Upstart Holdings Inc. (NASDAQ:UPST) Q1 2022 Earnings Conference Call May 9, 2022 4:30 PM ET
Jason Schmidt - Vice President, Investor Relation
David Girouard - Co-Founder, President, CEO and Chairperson of the Board
Sanjay Datta - CFO
Conference Call Participants
Ramsey El-Assal - Barclays
Pete Christiansen - Citi
Simon Clinch - Atlantic Equities
Andrew Boone - JMP Securities
Vincent Caintic - Stephens
Mike Ng - Goldman Sachs
David Chiaverini - Wedbush Securities
Arvind Ramnani - Piper Sandler
Shebly Seyrafi - FBN Securities
Sandy Beatty - Morgan Stanley
Nat Schindler - Bank of America
Good day, ladies and gentlemen, and welcome to the Upstart Q1 Fiscal Year 2022 Earnings Call. Today's call is being recorded.
At this time, I would like to turn the call over to Jason Schmidt, Vice President, Investor Relation. Please go ahead, sir.
Good afternoon and thank you for joining us on today's conference call to discuss Upstart's first quarter 2022 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'd like to remind you that shortly after the market closed today, Upstart issued a press release announcing its first quarter 2022 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 2022 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. [Operator Instructions]
Later this quarter, Upstart will be participating in Barclays Emerging Payments and Fintech Forum on May 16; Citi Beyond the Basics Conference, May 24; BofA Securities Global Technology Conference, June 8; and Morgan Stanley Technology, Media and Telecom Conference, June 14. We will also be holding our Annual Stockholders Meeting on May 17.
Now I'd like to turn it over to Dave Girouard, CEO of Upstart.
Good afternoon, everyone. Thank you for joining us on our earnings call, covering our first quarter 2022 results. I'm Dave Girouard, Co-Founder and CEO of Upstart. I'm pleased to say we're off to a great start in 2022. The Upstart team just delivered our seventh consecutive profitable quarter and our fourth straight quarter with triple-digit year-on-year revenue growth. As the recognized innovator in AI lending, we continue to expand our leadership position in personal lending and are now off and running in our auto lending product as well.
Despite the macro headwinds that appeared over the first quarter, we saw loan transactions of more than $4.5 billion, a record for the Upstart platform and perhaps for the industry as a whole. At the same time, we added a huge number of lenders and car dealerships during Q1. Today, we have more than 500 dealerships on Upstart as well as 57 banks and credit unions, which is up from 42 when I last updated you in February. At this point, we're adding about a lender per week. This is real progress, considering we had just 10 lenders on the platform when Upstart IPO-ed in December 2020.
Additionally, we now have 11 lenders with no minimum FICO score in their credit policies, up from 7 the last time we spoke. I'm confident that our momentum and pipeline for both dealerships and lenders has never been stronger.
We continue to make rapid progress with our auto refinance product as well. In the first quarter, we transacted more than 11,000 auto refi loans on our platform, almost twice as many as we did in all of 2021. We also launched our first AI model for auto refi that is partially trained by our own auto lending performance data. This kicks off the process of building and deploying increasingly accurate versions of our model, which is our primary source of competitive advantage in the market. In Q1, we also more than doubled the rate of instant approvals for auto refi applicants, another major step toward increasing funnel throughput and delivering a differentiated product experience.
Of course, in the recent weeks and months, it's become apparent that 2022 is shaping up to be a challenging one for the economy and for the financial services industry in particular. In my remarks in our February earnings call, I mentioned that the Omicron variant, the clear signs of inflation and the Fed's plans to counter it and the market rotation out of high-growth technology. Since then, it's become clearer just how aggressive the Fed will be with interest rates in order to combat a level of inflation that we haven't seen in decades. The two year treasury note, which is the most relevant industry benchmark for our business, has risen more than 200 basis points since October. And of course, the war in the Ukraine and the zero-COVID policy in China have only increased the risks and uncertainties facing the global economy.
As I said in February, lending is a cyclical industry and always will be. So we expect volume and pricing in our platform to vary accordingly. As a result of increased risk in the economy as well as the corresponding higher returns demanded by banks and credit investors, the average loan pricing on our platform has increased more than 300 basis points since October. In addition to increasing rates for approved borrowers, this also has the effect of lowering approval rates for applicants on the margin. Given the hawkish signals from the Fed, we anticipate prices will move even higher later this year, which will have the effect of reducing our transaction volume, all else being equal.
But if you've been following Upstart for a while, you know that we've been through several disruptions in our industry over the years, and each time, Upstart gained market share and emerged a stronger company. When the economy gets turbulent and nimbleness is at a premium, the advantages of a founder-led company with a closely knit and tenured leadership team become apparent. And that's what you have in Upstart: three founders involved in the business day in and day out and a proven leadership team, half of which have been with Upstart almost since inception.
I'm proud of how Upstart performed in the last two years, particularly during an economic cycle with no precedent. In the worst year of the pandemic, 2020, Upstart grew revenue 42% and generated a modest profit. And of course, our growth rate and our profits since then have been extraordinary by any measure. Even in this challenging environment in 2022, our guidance for full year revenue implies a growth rate of 47% over 2021, and we expect to be cash flow-positive.
With respect to credit performance, we're pleased how our models performed on behalf of our lenders during this tumultuous period. While not perfect, our model significantly outperformed traditional FICO-based risk models and learned quickly while doing so. For Upstart loans originated and funded by our banks and credit union partners, we saw significant overperformance since the beginning of COVID, which has normalized to on-target performance in recent months. There has been no meaningful underperformance of returns with any of our more than 50 lending partners since the program's inception in 2018 despite significant periods of economic disruption.
For loans funded by institutions in capital markets, we've observed more volatility, which is natural given the broader risk aperture. The unprecedented level of government stimulus caused the majority of these post-COVID vintages to overperform significantly. The abrupt termination of these stimulus programs has caused some of the more recent vintages to underperform. And finally, we're confident that our models are currently well calibrated to the latest consumer credit conditions, performing in line with expectations and are more accurate than at any time in our history.
Let's turn now to our new product efforts. One of our most important initiatives for 2022 is the accelerated rollout of our auto retail product. Since acquiring Prodigy in April of 2021, we expanded our dealership footprint from about 100 rooftops at the time of the acquisition to more than 500 today, making Upstart one of the fastest-growing auto retail software in the industry. Upstart's active dealership footprint over the last 90 days spans 35 different OEMs, including Toyota, Subaru and VW. At this point, we're also well into Phase 2, which is the introduction of Upstart-powered loans into our auto retail software. This represents the next critical good step in modernizing the car buying experience.
Without question, our early progress in delivering loans through our retail software has exceeded our most optimistic expectations. While lending has enabled in just a handful of dealerships in California, the uptake and win rate for the loan product, technically termed a retail installment contract, has been far better than anticipated. Our auto teams are working quickly to smooth some of the product edges, filling in a few missing features and completing integrations with various legacy dealer systems, all in the interest of moving toward a broad-based rollout. Our goal has enabled lending in a few dozen dealerships in four states this quarter, representing about 25% of the U.S. population, followed by a full nationwide rollout in Q3.
Based on what we now know, we expect the auto retail lending business to contribute meaningfully to Upstart's monthly transaction volumes by the end of the year, setting us up for a significant ramp in 2023. As I've said before, auto retail is perhaps the largest of all buy now, pay letter markets. So this is one of the most exciting developments in Upstart's history. You should feel confident that we have a lot of executive attention on getting it right.
I'm also pleased to share that we began publicly testing our small dollar loan product in the past few weeks. I first mentioned this to you in our earnings call last November. It's designed to help consumers with unexpected and immediate cash needs, think a few hundred dollars repaid in just a few months. But it's also important to remember that 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. This is a strategic initiative to our mission to improve access to credit, and we believe it will accelerate the pace at which we can bring more and more marginalized Americans into the mainstream banking system.
What I told you in November that we aim to launch the small dollar loan before the end of 2022, our small dollar team set an aggressive goal to launch the product by the end of Q1, and I'm pleased to report that they achieved this ambitious goal. Of course, we still have lots of work to do to realize the opportunity in small dollar lending, but the team's ambition is inspiring.
Additionally, I'm happy to share that our small business lending team is likewise making impressive progress and is aiming to have their product in the market within a few months. The first version of our SMB pricing model will include more than 500 variables about both the applicant and the business. It will also feature our loan month modeling framework, which is one of the most impactful innovations added to our personal loan product a few years back. Our initial testing suggests that version 1 of our SMB model will deliver higher accuracy as measured by Area Under the Curve, or AUC, than peer models that have been in the market for years. We'll begin to cautiously test this new product in the second half of the year.
I'm excited about our SMB product for two reasons. First, business lending is central to far more banks than is consumer lending. So our bank partners are ready and waiting for this. And second, despite the interest banks have in business lending, the FDIC data suggest that 77% of large banks and almost 90% of small banks have no online application process whatsoever. We're also hard at work on some fundamental upgrades to the infrastructure that underpins our AI models and how we develop them.
It's important to realize that the surface area over which we're implementing AI has expanded dramatically. First, we're now working on seven or eight unique models that target different aspects of credit targeting and origination. And we're implementing these different models across five different credit products as of now. Second, the amount and types of data used to train our models has grown exponentially and will continue to do so. As such, the time and processing power required to retrain our models has similarly increased.
So naturally, the opportunity to improve the infrastructure we use to build, train and deploy AI models is enormous. And an effort broadly referred to internally as Machine Learning to Heaven, or ML2H, we're working to dramatically upgrade this infrastructure. Our goals with ML2H are to allow hundreds of research scientists to seamlessly and securely build new models and add data to existing models, train and test them in an automated fashion and deploy them across the entire model ecosystem simultaneously. The system we're working toward will provide maximum leverage to our research scientists, productizing and automating how new models are trained, tested and deployed.
Another important area we're investigating is the means by which our AI models include assumptions about the macro economy. While our models have long considered the current macro context at the time alone is priced, we've consistently said that we aren't and don't aim to be macro forecasters, and yet macro events will always have some degree of impact on the performance of Upstart-powered loans. So given our product is designed to target a particular return to lending partners, that implies there's always some view of the macro future inherent in our models.
Given this reality, we intend for our product to explicitly share the macro adjustments that are embedded in the models, and furthermore, to allow our partners to put their own macro assumptions. This will provide significantly more transparency to our lending partners and will also put our focus squarely on risk ranking, which is the heart of what makes Upstart's models unique.
A few weeks back, we celebrated Upstart's 10th anniversary. It was a wonderful opportunity to remember all we've been through, to stop for a moment to reflect on how we got where we are today and to show the gratitude that I feel for all those who have been part of that journey: our employees, past and present; our investors and partners; and of course, our friends and families that made this all possible. Some of the old timers, the OG Upstarters, if you will, shared some of their favorite moments in Upstart history. We talked about the street curve we sat on from lunch each day in our Palo Alto office because we had no better place to gather.
When I had the chance to speak to the team, I told them that I'm not particularly adept at celebrating the past. It's just not me. I'm far too excited about and paranoid about the future to spend too much time toasting to our success in the past. As we shifted towards talking about the future, I told our team we need to act with urgency today with a healthy dose of paranoia. We're a company grounded in reality with our eyes wide open as to the evolving risks we see in the industry and in the world. And at the same time, we have to pair this urgency about the present with optimism and absolute determination about the future.
Fortunately, most of our leadership team has been here. So we know the drill and are confident that we can navigate whatever 2022 and beyond might hold. And we know we can blame that urgency for today with the optimistic eye on the horizon because although we serve a cyclical industry, we represent a secular change that the financial services industry desperately needs. Artificial intelligence will reshape the economics of lending in ways that will reverberate for decades. We're today pursuing opportunities that represent more than $6 trillion in annual origination. So there's little question about the scale of the addressable market. We see a clear path to building a company with more than $10 billion in revenue in the coming years and are maniacally focused on achieving that goal.
Thank you. And now I'd like to turn it over to Sanjay, our Chief Financial Officer, to walk through our Q1 financial results and guidance. Sanjay?
Thank you, Dave, and thanks to everyone for being with us today. Just to quickly call out the key financial trends before diving into the more detailed numbers. On the top line, origination volumes and revenue from fees were both slightly up from last quarter, which was encouraging given the seasonal drop we typically see in Q1. Profitability was ahead of guidance but as anticipated, down sequentially from last quarter, partly due to the ramp of auto lending.
We've continued to deploy our balance sheet assertively in the service of R&D in both auto lending and new segments of personal lending as well as using it to smooth fluctuations in funding of corporate loans. And as Dave outlined, the macro environment has become an increasing headwind to growth this past quarter with both rising interest rates and rising consumer delinquencies putting downward pressure on conversion.
With these dynamics in mind, here now is a summary of our numbers. Net revenues in Q1 came in at $310 million, up 156% year-over-year. Revenue from fees constituted $314 million of that amount, representing 101% of overall revenue and up 9% sequentially from last quarter. Net interest income was a negative component of net revenue this quarter as the loan assets on our balance sheet which we mark-to-market each quarter sustained declines in valuation due to the rising interest rate environment.
The volume of loan transactions across our platform in Q1 was approximately 465,000 loans, up 174% year-over-year and representing over 350,000 new borrowers. Average loan size was up 18% over last quarter, an indication for us that fundamental loan demand from borrowers is back on the rise after being suppressed for more than a year due to government stimulus.
Our contribution margin, a non-GAAP metric, which we define as revenue from fees minus variable costs for borrower acquisition, verification and servicing, declined from 52% in Q4 to 47% in Q1, a level which was nonetheless 100 basis points above guidance. Our declining contribution margin was almost entirely a function of the expected ramp in auto lending, which remains contribution-negative at this early stage. Without the effect of auto loans, our contribution margin for personal lending would have clocked in at a robust 51%.
Operating expenses were $275 million in Q1, growing 13% sequentially over Q4. Sales and marketing and customer operations spend, both typically viewed as variable costs, each outgrew revenue this quarter at 16% and 18% Q-on-Q, respectively, due to the additional onus of acquiring and onboarding auto loans. Engineering and product development grew 8% sequentially due to slower hiring than targeted and remains our priority area of investment. Growth in general and administrative spend grew 3% sequentially.
Taken together, these components resulted in Q1 GAAP net income of $32.7 million, up 224% year-on-year and above our guidance but down 44% sequentially from Q4. Similarly, adjusted EBITDA exceeded guidance at $62.6 million and grew 198% year-on-year but slid 31% quarter-over-quarter. Adjusted earnings per share for Q1 was $0.61 based on a diluted weighted average share count of $95.5 million.
We ended the quarter with $1 billion in restricted and unrestricted cash, down from $1.2 billion at the end of last year as more of our capital base flowed into loan assets in support of R&D programs, primarily in auto refi and some newer segments of personal loans. Additionally, we have started to selectively use our capital as a funding buffer for core personal loans in periods of interest fluctuation where the market clearing price is in flux. Our balance of loans, notes and residuals at the end of the quarter was consequently up to $604 million from $261 million in Q4.
Since our prior earnings release, the level of uncertainty in the macro environment has continued to grow. After remaining at historically low levels for the past 18 months, loan default rates rose quite abruptly towards the end of last year and are now back to, or in some cases above, pre-pandemic levels. This is a dynamic we have observed consistently across the full breadth of our portfolio although one which appears to be disproportionately impacting higher-risk tiers, which are generally composed of borrowers who one might assume have a greater exposure to loss of government stimulus.
As a way to keep investors abreast of such credit trends, we have introduced new information in our investor materials, which shows in aggregate for all historical vintages the in-period loan defaults compared to the aggregate defaults that were predicted across those vintages at the time of their origination. The drop and subsequent reversal in developed trend that is shown on the chart are in our view a function of the injection and subsequent waning of the government stimulus. And as a consequence, virtually all of our pre-2021 vintages will substantially outperform their return targets while the two or three vintages most adjacent to the reversal in trend at the end of 2021 are set to underperform.
Separately, interest rates have continued to climb in response to inflation signals and Fed tightening. The combination of inflation and monetary tightening imply the nontrivial risk of a recession potentially later this year. Given the general macro uncertainties and the emerging prospects of a recession later this year, we have deemed it prudent to reflect a higher degree of conservatism in our forward expectations.
With this as context, for Q2 of 2022, we are expecting revenues of $295 million to $305 million, representing year-over-year growth rate of 55% at the midpoint; contribution margin of approximately 45%; net income of negative $4 million to $0 million; and adjusted net income of $28 million to 30 million; adjusted EBITDA of $32 million to $34 million; and a diluted weighted average share count of approximately 96.2 million shares.
For the full year 2022, we now expect revenue of approximately $1.25 billion, representing a growth rate of approximately 47% from the prior year, down from $1.4 billion guided last quarter; contribution margin of approximately 48%; adjusted EBITDA of approximately 15%.
Thanks once again to everyone at Upstart who is working hard to move our mission forward. And with that, Dave and I are now happy to open the call to any questions.
Operator, back to you.
[Operator Instructions] We take our first question from Ramsey El-Assal with Barclays. Your line is open. Please go ahead.
My first question is on the conversion rate. It came in a little lower than our model. And Sanjay, I know you mentioned that some higher delinquency rates might have -- might be putting some pressure there. So I guess, first in quarter, maybe talk about the sort of puts and takes with the conversion rate. And then also, what should we be expecting as we move forward throughout the year?
Ramsey, this is Sanjay. Yes, thanks for the question. I think it's related what we just described, which is two things: the rates -- the interest rates in the economy go up, and consequently, referrals for banks and for institutional buyers go up. And similarly, as we've talked about, between roughly November and February, delinquencies in the economy reverted as well. They have been unnaturally low for about 18 months. And with the sort of the waning of the government stimulus, in our view, those trends have reversed. And those two things result in higher rates, interest rates quoted to consumers, and that results in lower conversion. So I think that's the totality of the story.
How that plays out for the rest of the year, I think, is a function of those two things. We view the reversal in the delinquency trends to be stable now and has been stable for about 60 days in our view. There is the -- as I said, there's the prospect of higher rate hikes or, I guess, higher interest rates. To now at the end of the year, there is a risk of something like a recession as inflation plays out and as the Fed continues to tighten. And so those two things could present further risk to the conversion rates. But I guess it would sort of depend on your view of how the macro environment will evolve for the rest of the year.
Fair enough. I also wanted to ask you about the ABS side of your business. You mentioned that you'd seen some dislocation that sort of worked its way back to something more normal. I guess, talk about the demand environment for your loans as we stand today. And then also maybe clarify whether you foresee needing to incur any kind of residual liabilities or any other kind of handholding measures and get -- in order to get those transactions kind of consummated or whether you think the environment is not really going to demand that as we move forward.
Yes. Sure. So the broader picture on loan demand, as you know, there's a variety of different channels by which the loans get funded. I think that the loans that are getting funded through bank balance sheets or credit union balance sheets, I would say, is among the more resilient because they have a cost of funds from their deposits that tends to be a little bit insulated from the way that the yield curves are moving the economy. And then we have buyers that are a part of the institutional world that are buying and holding on balance sheet. And I would say those are a little bit more exposed to interest rate movements in the economy but still somewhat resilient. And then there is the amount of the loans that are getting from those buyers sent to the ABS markets, and that's a fraction that has changed historically as a result of how constructive the ABS markets have been.
Obviously, the ABS markets in 2021 were historically constructive. So you could make a very healthy gain by purchasing a loan and selling it into the ABS world. And so a lot of that activity happened. That opportunity, as rates have tightened and this cost of funding has gone up, is far less today. And so the amount of product getting sent to the ABS markets is a lot less in 2022 than it was in 2021.
But I don't think that really changes our equation with respect to how we participate in the ABS markets. We've always been a little bit at arm's length in the sense that it's not our balance sheet that's contributing meaningfully to those ABS deals. It is loan buyers and investors and other institutions that are making the decision to contribute or not. And we don't really retain any residual risk in those deals, and it's not a plan to do so going forward.
We take our next question from Pete Christiansen with Citi.
Dave, Sanjay, I was just hoping you can elaborate a little bit on that last comment, Sanjay, made about, I guess, credit trends seem to be stable over the last 60 days. If you look at some of the rollovers on delinquencies, I guess there's some concern that you could trigger some of the ABS C&L thresholds. Just wondering if there's a concern there at some point. And could this impact your ability to attract funding? Then just to my second question, Sanjay, you mentioned you've upped the balance sheet risk here a little bit. How far are you willing to go in terms of supporting new loans and putting warehouse liabilities on the balance sheet?
Yes. Pete, thanks for the question. This is Sanjay. Let's see. Your first question is about pattern of diligences and ABS triggers. The delinquencies just sort of in absolute level, as I said, sort of they renormalized to pre-pandemic levels fairly abruptly starting in October and November. I think that pattern stabilized in sort of roughly February. To us, March and April were very stable months.
The question of triggers on ABS deals, so it's a somewhat technical thing. But what will breach triggers on an ABS deal aren't the sort of in-period delinquencies, but it's the charge-offs from delinquencies of prior months. And we've done some modeling on that. I think that with respect to our large ABS deals, I don't think there's much concern of breaching triggers. We do some smaller monthly sort of pass-through issuance. And there is a possibility that those triggers will be breached.
That's a somewhat technical thing. It basically means that the resid holders are locked out of cash flows for a couple of months while the bondholders get replenished, and then the cash flows begin to flow again. So it would impact their returns. And it's not something that rises to the level of seriousness of us worrying about ability to traffic those deals. And I think it's sort of a temporary technical thing that will cause some interruption of cash flows between bondholders and residual holders.
So -- and then your second question is how we plan to use our balance sheet. And as I said, I mean, historically, our balance sheet has been almost exclusively for the purpose of R&D. We have used our balance sheet in the last quarter to do what I call sort of a market-clearing mechanism. And by that, what I mean is when interest rates in the economy change quite quickly, I think it would be fair to say that our platform, its ability to react to the new market-clearing price, it's probably not as nimble as we would like. It's somewhat manual. It requires a bunch of conversations and phone calls.
And so when interest rates smooth and investors are -- so each deciding what their new return hurdles are, there can be a gap or a delay in responding to funding. And that's a situation where we've chosen to sort of step in with our balance sheet and almost sort of bridge to the new market-clearing price. And is that is happening often and abruptly. We've been sort of playing that role with our balance sheet.
I don't view that to be a long-term or necessarily sizable activity for us. I think that developing the mechanisms to respond more nimbly to new price discovery as rates change is something that's on our road map, and it's something that we want to start to invest in so that it can happen in a much more automated way. At the end of the day, we view our platform as being a platform that responds to risk and rates in the environment. And so the faster we can do that, the faster we'll be able to deliver the new returns in any given scenario to the investors and not have to bridge it with their own balance sheet. So I kind of -- I would view it through that lens.
Next question from Simon Clinch with Atlantic Equities.
So I just wanted to -- in terms of your guidance for the year, you're explicitly, I believe, building in a recession scenario late in the year. So I was wondering if you could help us just sort of bridge the gap to sort of what's changed in terms of the mix of your loans. You talked about before having embedded $1.5 billion of auto loans for the year. What's changed there? What's changed on the personal loan side? And then I have a follow-up just probably on the last points you were just making.
Sure. This is Dave. I might actually jump in, at least once here. Yes. I mean, I think it's actually fairly simple. We expect less volume than we would have a few months ago based on pricing in the marketplace being higher, and that's a function of both underlying base rates being higher as well as the risk in the environment and the risk premium that the lenders or investors are demanding. So you put that all together, and it's been an increase in sort of average rate to the consumer of several hundred basis points.
And I think we've always said, we're not, in some sense, a terribly interest rate-sensitive thing. And we've kind of said interest rate changes of 50 basis points or 100 basis points are something that could well be offset by improvements to our platform. But in this case, it's much more significant than that. And of course, in the last few months, we've seen -- and it is probably 300 basis points or higher. And so that's what's giving us our guidance for the rest of the year.
We aren't -- we're not -- in no way predicting a recession or anything like that. It's not really our job to try to do that. We're just reflecting the prevailing rates in the marketplace and the loan transactions that, that typically translates to, and that's what you're seeing there.
And just to follow-on on the last point. I'm kind of surprised to hear that you're using your balance sheet to put some loans on there, which aren't just for R&D purposes. And it strikes me as it's just not a normal course of business for you given you're a platform business for. So I was just wondering what kind of message that might send to your bank partners or to others in the system. Just curious about that.
Look, I think it's -- Sanjay kind of explained this. Generally, we view ourselves as a marketplace where ultimately, price discovery happens and loans are funded or not funded when they make sense by our bank partners or by capital markets, institutional investors, et cetera. Some of this works very fluidly, particularly on the bank and the credit union side, where they have very direct controls to change their return hurdles, et cetera. We don't have those mechanisms in place as well on the other side. It's more of a manual process.
So when something changes as quickly as it did in interest rates and the risk premiums in the market changed very rapidly, then we stepped in to sort of bridge that, but it's sort of a temporary thing. And as Sanjay said, it is an intention of ours to make the system more automated and more fluid so we don't have any need to do that. It's not part of our business to hold loans generate net interest income from loans in our balance sheet, but we certainly want to make sure there's fluidity in the system. And we're definitely going to do some more work so we can do that without any of our balance sheet participation in that.
And Simon, just to maybe put the numbers into context, but I think the amount of the total platform loans that ended up on our balance sheet this quarter was still a single-digit percentage. And of that amount, probably close to three quarters of it is still -- was still R&D-style spending on predominantly auto loans and other new products and segments. So it was still a relative minority or a relatively small percentage, but it is just sort of an important new thing that we haven't been doing in prior quarters just because of the fluidity of the environment.
We take our next question from Andrew Boone with JMP Securities.
As we think about rising interest rates that you're offering, can you just help us understand the spread between what consumers are replacing, right? So if I think about credit card rates going up and other credit products also rising, are you guys implying that the spread between what is being refinanced is just broadening out and widening? And can you help me understand that?
And then secondly, as we think about auto refinancing, there are clearly some headwinds going on there with auto prices coming down, rates coming higher. How are you thinking about the auto refi business for 2022? Can you just double-click on that?
Sure. Thanks, Andrew. On the first question, I mean it's actually fairly straightforward. And it's not even as -- you don't have to think as deep as like what a person might be refinancing through an Upstart-powered loan. It's really as simple as when the consumer rates go up, that means on the margin, a whole bunch of people that would have been approved are no longer approved. So there's a whole bunch of just loans that never happened at all. And there's a bunch of people that are still approved, but the interest rate is a few percentage points higher, and a certain fraction of them are going to decide that's not the product that they want. They don't need it. It's -- in many, many cases, it's a discretionary loan where they're buying something or paying off something that they don't necessarily have to.
So there's a fair degree of price sensitivity. And just put those together, when average rates go up, you're going to see less volume. When average rates go down, you're going to see more volume. There's a lot of things we're, of course, doing in the mix of that to make the funnel more efficient and more performance. But all else being equal, of course, if rates go up, those are the effects that you're seeing there.
On auto refi, I would just say there's a lot in flux because it is a new product and because it's also a refi product, meaning it has interest rate sensitivity to it. So a little hard to judge how those things will balance over the course of the year. And that's kind of -- it's a new product for us. We don't have as much history as we do in personal lending. But certainly, again, it's an interest rate-sensitive product, so that works against us. But we're also making some pretty rapid progress in terms of automating, people refinancing loans, getting much more focused on digital signatures away from wet signatures, et cetera, et cetera. So there has been some really good progress there, but how those will trade off over the course of the year is just a little hard to predict right now.
We take our next question from Vincent Caintic with Stephens.
I appreciate all the products you're rolling out and the talk about the addressable market, but I guess the biggest investor worries on the funding side and not having the funding to fuel your origination volume. So two parts question. So first on that balance sheet, $600 million. Understanding it's not supposed to be a long-term thing, but if you can maybe size what your appetite is. And I guess my understanding was that only loans that were approved by the funding partners were to be originated. So I'm just wondering what's different there.
And then secondly, if you can talk about kind of the funding partnership. So how is interest? That -- I saw that you had 15 new bank partners this quarter. So if you could talk about that. And then the forward flow agreements, if you could talk about your conversations with your institutional investors.
Yes. Sure. Maybe I'll start off on the institutional side. I don't think we have a specific target or number in line with respect to how we're managing our balance sheet. I mean it's sort of an ongoing trade-off between wanting to spend those dollars in R&D and new products to sort of incubate new models and calibrate new models versus buffering the core business through interest rate shocks while we're waiting sort of investing, making the platform sort of more nimble to reacting to finding those market-clearing rates. And how we balance the two is I would say a bit of a fluid equation. So I don't think we have an easy heuristic on how we would prioritize the two.
But certainly, in Q1 with all the volatility we saw even though I would say that the majority of our balance sheet spending was still of the R&D flavor, just making sure that the core business was stable, continue to be an important component for us.
And I don't know, Dave, do you want to talk a little bit about the lending partners side?
Yes. We actually had a phenomenal quarter, significantly our best ever in terms of signing up new lending partners and deploying them as well. So doing exceptionally well, particularly with credit unions. So I think both in terms of what we signed up and deployed onto the platform as well as what our current pipeline looks like has never been stronger. So we're extremely pleased with the progress we made. We're adding about 1 lender every week. And I think there's certainly an opportunity to accelerate from there. So lender adoption has certainly been a highlight.
And just a follow-up question on the guidance. So the revenue adjustment to your guide, is that all transaction volume? Or is there any change to assumptions to your take rate? And if you wouldn't mind, if you do kind of have an assumption for your transaction volume, that would be helpful.
Volume? But -- I think I would say the entire difference is transaction revenue-related. We don't have any explicit assumptions on balance sheet for any -- of that nature.
We move to Mike Ng with Goldman Sachs.
So I just had a follow-up on the revenue outlook change. It sounds like it was mostly driven by a lower origination outlook due to lower demand from rising rates. I just wanted to see if you're seeing any change in the lending parameters or shrinking of the credit box due to what's happening on the funding side.
Mike, this is Dave. No, I don't think there's any significant change in that regard. I would say, generally, most -- on the banks and credit union side, there's been very little move in any direction. And I would say in some cases, they've raised target returns a bit, but I think it's just at the margin. So largely, that side of the house has not changed significantly. It's really on the other side where expectations of investors and -- et cetera, has gone up in terms of return targets. So no, it's not really any real change with respect to the credit box of the various lenders.
And I just wanted to follow up on some of the discussion around the loans on the balance sheet. What's your assumption for where that goes for the rest of the year? Do you plan to simply allow those to mature? Are you going to kind of sell them into the market when you have the opportunity? Just any thoughts there would be great.
I would say those are being held as held-for-sale loans. Obviously, it will be a function of what that secondary market looks like. And we'll make an economic decision on them, but I think the preference would be to get liquidity and get cash back on the balance sheet rather than to hold the loans and earn interest income.
We take our next question from David Chiaverini with Wedbush Securities.
I wanted to ask about the recent change to your loan modification policy in which you made it easier for Upstart borrowers to obtain forbearance and these delinquent borrowers would be considered current on their loans. Can you discuss why you made this change? And are delinquent borrowers automatically entered into the forbearance program? Or do they have to opt into it?
This is Sanjay. Yes. So on loan modifications, I guess a simple way to describe it is -- and even independent of the macro environment, we had -- certainly through the early days of COVID when unemployment shot up and there was a lot of requests for forbearance, we have a lot of data that led us to conclude that you're always -- or you're generally better off creating -- in times of, call it, macro stress creating flexibility for borrowers and helping them work through those periods of turbulence.
I think in rough terms, if you charge off a loan, you're probably getting something in the order of $0.10 to $0.12 a dollar on the recovery. If you run an effective forbearance program, you can sort of get sort of on your $0.30 to $0.40 on the dollar. So we had had plans anyways to sort of begin to optimize the way we manage modifications in forbearance. But I would say that this macro environment accelerated those plans because it's made it that much more valuable to the borrowers.
It's not an automatic enrollment thing. It still does require an application and review. But I would say we have all done -- and we loosened our standards to make it a program that was sort of more widely applicable.
And my follow-up question relates to the fair value adjustment. It looks like in 2018 when interest rates were rising, the fair value adjustment was about a $40 million headwind to revenue. And then when interest rates fell, the fair value adjustment was about a $30 million tailwind in 2021, which probably would fall right to the bottom line of EBITDA. Could you discuss how we should think about that line item this year and how it impacts your EBITDA forecast given the rising rate environment?
Yes. Sure. I guess -- so that's always been, I would say, a small component of volatility on the P&L. As you know, it's sort of deprioritizing our business model.
The reason of the magnitude for it in 2018 was a bit of a technical one. In the early days of our securitization program, we were doing risk retention ourselves because we were a relatively new entrant into the market. We couldn't convince anyone to do it for us. So we had to consolidate a lot of those securitizations onto our balance sheet. And as a result, there are sort of big positives and big negatives. We didn't actually have any economic exposure to those deals. So some of that is a bit optical.
I would say, in 2020 and 2021, you see our balance sheet reduced quite meaningfully. Some of that is just the deconsolidation of those ABS deals. The way I would think about it going forward is it's sort of a function of the scale of our balance sheet and what's going on in the economy. So for right now, our -- the scale of our balance sheet is a little bit bigger than it was last year. And obviously, when you put loans in the balance sheet, there's two different accounting treatments. Our particular accounting treatment is we mark our loans to market every quarter. And so when interest rates -- you're holding a loan and the interest rates are going up, you'll take a fair value hit. And so that's what's happening right now.
I guess the answer to your question depends a little bit on what happens to interest rates, obviously. If they continue to go up -- and in particular, as Dave mentioned, the one that's particularly relevant to us is the two-year treasury. So it's about 250 basis points higher than it was late last year. If that's topped out and it's stable, we shouldn't see any further fair value devaluations. But for holding loans on our balance sheet and that number continues to go up, there will be further devaluations.
So I guess in the grand scheme, I kind of view this as it's going to create some positive and some negative volatility from year-to-year. I think in the long-term sort of framework of the business, we don't really view it as meaningful to the business opportunity we're trying to execute against.
We move to Arvind Ramnani with Piper Sandler.
I just wanted to ask about some of the topics you've talked about. You're putting loans on your balance sheet. And I know, Sanjay, you clarified most of it is more for R&D, but there is some kind of some portion -- I don't know if you're able to sort of quantify what's R&D versus kind of just taking on like kind of regular risk. And then the second part of the question is, under what circumstances should we see that number go from the single digits to double digits? Is that possibility on the table as we look out for the rest of the year?
Sure. Yes. I guess the answer to the first question, Arvind, is I think in Q1, probably about three quarters of our loans are R&D. The majority of what we balance sheet was still in the auto segment, which we consider to still be sort of the R&D phase.
What is the -- what are the potential future scenarios? It's a good question. To some extent, it depends on the -- it's a function of the level of dislocation in the economy and the progress we've made in, I would say, automating our platform's ability to react to new market-clearing rates. So in a world where our platform exists as it does us today and there's further sharp shocks to the interest rate, and what's happening is the prices on our platform are not reacting in real time to the return requirements of the investors as they're changing. That's the situation in which our balance sheet has sort of stepped in to create stability.
So I would hope that, on the one hand, we're creating more of an automated mechanism in our platform's ability to react to those changes in the economy so they would happen in real-time investors, maybe at the limit could just set their own rates as they change. And it sort of depends also to some extent on the stability of the rates in the economy. So those are a bit hard to judge, but it's sort of a real-time equation for us as the environment changes.
And then just a follow-up on auto. Remember, we had met in New York. Maybe it's been a couple of months now. One of the topics you discussed was auto. Kind of -- you kind of looked to kind of take the large proportion of the auto loan to kind of work with the banking partners. Is there like a time line change just given circumstances changed quite a bit in the last couple of months? Or do you still feel like you can stick to the original sort of time frame?
I don't think that's changed. I mean we've kind of thought of this year for the refi product, we are bringing lenders on. I think we have 12 signed up now. So our first priority is bringing lenders onto that platform, and we'll do that through the second half of this year. The auto retail product is different and much earlier, and I would view that as something where we'll move toward bringing lenders and -- on toward the beginning of next year. So there are different time frames, but I don't think our expectations for either of those products have changed.
Moving forward to Shebly Seyrafi with FBN Securities.
So I was impressed by your 18% sequential growth in your average loan amount. And so it hit $9,700, but you were at $13,000 at the beginning of 2020. Do you think you can approach that number anytime soon?
Shelby -- oh, go ahead, Dave.
Well, why don't I start. We don't expect that -- I would not expect it to go back to $13,000. I think, basically, our -- is our systems get better, proving more people at the margins, you're going to trend towards smaller loan sizes. There's a lot of good things about having better support for smaller loan sizes in terms of what people need. We're now launching the small dollar product, which is, of course, even smaller.
So I don't see a dynamic, other than maybe a short-term thing here and there, that would drive loan size up significantly. It can be just driven by demand, and we may get some high-quality, very prime bank lenders on the platform, which would typically have larger average loan sizes. So it can be some puts and takes like that. But if you looked across the whole platform and you look across a significant period of time, it's actually kind of hard to imagine loan sizes going back to where they would have been back then.
And my follow-up is that the loan amount on your balance sheet is about $600 million. So can we -- or can you say at least that, that was a high watermark for the year or that's going to be the high watermark for the year? Or do you think you can grow meaningfully over the next couple of quarters?
I know that we could say that that's the high watermark. I mean our plan's to continue with R&D programs. The demand there is growing. As we get out of the auto refi sort of balance sheet program, we're going to start to shift some of those dollars to retail as they become more significant later in the year. And as Dave said, we think we'll be in market with small business lending.
These are all programs that investment markets and banks are going to want to see some amount of curves and history before they start investing significant dollars so that they -- and argue will require some incubation. And the other side of the equation is the economy does remain very fluid right now. And if we do need to step in with our balance sheet to sort of to stabilize the core business, I think that's an important tool for us. I don't think we would ever do it in any way that started to exhaust our capacity or even come close to it. But it is a tool that -- although it's not our objective, it's an important stabilizer of the business as -- when the waters are choppy, so to speak.
Next question from James Faucette with Morgan Stanley.
This is Sandy Beatty on for James. Can you remind us how the user limits impact what your platform can do, particularly for those lower-end consumers? And I guess, are there scenarios where you run into issues here just in terms of capping what would otherwise be approved as APR offers move higher? And how does that impact your market share trajectory?
Sure. So basically, there's 50 -- some 50-plus lenders on the platform. They each have their own sort of statutory arrangement they operate under. It could be state-chartered, could be nationally chartered. So they have different rate caps that they observe themselves, and we have no say on that whatsoever. If you look across the entire platform, we have made a decision a long time ago that we don't support any lender originating loans above 36%. So that's the statute that all sort of nationally chartered banks are under. And we just kind of made the decision a long time ago, that's where we want to put the limit for the platform overall. So we can control that.
So the bottom line is, we -- when -- there were a lot of people that three months ago might have been approved close to 36% that today would not be approved at all. So for sure, when rates go up and return hurdles go up, it has the effect of pushing people out of the approval band and into the decline band. And that's just the nature of the business. And when the model gets a little smarter, it will approve some of them and it will disapprove others that it might have approved before. And that also is the nature of the business, the nature of the product getting better.
But that's the bottom line. We don't per se have any say in what a bank or credit union's maximum APR is. That's a function of both their own business model and their own regulatory regime that they operate under.
And just as a quick follow-up. Take rates, in a tighter funding environment with downward pressure on conversion, is there a scenario where we could see changes to unit economics just within that funding backdrop? Just wanted to make sure I ask.
Sandy, it's Sanjay. I guess the short answer is yes. As macro environment becomes more challenging, you might see us become more conservative in how we're managing the business, and that could result in higher take rates or bigger, larger unit economics from lower marketing and operations costs.
We take our final question from Nat Schindler with Bank of America
So you mentioned that the -- you outperformed heavily during the pandemic because of stimulus. It makes sense. And now your loans are showing signs of underperformance, particularly in the worst -- the lowest-quality loans, the lowest-FICO score loans, I'm guessing. You do the bulk of your low-FICO score lending really started late last year. And what does that mean for your -- the performance of your entire portfolio of loans that would have been 2022 or later? Are we looking -- or maybe even Q4 2021 originations and later? Are those going to be even significantly worse than you were suggesting because they're weighted towards those lower-quality loans?
So now a couple of things of clarity because the statements you're making are entirely accurate. First of all, most of the effect we saw from the decline from the sort of retraction stimulus was across the board, all flavors of loans. So I mean there was some marginal difference between the highest-risk loans and the lowest-risk loans. But largely, it was an effect that we saw that was universal. And for us, it's a very clear sign that it's a macro effect. So that's one important thing to say.
Another important thing to say is actually has nothing to do with FICO or whatsoever. It's really by our own risk grades that we look at these things. FICO -- and you can actually see this in some information we put into the investor deck. Our risk separation was dramatically more than FICO. So it's not really related to FICO scores whatsoever.
But there you have it. I mean, I think that we're very happy with how the model is performing now. It has been stable, as Sanjay said, for the last 60 days. I think the bottom line is when you go through a period where the government essentially pays everybody's salaries for most of a year and then suddenly stops, it's a pretty damn difficult thing to calibrate exactly right. And so I don't think we're exactly right. But our risk ranking has been exceptional. The product is stable today. And none of our banks and credit union partners have seen any underperformance. So overall, we're actually feeling quite good.
What's the risk something like? Oh, sorry.
Right now. I was just going to maybe again describe. I kind of -- I know you're -- sort of the direction you're heading in. But the effect was broad. As Dave said, it was maybe slightly more on high-risk loans than low-risk loans. But the real relationship to cohort performance is on timing. And by that what I mean is, if you're sort of the -- if you're an investor buying happily or a bank sort of originating happily across the past couple of years, what you'd see is probably 12 vintage quarters of overperformance. The ones that were timed right around the injection of the stimulus, so call it, late '19, '19 to sort of early 2021, those will overperform dramatically, so maybe kind of like 2 times the return target.
And then the sort of the two or three vintages that were right around the time of what we call the reversal of the loss trend, and you could see this in our investor deck, it was sort of like the Q2 and the Q3 2021, will marginally underperform into the tune of, we'll call it, 20%. So really, the underperformance we're talking about, it's tempting to kind of want to relate to the change in mix we've had over the last year, but really what it is, is it lines up almost exactly with the injection and the sort of the evaporation of the government stimulus. And as Dave said, there's marginal differences between the low and the high end of risk. But I think that's the secondary story. It's not the main driving effect here.
Would then it be reasonable to assume if student loan forbearance ends as expected that it would have a compounding effect on this? It would be the opposite of stimulus?
If forbearance ends, meaning if they start demanding the repayment of student loans? Or if they forgive student loans?
Yes. They start to main repayment again as expected. I mean it's still unknown what they're going to do, but keeps getting delayed. So...
Yes. I think that would be another example of reversal of stimulus, right? They've sort of allowed forbearance for a long time. And so I think it would just maybe be a continuation of the trend I'm describing. I think the majority of the excess balance sheet in our economy has been unwound. You'd see like the personal savings rates, they shut up in a couple of different instances over the course of 2020 and early 2021. And those savings rates are right back down to where they were before. So there may be some -- maybe not all of the stimulus has been rained out of the economy. You could sort of describe it that way, but we think a lot of it, if not the majority of it has. And so there may be a bit more to go. But as I said, we've seen stable trends now for a couple of months.
Is there any demographic -- kind of are you more demographically targeted towards younger people who are likely to be -- to have student loans? Or are you more broad or you targeted somewhere else? Just wondering how much -- how big of an impact that could be on your borrowers.
I would say in the past, we definitely like the inception of our company, largely it was around tensile borrowers, but I don't think that's the case anymore. I think we have a pretty broadly distributed portfolio across the risk grades and the demographics now. So I don't know. I think it's the -- the biggest platform in this space will probably be -- I think you could imagine trends will be roughly proportional to the general population talk at this point
It appears there are no further questions at this time. I'd like to turn the call back to Dave Girouard for any additional or closing remarks.
I just want to say thanks, everybody. We are actually quite pleased and quite happy with the results. Definitely appreciating also that 2022 is a complicated year in the economy and a lot of open questions, but we are exceptionally confident in the strength of the business and is optimistic about our future as we have been. So thanks to everybody for joining us today.
That concludes today's call. Thank you for your participation. You may now disconnect.