans whose poor credit scores do not accurately reflect their true creditworthiness. This increased breadth of borrower profiles in both directions has driven up the volume of applications at the top of our acquisition funnel by a factor of three over the past year. but has also exerted downward pressure on the average platform conversion rate, which at 23.0% is up 780 basis points versus last year, but down sequentially from 24.4% last quarter. Borrower segments that are relatively newer to our models will initially tend to convert at a lower rate than those segments for which we have longer history. The percentage of loans fully automated on our platform is also down slightly to 67% in Q3 from 71% last quarter for similar reasons. Newer borrower profiles will tend to have more conservative rates of instantaneous approval until we develop a longer history and greater loan volume for our models to train on. As a compounding factor, this past quarter we experienced a large coordinated effort to obtain loans fraudulently from our platform. We bore no meaningful financial impact from this activity, but our increasingly public profile as a company leads us to expect that episodes of this type will become increasingly common. This new reality has motivated us to implement additional protections to our origination processes, and these defensive measures have contributed modestly to lower automation rates. Since fraudulent applications also inflate the denominator of our conversion ratios, we have begun removing those rate inquiries identified by our platform as likely fraudulent from our conversion rate calculation. Such fraudulent applications had an immaterial impact to the conversion rate in prior quarters. Our contribution profit, a non-GAAP metric which we define as revenue from fees minus variable costs for borrower acquisition, verification, and servicing was $95.9 million in Q3, up 184% year-over-year and representing a 46% contribution margin, down from a margin of 54% in the year prior and slightly above our guided level of 45%. As discussed on past earnings calls, this margin was expected to moderate concurrent with the scaling up of our acquisition programs, as well as from the margin impact of our growing auto loan volumes. Q3 operating expenses were $200 million, up 276% year over year, or 239% when netting out the impact of stock-based compensation. Investment in engineering and R&D remains our priority, growing 272% to $37 million in Q3. We plan to continue investing heavily in our technical workforce as we ramp investments in machine learning, auto retail, fraud, and as we scramble new teams to begin tackling opportunities in segments such as the small dollar lending, small business lending, and mortgage initiatives discussed by Dave. General and administrative spend grew slower than revenue in Q3, increasing 241% year-over-year to $34 million. The other expense categories of sales and marketing and customer operations were largely driven by variable cost increases supporting revenue growth. Our Q3 gap net income was $29.1 million, up 201% from the same quarter of the prior year. Adjusted EBITDA came in at $59.1 million in Q3, up 283% from Q3 of 2020. Adjusted earnings per share for Q3 was 60 cents, based on a diluted weighted average share count of 96.1 million. Turning our attention to the balance sheet, we ended the quarter with $1.2 billion in restricted and unrestricted cash, up from $618 million at the end of last quarter. This balance reflects the proceeds from the convertible senior note offering completed on August 20, which resulted in an additional $587 million of proceeds, net of debt issuance costs, and purchase of capped calls. In terms of loan assets, we carried an aggregate balance of loans, notes, and residuals of $140 million, up from $95.3 million in Q3 and down from $145 million at the end of the same quarter in the prior year. While the fraction of overall platform loan volume funded through our own balance sheet remains low at sub-3%, the absolute dollar volume of loans we carry is edging upwards as we use our balance sheet to support the scaling of our growing auto product as well as our expansion into the lower credit score segments of personal lending. In terms of macro outlook, we are seeing the early signs of a return to the pre-COVID consumer profile, with personal savings rates in the economy now having fallen back to pre-COVID levels and credit card balances steadily edging upwards to within 90% of pre-COVID levels. We expect the continuation of this trend to eventually lead to an increase in consumer default rates, consistent with pre-COVID levels, and we believe that any issuer who has not priced this in is likely to experience a deterioration in the performance of their returns. We also expect these macro dynamics to ultimately lead to an increase in borrower loan demand, although this has yet to manifest in our results and remains upside to our forecast, as the exact timing is unclear. With this backdrop in mind, for Q4 of 2021, we are expecting revenues of $255 to $265 million, representing a year-over-year growth rate of 200% at the midpoint, and bringing our full-year 2021 revenue guidance up to $803 million versus the previously guided number of $750 million. Contribution margin of approximately 47%, net income of $16 million to $20 million, adjusted net income of $48 to $50 million, adjusted EBITDA of $51 to $53 million, and a diluted weighted average share count of approximately 96.7 million shares. I would like to reiterate Dave's gratitude to all of the talented teams at Upstart, whose hard work makes all of these results a reality. and to also wish everyone a happy upcoming Veterans Day. On behalf of Upstart, we want to express our gratitude to the women and men who serve our country and to those who have sacrificed to keep us all safe. With that, Dave and I are now happy to open the call to any questions. Operator?

Thank you. If you would like to ask a question, please signal by pressing star 1 on your telephone keypad If you're using a speakerphone, please make sure your mute function is turned off to allow your signal to reach our equipment. Again, that is star one if you would like to ask a question. We'll pause for just a moment to allow everyone an opportunity to signal for questions.

And once again, that is star one if you would like to ask a question.

And we'll take our first question from Arvind Romani from a private investor.

Hello.

Arvind, your line is now live.

Okay, great. Thank you. Sorry, I was on mute. Yeah, I wanted to ask you a question about the quarter. You know, certainly when I go back to last sprint, you know, you all raised revenue guidance by $150 million. And this sprint, the raise is smaller, roughly about $60 million. And, you know, certainly now you have two quarters to play it as opposed to last quarter. We had three quarters to kind of play with, but still the magnitude of the beat has diminished. I mean, are you kind of entering a phase where, you know, you're sort of what you're guiding to and what you'll deliver? You know, do you expect that sort of accuracy to increase and thereby that magnitude of beat to be lowered as a result.

This is Dave. Yeah, I would certainly expect that. I mean, our goal is to provide accurate guidance. And so for sure, you know, the more we can understand about the outquarters in our business, the more I would expect us to be reasonably conservative, but trying to provide, you know, real transparency into what we know and expect about our business. So I would just attribute that to us having a good sense. I mean, our business is growing at a phenomenal rate by any measure, being triple-digit. And then some revenue growth while being profitable is quite unique in the fintech world. So we're quite pleased with that. And to the extent we can reasonably predict that in the future, we would view that as a positive.

Great, great. And then, you know, on the call, you talked about – certainly the auto business and these essentially kind of flag, you planted the flag on the adjacent businesses within lending, home buying, and a couple of those. How should we really think of unit economics of these? I'm just trying to really kind of size up the financial impact because I would assume like personal loan is very different than auto, which is very different than home. Are you able to kind of provide some kind of financial metrics of how we should start to think about this as we think over the next like two, three years of your financial model?

Yeah, maybe I'll just try to give a summary view on that question and then Sanjay can chime in with some more specifics. But generally, you know, we move towards markets where we see significant inefficiencies. And to us, that means we could build a better product for consumers and for lending partners and also make a profitable product for upstart so you don't you don't likely see upstart chasing low margin zero margin money losing parts of the of the credit industry because they're not of interest to us so the unit economics of each product may change may be very different in certain ways but generally speaking we're obviously looking for strong contribution margins and things that can you know help us to grow so If you want to know what sort of company Upstart is, it's one that focuses on obvious inefficiencies where our AI models can deliver kind of excess economics that can make a win for everybody. But the specifics of each product are almost by definition different, the nature of each product. But I think we're very confident we will continue to drive top-line and bottom-line growth. Maybe let Sanjay add to that.

Yeah, sure. Hey, Arvind. I guess I would say auto is sort of the product that's in flight right now, and you can sort of think of it into two different categories. One is auto refi, and the unit economics on that product are very analogous to what we have in our personal loan category, meaning the banks that originate the loans will pay us a fee and will incur an acquisition cost and an operations cost, and that'll lead to a contribution margin. And look, the contribution margin... as with personal loans, will be a function of how strong our conversion funnel is. So it'll be something that improves over time. And it'll probably be on some journey similar to what we've seen in personal loans, but with obviously a few years behind. And then what do you think about the other category of auto that we're getting into now with auto retail lending? It's purchase auto at the point of sale in the dealership. And we expect those economics to look better because There's less acquisition costs involved because you're at the point of sale. And frankly, some of the operations required to originate the loan are handled by the dealer. So we view that to be a higher margin product. And then once you get beyond auto, frankly, it's just way too early to tell. I think we're in the very preliminary stages of thinking about business lending and home lending. So I don't think we have anything much to really report on that front. But I think as we get into those... Those areas, as Dave said, we're really looking for the opportunity to create value and earn profits. So that will be the expectation.

Terrific. And just a quick follow-up on that. As we think of – I'm not looking for guidance on 22 and 23, but as we start to think of those years, how should we think of revenue contribution from auto? Is it too early to ask that question? But certainly starting to think about 22 and 23 numbers.

Yeah, probably a bit early to ask that question with any specificity, Arvind. I mean, I think we believe that they will be meaningful. We believe auto will be a meaningful contributor to our financials next year, but we don't have specifics on what exactly that means.

Perfect. Thank you very much. Thank you.

And our next question comes from Randy L. Assal with Barclays.

Hi, gentlemen. Thanks for taking my question this evening. I wanted to ask about guidance, and it seemed like the fourth quarter guidance came in, you know, well more than the Q3 beat. It almost feels as if there was a little, maybe a pocket of a little sort of lower performance in Q3, but it seems like you're building momentum into Q4. Can you kind of, you know, elaborate on what's giving you confidence to raise the guidance so much more than the beat in Q4?

Hey, Ramzi, this is Sanjay. Yeah, go ahead, Dave.

No, go ahead, Sanjay.

Oh, sorry. Yeah, look, I would say sort of something we've said more generally over time, that what's really driving our performance is improvement over time in the underlying models, and that doesn't happen in a smooth way. It sort of happens in fits and starts, and so if you look at where we've sort of focused our time in Q3, we've had some good wins, but frankly, there was a lot of and focus we put into areas like fraud mitigation, just because that became an interesting topic for us in Q3. But, you know, we see the progress we've made so far in the quarter and what we have remaining in the roadmap for this quarter, and I think we're very optimistic about it. So I think you see that play out in our guidance, and you'll see that play out in our revenue patterns over time as well.

Okay. And could you give us more color or any detail on the fraud issue you faced? And is it contained? Is it ongoing? Should we expect any impact next quarter? Anything, any details you could provide there would be helpful.

Sure, Ramsey. This is Dave. So, I mean, just to be clear, we sort of deal with fraud constantly and pretty much since the beginning. In the third quarter, we saw a pretty significant and at scale organized effort, which in the end didn't have any, you know, Michael Siafina, M.D.: : material financial impact on us and we were able to successfully defend against it, but it didn't require us to put some resources against it, and you know we sort of came out of it with more defenses than we went into it, but. Michael Siafina, M.D.: : So we just see this as you know, but everything we can learn just a result of being a higher profile company and something we can pretty much expect going forward but. In the end, it did not have a financial impact on us. It's sort of, I guess what we would just say, an ongoing cost of doing business. We're pretty happy that we have not suffered significant losses from fraud in our history and feel very good about where our fraud rates are and have been. No, I don't think it's really anything worth thinking about or building into a model or anything else. It's really just part of naturally being a consumer-facing financial technology company.

Got it.

All right. Thank you very much.

And our next question comes from Mike Ning with Goldman Sachs.

Hey, good afternoon. Thanks for the question. I just have two. First, I was just wondering if you could talk a little bit more about the small dollar loan product. You know, how does that fit into the broader strategy? Does it help you accelerate market share gains within the personal loan market or do you see some other opportunity? And then second, I was just wondering if you could talk a little bit about the loan sizes, the average loan sizes in the quarter and how we should think about that going forward into 4Q and then into 2022. Thank you.

Sure, Mike. This is Dave. I'll cover the small dollar question and maybe Sanjay can speak to loan sizes. You know, we just see it as an opportunity for a win across the board. And first of all, for consumers, it's just an area that is not particularly well served with an affordable product that's not exploitive of the consumer. So, you know, loans for a few hundred dollars for a few months are just it's just an area ripe with exploitation. And we can do it right. We can do it better in a way that's not exploitive and provides financing to a consumer for short term when they need it. And we can do it. under the envelope, you know, that banks operate, uh, meaning under the 36% rate limit and, and, uh, with much more affordable products that don't create debt cycles that are harmful to consumers. So first and foremost, it's a huge step forward for the consumer, which is very well aligned with upstart's mission and, and, and frankly, very central to it. Uh, secondly, banks themselves, uh, have a lot of pressure to better serve low and moderate income Americans. I mean, there's literally been, you know, messages sent from regulators to banks that they really need to better serve consumers with small dollar loans for when immediate financing is needed. And most banks have just struggled to be able to do that in a reasonable and affordable and responsible manner. So it just isn't out there. But we've heard a ton from our bank partners and credit union partners about wanting to have a simple low dollar product that can help them do the right thing by their customers. So it's a win there. And finally, for Upstart, it's a huge win for us as well. It's much less about direct profits on this product than it is about the ability to have our models learn much more quickly and to bring people who are at the margins and our models might not be able to approve today for a larger loan, but could approve them for a much smaller loan. And that gets them to a path of being in the mainstream financial system more quickly. And it helps our model learn more quickly and So suddenly down the road, one of our bank partners could offer them a car loan or a mortgage or some other product. And that to us is an enormous win if we can bring millions of people who are currently outside the mainstream bank-oriented financial system and bring it inside of it in a helpful way. And that's exactly what we want to do with a small dollar product. I'll let Sanjay take the question about loan sizes.

Sure. Hey, Michael. Yeah, with respect to loan size, there's really maybe two main things going on. One of them is optical and one of them is real. The optical one really has to do with the mix of our portfolio. And in particular, as Dave described, more and more banks are eliminating their cycle guardrails. And what that is allowing us to do is go down spectrum in terms of what we can originate and underwrite. And as we do that, obviously, we're working with a borrower that is receiving smaller dollar sizes just because they're at the riskier end of the spectrum. So as we do that, sort of the mix of the portfolio is shifting. And so that's having an impact on average loan size. So that's something we would call optical. The real effect is that loan sizes in real terms have been coming down pretty consistently since pretty much the end of last spring. And that has been true pretty much through the end of this past quarter. And we attribute that, or I guess we call that sort of suppressed loan demand, and we largely attribute it to the stimulus in the economy. And so we don't have a crystal ball as to how this is going to propagate going forward, but I do think that as I called out in my remarks, I do think we are starting to see the signs in terms of the macro economy of a normalization in the economy with respect to savings rates, with respect to credit balances. And so if that is really accurate, what we would expect to see is a stabilization of loan size and then ultimately a gradual return to larger loan sizes as savings rates go down and credit balances come back up. So that is, I wouldn't call that a prediction. That's maybe more of a thesis. But that's sort of where we stand right now with respect to loan size.

Great. Thanks, Dave. Thanks, Sandra. Thanks, Mike.

And our next question comes from John Hecht with Jefferies.

Hey, guys. Thanks very much for taking my questions. The first one is just about the funnel. Maybe can you talk about kind of the channels and sources of new customers? I believe you had some good momentum in recurring customers coming directly to your site. So any comment or characteristics with the development of the Mixa on the front end?

You know, John, we had very fast growth and I would generally say it was across all our channels. So it wasn't channel specific, which we do is a good thing for certain, you know, having some of our banks drop FICO scores suddenly means their partner sending us traffic that they wouldn't have sent us before. That doesn't always convert as much. So there's effects like that. But generally speaking, you know, the partner traffic has been great. We've seen great progress in digital. We are without question growing quickly in repeat loans or kind of lifecycle marketing, which is what we call it internally. So there's a lot of those dynamics. Direct mail continues to be a very productive channel for us. So when our funnel gets better and our aperture of who we can lend to or who our bank partners are lending to gets better, then it just typically goes across channel. And that is definitely what we've seen is cross-channel improvements.

Okay, that's helpful. Thanks. And then maybe talk about the investor side. I know there's been a dramatic, at a global scale, an increase in investor demand. Is that still the case in terms of kind of capital allocation? And then maybe any commentary on the pipeline of banks and credit unions that we should expect over the near term?

Sure, maybe I can start with that. Maybe I'll just give the high-level...

Sorry, Sanjay, maybe I'll just do the high level first. As I think you know, John, we have a lot of banks who are originating loans for their own balance sheets. We have some banks that are selling to 100 plus, 150 plus capital markets partners, and that creates this very liquid environment that we work in. I'll just speak maybe to the bank side. I mean, we definitely have very rapidly growing interest from bank partners for some of the reasons we've said in the past, which is banks tend to be very heavy in deposits and light in loans. And that's an environment that works really well for us. I would say we just continue to see increased interest. Our ability to get banks onto our platform quickly has definitely improved. I mentioned a In my remarks, we got a bank partner on in less than 50 days, which is a new record for us. So the demand is there. I think the confidence in our models from the bank partners is there. The returns to the banks have actually been excessive, meaning higher than they would have expected on loans, principally because the macro environment has really suppressed default rates for us, probably like a lot of others. The banking side is very, very strong, and we feel very confident we're going to continue to add banks and credit union partners at a rapid clip. And maybe Sanjay can speak to the investor side.

Sure. Yeah, thanks, Dave. Yeah, what I'd say for those loans that flow through the banks and into the capital markets, I would maybe point out a short-term dynamic and maybe a longer-term or more secular one. The short-term dynamic, as you know, is like loss rates are naturally low right now in our opinion. And the ABS markets, I believe I would call them highly constructive. And of course, there's a lot of capital right now chasing yield. And the combination of those three things has made it such that there is a lot of demand for loans out there. We think that will change and normalize. I think the things that we think about in the more medium to long term, which are important to us, are two. One, I do believe over this period of time, the sort of whole category of digital personal loans has really, I think, gained a lot of credibility as a legitimate source of yield. And you're starting to see institutions that are much more conservative and traditional come into this space. And so I think there's been an increasing acceptance into the mainstream of investment finance on the credit side. And then the second more important one is, you know, how do we look within that? And I think that, you know, Upstart continues to have a lot of credibility for the performance of the credit and our ability to underwrite the risk. in a way that I think differentiates itself from other products in the market. So we think the combination of those couple of things will be such that even when we're past the current, let's call it sort of distortions in the market, we'll be in pretty good shape with respect to demand.

Great. Thanks very much for the commentary.

Thank you, John.

And our next question comes from Nat Schindler with Bank of America.

Yes. Hi, guys. I just want to talk a little bit about what you guys mentioned about longer term, about how you could use your platform on other loan types. Obviously, personal loans and auto loans are sold in the ABS markets. If you can prove you're better than FICO, great. You're going to get a premium on that, and you're going to be able to sell your loans better, and you can get funded well. But when you mention things like homes and mortgages, basically sold entirely into GSEs. How could your system work in that world where Fannie Mae is basically deciding who gets a mortgage?

Hey Nat, this is Dave.

It's a good question. And let's just say mortgages are funded in lots of ways, first of all, and some are qualifying, some are not qualifying the GSEs. Rules can change over time. So there's just a lot of ways this market could go. But having said that, if people don't have access to a reasonably priced mortgage today at all, we certainly think there's an opportunity to serve them better. And again, whether funding is from bank balance sheets, whether it's from securitization markets, whether it's from GSEs, or whether it's some combination of those things, is certainly something we'll look at over time. But we feel very confident, given what we do and who we are, that better underwriting somebody who wouldn't otherwise fit easily into a qualifying mortgage using the models that most mortgage originators use today, there's a very large number of people who aren't well served by that market, and we think we can, together with our bank partners, build a better product, and that's what we expect to do. All right.

Great. Thank you. Thanks, Matt.

And that concludes today's question and answer session. I would now like to hand it back over to our speakers for any additional or closing remarks.

I just want to say thank you to everybody.

It's been a great quarter for us and a great year since the IPO. We are really excited about what we've done, but even more excited about What's in front of us, I think we've kind of laid out some of the framework of where we're going, but we feel the company has never been stronger and never more excited about our prospects and what we can do in the financial services industry in the years to come. So thanks all for joining us this quarter, and we'll see you in a few months.

This concludes today's call. Thank you for your participation. You may now disconnect.