Better Home & Finance Holding Company (BETR) Earnings Call Transcript & Summary

March 13, 2026

NASDAQ US Financials Financial Services earnings 64 min

Earnings Call Speaker Segments

Tarek Afifi

executive
#1

Good morning, and thanks for joining us for Better Home & Finance Holding Company's call covering both the fourth quarter and the entire 2025 fiscal year. I'm Tarek Afifi, part of Better's Corporate Finance team. With me today are Vishal Garg, our Founder and Chief Executive Officer, and Loveen Advani, our Chief Financial Officer. Alongside this call, I'd ask you to take a look at the earnings release for the fourth quarter and full year, which you can find on our Investor Relations website. An investor presentation is posted there as well. Some of what we say today could be considered forward-looking statements under federal securities laws, built on our present expectations and assumptions; those expectations and assumptions carry risks, uncertainties and other factors, which we describe further in our SEC filings, and those could make our actual results come out materially different from our historical results; aside from what the law requires, we take on no obligation to update forward-looking statements. During this discussion, management will go over certain non-GAAP financial measures, which in our view are relevant when evaluating the company's financial performance. You shouldn't treat these non-GAAP financial measures as substitutes for our GAAP results, and they ought to be read alongside them. Reconciliations of these non-GAAP financial measures to GAAP financial measures appear in today's earnings release and investor presentation, both of which are up in the Investor Relations section of Better's website and, once filed, in our annual report on Form 10-K filed with the SEC. Additional information as of, and for the period ended, December 31, 2025, will be available once we file our annual report on Form 10-K with the SEC. At this point, I'll hand things over to Vishal.

Vishal Garg

executive
#2

Thanks, Tarek. Morning, everyone, and welcome to the call covering our fourth quarter and full year 2025 results. First, I want to extend a warm welcome to Loveen Advani, who is joining us as our new Chief Financial Officer. Loveen brings seasoned strategic and operational finance leadership, with a proven record of steering companies through growth and transformation. Time and again, he has shown he can bring strategy, capital allocation and execution into alignment. As we pursue our strategic and financial priorities in this next chapter of anticipated growth, his experience and leadership style will be instrumental. Beyond that, I love him because he rolls up his sleeves and gets his hands on keyboard. At the outset of our relationship, he forwarded a model to me, and the two of us started collaborating on it, just the pair of us, during late-night hours. For the next phase of this company's Blitzscale expansion, that is precisely the sort of CFO that is required, and having Loveen here makes us extremely, extremely glad. Better is a vertical AI platform that is thoroughly transforming and revolutionizing the home finance sector. We are constructing the AI native frontier of consumer finance, and by doing so, we allow companies that have enormous customer bases to offer mortgages and HELOCs to their customers in an AI-first manner, while also empowering the existing network of local retail mortgage originators. Adoption throughout the ecosystem verifies that this change is genuine and speeding up. That's what the Tinman AI platform can do. Over the last ten years, we've created an AI-driven matching engine — the first of its kind — that links consumer credit data, income data, asset data and property data to the preferences of about 40 different investors on our platform, which lets us approve mortgages and home equity loans almost instantly. What you get is a process that's quicker, less expensive, simpler and simply superior. We're undergoing a real transformation: once a direct-to-consumer mortgage business for consumers who came to Better.com, we're now an AI-native mortgage platform serving the whole mortgage industry. During the last ten years, we assembled the tech, the physical operations, and the ties with investors needed to produce home loans more quickly and at lower cost than competitors. Now we are using that base and stretching it throughout the whole ecosystem — giving partners who bring huge customer rosters the tools to power their offerings, and letting neighborhood retail brokers and originators expand at a scale that previously could not be achieved. You can see that change reflected in our numbers and in the traction we're generating alongside our enterprise partners. Such partnerships tend to be sizable and intricate, requiring longer sales and setup timelines than anything handled within our D2C operation, and cultivating them isn't work we can accomplish on our own. This takes close teamwork with our partners throughout the whole process — from getting integrated and brought on board, all the way through conversion, tuning, and growing the product. How quickly things ramp up is something we share, and we're side by side with every partner to help them scale. What we're achieving is genuine. The early numbers look quite encouraging, and we've never been more excited about what lies ahead. Let me walk you through what we're seeing across each of our key partnerships. As you know, we launched the largest platform collaboration ever undertaken by Better together with Intuit Credit Karma, a leading personal finance company serving more than 40 million monthly active users. In the past year by itself, Intuit Credit Karma handled 47 million tax filings and surpassed 140 million members. Actually, over 80% of Americans who obtained a mortgage during the last year belong to the Intuit Credit Karma platform. Via this collaboration, we're merging the extensive scope and richness of Credit Karma's member information—covering credit, income and property details like complete credit bureau reports, tax returns and thorough home valuations—straight into the Tinman AI platform. As you may recall from our public announcement, Credit Karma aims to deliver $1 trillion in interest savings for its members on their mortgages. This is a big job. Our joint offering has been cutting borrowers' lifetime interest costs by about $25,000 on average since we rolled it out in October 2025, and reaching Credit Karma's goal would require funding 40 million mortgages. After working together for over 9 months, we launched on the Credit Karma app in October 2025, and growth has come fast since then — so far we've reached fewer than 1% of their monthly users whom we believe are eligible for the product. The potential here is enormous, and our main priority is to deepen the integration of the Tinman AI platform throughout the different Credit Karma consumer touch points to more effectively address the complete needs of its whole member base; additionally, via our Tinman AI platform, we keep making strong strides in expanding our platform to support local retail mortgage lenders, equipping them with the infrastructure to establish and grow their operations atop our technology; we keep scaling NEO with its local loan officer teams throughout the United States seeing fast expansion. So what that means is retail mortgage lenders can set up shop on Tinman with basically zero cost to bring in customers through this channel. Watching the NEO team grow has been incredible — they came in at a $1.5 billion run rate and wrapped up 2025 at $2.4 billion on the Tinman AI platform. What Tinman AI has shown is that it takes friction out of the picture, giving originators room to grow the right way, and 2025 saw 28 new loan officer teams join the platform. Once fully rolled out, within 6 months, NEO saw funded loans per mortgage adviser climb 91%, per processor rise 17%, and per underwriter go up by nearly 50%. Retail mortgage teams nationwide are noticing these improvements and are departing their current platforms to come to the Better platform and to begin our shared mission of making retail home lending more affordable, quicker, simpler and simply better. Next, our top 5 U.S. A nonbank mortgage loan originator partner launched on the Tinman AI platform this February, with only 2% of its loan officers taking part. Over the next few months, our goal is to bring that up to all 3,000-plus loan officers. According to early feedback, the loan officer experience for Tinman users is superior to what they had with the previous implementation on their legacy software stack. We think that once this rollout reaches all of their loan officers, this partnership is going to be a game changer for both companies. It gives Better a big opportunity for platform volume, and it also hands one of the biggest mortgage originators in the country an advantage over competitors in how they take care of their customers. Beyond that, Finance of America, which is a top reverse mortgage lender and can reach millions of customers who are usually rich in home equity yet short on cash flow, is just starting to ramp up. Working together, we're introducing the first HELOC and HE loan products for their customers, all driven by our Tinman AI. Our collaboration with Finance of America should let us serve a demographic that historically hasn't been well served by the senior market, and we're optimistic about that. Over the next several quarters, we anticipate meaningful outcomes from this partnership.

Vishal Garg

executive
#3

And then, we shared some big news: we've rolled out the very first conversational engine for credit decisions covering mortgages and home equity loans, wired straight into ChatGPT via our Tinman AI app. Now, loan officers, banks, and fintechs can get credit outputs ready for decisions in just 47 seconds, which cuts origination timelines by 21 days on average. Better is the sole application permitted to show credit decisions inside ChatGPT, driven by our proprietary MCP technology layered on Tinman. Tinman is able to underwrite around 95% of mortgage and home equity loan types instantly, and any institution holding a ChatGPT enterprise license can roll it out; no traditional aggregators, no markups. That creates a major new distribution avenue and an obvious route to grow into a direct-to-consumer avenue eventually. As you may recall, OpenAI and ChatGPT have more than 800 million users worldwide and more than 80 million users in the United States, with that figure climbing quickly. We see this as the third version of the Internet, and we're first to market with a clearly differentiated offering compared to the other folks who've launched apps on OpenAI and ChatGPT, and with the capability not to offer a marketplace or provide a solution that then forces consumers to exit the platform, but instead to deliver a solution that lets consumers complete the whole transaction right inside their ChatGPT interface. Since our OpenAI announcement, we've witnessed an enormous instant reaction from across the financial services industry. Within days of putting out a short demonstration video last week, we've gotten inbound interest from more than 40 financial institutions, mortgage companies, banks, fintechs, all contacting us at the most senior levels to ask for a demo and collaborate with us on deploying our ChatGPT application. For instance, a bank CEO in the South got in touch after watching the announcement. They're looking to expand their mortgage business, but not through the approach they'd attempted previously—hiring big teams, constructing fixed infrastructure, and shouldering the operational burden that follows. What clicked for them was how simple the ChatGPT app is and the notion that any loan officer at any branch can immediately qualify a consumer for a mortgage via a conversational interface, with minimal setup time, minimal training, maximum reach. That's precisely the challenge we set out to address. The mortgage industry has been stuck for a long time in a cyclical model, ramping up headcount when markets are good and slashing it when they're bad, with fixed costs that hurt originators when volumes fall. Tinman completely transforms that dynamic. The infrastructure we've built and demonstrated with our existing partners can be rolled out for any bank, fintech, or local originator team. We're providing institutions the flexibility to expand their mortgage business without the operational burden that has historically made such growth so hard to maintain. We have two strategies when it comes to go-to-market on the Tinman AI platform. To begin, we aim to possess what lies ahead through alliances such as those we've established with Credit Karma and OpenAI, in which we're creating fresh methods to connect with tens of millions of buyers that are considerably simpler and quicker for them to utilize, while employing our technology to build a customer experience and value proposition barrier that nobody else in this sector is able to equal. Next comes advancing the path ahead, which is what we've accomplished alongside NEO, Finance of America, and the top 5 mortgage originator. Better serves as the vehicle through which these neighborhood market specialists and sizable established mortgage originators possessing strong ties can keep assisting both their clients and referral partners. Thanks to Better's collaboration, NEO is turning into one of the quickest-expanding retail lenders nationwide. The personnel stayed the same. The connections stayed the same; just the technology platform shifted. I'll now discuss our financial highlights, and Loveen will offer more specifics shortly. During Q4 of 2025, we produced $1.5 billion in funded loan volume plus $44 million in revenue, marking year-over-year gains of 56% in loan volume and 77% in revenue, respectively. Such expansion occurred across all 3 of our primary product categories: refinance, purchase and HELOC. Our Tinman AI platform produced $646 million in volume during Q4, accounting for more than 40% of overall volume and exceeding our earlier guidance of $600 million. This better-than-expected result mirrors partner demand and their rising trust in our platform. Though Q4 is consistently weaker due to seasonality, our year-over-year expansion beat that of the sector average, which stayed fairly flat. Based on MBA figures, during Q4, overall residential funded loan volume rose 4% year-on-year, whereas Better's funded loan volume climbed 56% across that same stretch. Across all of 2025, we achieved $4.7 billion in funded loan volume plus $165 million in revenue, rising 32% and 52% year-over-year, respectively; we secured this expansion even with a roughly $1 billion obstacle stemming from the end of our Ally partnership, demonstrating how resilient our model is. We stay on course to hit $1 billion in monthly volume by May 2026 and to achieve adjusted EBITDA breakeven by the close of Q3 2026. To succeed in a commoditized marketplace, you must prevail on 3 fronts: customer acquisition cost, operational cost and cost of capital, which we refer to as the 3 pillars of competitive advantage. Regarding customer acquisition, our approach flips the conventional origination model. Instead of purchasing customers on an open market, our partnerships are arranged so that customers come straight to us. Credit Karma's more than 140 million members, NEO's 70 local branches and 140 mortgage advisers, along with our top 5 nonbank originator working with over 3,000 local mortgage advisers, amount to embedded distribution at scale, a structural CAC edge that rivals find incredibly hard to copy and one that's difficult to maintain absent a technological moat. When it comes to what we spend to operate, Tinman takes care of as much as 80% of the repetitive work involved in producing loans, and our Betsy tool settles underwriting problems right away — it gathers loan facts and guidelines and puts together communications within seconds. What that gives us is a platform capable of scaling production by means of AI efficiency and growth, with no added overhead. What it costs us to process a loan, underwrite it and close it — and here I mean mortgage loans and HELOCs counted together — comes to roughly $800 per loan, which is way below what anyone else in the industry pays. We think that with the initial launch, our Home Token is going to let us book an additional $500 in revenue for every funded loan. And as we scale that up, we believe that over the long term, we're going to be able to realize meaningful gains in loan revenue, plus in funding cost to the consumer and interest rate to the consumer, and we believe that will turn into a meaningful competitive advantage and moat thanks to the efforts we have invested. Regarding cost of capital, we keep enhancing our warehouse terms while we work on growing capacity so we can support the growth of partnership volume.

Vishal Garg

executive
#4

At the same time, we're pursuing a tokenized credit facility that's secured, through a stablecoin ecosystem, which we believe could bring funding costs down by as much as 100 basis points after it's in place — a structural funding edge that any traditional mortgage originator would find hard to replicate. The groundwork for this moment has been laid by us over the last 3 years. Honestly, the most recent time I felt this level of excitement about where Better is headed was March 2021. Once again, we can see a path to becoming the biggest mortgage company in America. This really is a turnaround we've spent years bringing to life, one made possible by rolling out AI throughout our whole business and by using the Tinman platform, which we began developing back in 2014. That's why we believe our moat is more durable than that of a traditional AI native firm going up against a traditional incumbent. We've created an end-to-end system that unifies 8 different systems across the mortgage industry into a single one, so what gets captured isn't only the rules but also all the context surrounding human decisions about the data and the rules; and learning from data across $110 billion of loans is what lets us keep pushing ahead, reducing our cost to produce, raising our conversion rate and building for our partners who are building the future; and we think we can keep doing this because the competitive advantage that comes from richer learning data only compounds as time goes on, with more transactions and more partners joining the ecosystem. With momentum, scale and a clear path to adjusted EBITDA breakeven, we're now solidly in our next phase of growth. Partnerships are growing, adoption is climbing, our platform is proven and our AI capabilities are best-in-class, and we're only getting started; with that, I'll hand things over to Loveen for a detailed walk-through of our financials.

Loveen Advani

executive
#5

Thank you, Vishal. It's a privilege to be here at Better during such a defining moment. Thanks to its differentiated platform, the company stands out as a leader in AI-powered home finance. I'm eager to work alongside Vishal and the team to deliver disciplined execution, strengthen financial performance, and build shareholder value. As Vishal described, we're undergoing a substantial strategic transformation — moving away from being a direct-to-consumer originator toward becoming an AI-native platform that powers the wider mortgage ecosystem. Financially speaking, this shift matters a great deal. Enterprise partnerships at this scale take longer to ramp, yet they offer much larger volume potential and substantially better marginal economics than what our legacy D2C model delivered. The encouraging part is that the shift is already visible in our numbers. Partner-driven business through our platform keeps growing at a fast clip, and it now represents an ever-larger share of what we do overall. Take a concrete illustration: across 2024, we financed $3.6 billion in total volume, and none of that came from Tinman's AI platform partnerships; by 2025, funded loan volume hit $4.7 billion, of which 35% came through our Tinman AI platform. As for 2026, we can see a defined path toward having over 60% of loan volume come from the Tinman AI platform side of the business. That represents a genuine overhaul of how our revenue is composed, and it demonstrates how this transition is being executed. Let me now take you through how we did financially in Q4 and across all of 2025. No matter what happens with the wider economy or the mortgage market, Better continues to create opportunities. We have a big market we can go after and hold under 1% of it right now, so we've shown our ability to expand whatever the macro backdrop happens to be. Starting with Q4 of 2025 compared with Q4 of 2024: funded loan volume increased 56%, reaching about $1.5 billion. Revenue was up 77%, hitting approximately $44 million. Most of that expansion came from getting more loans funded via partnerships on our Tinman AI platform. If you break loan volume down by product, refinance climbed to 8%, purchase was up 22%, and home equity gained 18%. As for channel, Tinman AI platform partners accounted for 44%, while direct-to-consumer made up 56%. Looking at the breakdown by product, purchase accounted for 49%, refinance made up 37%, and home equity represented 14%. When you compare the full year 2025 against full year 2024, funded loan volume rose 32%, reaching roughly $4.7 billion. Revenue climbed 52%, hitting approximately $165 million. The launch of our Tinman AI partnerships, along with ongoing expansion in our direct-to-consumer business, powered these results. On a product basis, refinance was up 119%, home equity gained 78%, and purchase improved 14%. In terms of channel, Tinman AI platform partners delivered 36%, direct-to-consumer contributed 62%, and the remaining 2% came from our former Ally partnership. Breaking it down by product mix, purchase represented 61%, refinance was 21%, and home equity accounted for 18%. Now let's discuss cost efficiency. During Q4, net revenue overall climbed 77% compared to the same period last year, and our expenses stayed more or less level. That shows obvious operating leverage — we're growing top-line dollars while marginal costs drop, thanks to efficiencies coming out of the Tinman AI platform. We continue to pare down our cost structure, yet we still ensure sufficient resources are allocated toward new partnerships. Those partnerships ought to fuel meaningful growth throughout 2026 and beyond. Unit economics in our direct-to-consumer business continue to improve. AI is now embedded across every facet of how we market and run things. Per-loan contribution margin increased 28% from the previous quarter, moving from about $1,800 to about $2,300 for each loan. We expect to keep lowering the expense of originating loans by converting more of them, spending less to bring in customers, and getting more out of our workforce. For the fourth quarter, adjusted EBITDA loss was roughly $24 million, which compares with a loss of $28 million in last year's Q4 and a loss of $25 million in the quarter right before this one. Even though we aim to keep reducing those losses each quarter, we are always balancing cost discipline against funding growth bets. Because the business continues to grow while marginal economics stay favorable, we get to adjusted EBITDA breakeven sooner; in our view, we have reached an important inflection point, moving away from being primarily a direct-to-consumer fintech and toward becoming a true AI platform for the mortgage industry. That underpins our confidence that we will reach adjusted EBITDA breakeven by the close of Q3 2026. Now a quick update on our balance sheet and capital position; we finished Q4 2025 with $227 million across cash, restricted cash, short-term investments, and assets held for sale. Our ties with financing counterparties remain solid, with three warehouse facilities providing $575 million of capacity as of December 31, 2025. We're grateful for our warehouse lenders' ongoing backing as we roll out Tinman AI throughout the mortgage ecosystem. Moving to our outlook; for total loan volume, we're projecting $1.4 billion to $1.55 billion in Q1 2026, with the midpoint representing 70% year-over-year growth from Q1 '25. Considering the rate at which our partners are expanding, we remain confident about reaching $1 billion in total monthly loan volume by May 2026. By the close of Q3 2026, we expect to achieve adjusted EBITDA breakeven. Several factors will fuel this: rising volume through both our Tinman AI platform and direct-to-consumer channels, stronger contribution margin on each loan, enhanced pricing, and reductions in corporate expenses. One thing worth noting is that these avenues for growth each have their own timelines for scaling, meaning the path to breakeven could see some unevenness. With that said, I'll turn things over to the operator for Q&A.

Ramsey El-Assal

analyst
#6

I'd like to ask about guidance. At the midpoint, your guidance implies Q1 loan volume is approximately unchanged compared with Q4, as I understand it. Clearly, there's a great deal of exciting activity happening at the company. I'm just curious whether you could take us through what's driving that. Partnership volume has been growing well on a quarter-to-date basis relative to Q4. Does that suggest you anticipate more modest growth on the direct side? Or which drivers ought we to be thinking about?

Loveen Advani

executive
#7

Eric, this is Loveen. I appreciate the question. The reason it's flat comes down to seasonality. If you turn to Page 17 in our investor deck, that's a point we addressed, and we've laid out the previous 6 quarters; so looking from Q4 '24 through Q1 '25, the number declined, correct? And in the current year, going from Q4 '25 to the Q1 '26 figure we're guiding to, it's level or marginally higher. That really illustrates the sort of expansion occurring in the platform.

Ramsey El-Assal

analyst
#8

Understood. All right, and I've got a quick second question. I'd like to ask about profitability. Clearly, your present objective is hitting adjusted EBITDA profitability before Q3 of this year wraps up. In what way should we consider — how are your views shifting regarding profitability over the medium term and the longer term, particularly given this speeding up of the move toward the partnership model? In what way should we think about your profit profile down the road?

Loveen Advani

executive
#9

Yeah, definitely. Look, I only came on board a month ago. Getting to profitability by Q3 2026 is the initial job, correct? Once that's done, we'll assess our growth opportunities together with incremental positive contribution margin, correct? So whenever we assess new partnerships, we'll be considering a contribution margin spanning 10% to 15% up to as much as 25% to 30%. And we'll sort of be examining that span as we sort of consider our growth opportunities.

Vishal Garg

executive
#10

I'd say the product breaks down into three separate categories. The first category covers what we handle on D2C. Then the second category is everything we do on the Tinman AI platform, where the loans get closed under our own name. That's the approach we take with NEO. That's the approach we took with Credit Karma. That's the approach we take with others as well. And then the third one concerns what the margin looks like on business where the lender closes under their own name; that's the approach we take with Finance of America; that's the approach we take with the top 5 mortgage lender, with the top 3 fintech. Across all of those, those lenders close under their own name, and we hand them the platform to make that happen. Their salespeople are processors, underwriters and the closers working inside our software. And so every one of those carries a distinct margin profile along with a distinct revenue per loan profile, right, based on how much work we're putting in there. Take D2C, obviously, we handle everything from customer acquisition to sales to processing, underwriting, closing and investor marketplace. In — but we handle everything else. And then with the pure stuff like processing, underwriting, closing and capital markets, sometimes we handle cap markets, sometimes we don't. And so it simply comes down to that, what revenue per loan will end up being and what the margin will end up being. As revenue per loan sort of declines, the margin actually widens since it increasingly becomes a situation where they're simply using the platform. So the platform alone business can reach 60% margin. The D2C business, as you can tell from a contribution margin standpoint, is a business with 20% plus margin measured on a contribution margin basis; so we're going to get to know that and define that. We feel quite confident about the guidance we're providing and especially the growth we're showing. But I think at growth rates of those kinds, you can't precisely know what people will purchase. And we're in the transformation phase of the business. So we'll know more across the coming couple of quarters.

Kartik Mehta

analyst
#11

Vishal, those partnership numbers point to a huge amount of demand right at the top of the funnel. What I'd like to know is which metrics you're watching as someone moves from preapproval all the way to a loan that actually gets funded, and how does that support hitting the $1 billion goal?

Vishal Garg

executive
#12

Sure. Kartik, when you look at it through the lens of the D2C business we've shared numbers on before, that figure comes out to roughly 5%. So imagine partner volume arrives on a cohort basis — say I receive $1 billion in pre-approvals. Of those, I wind up funding somewhere near 5%. That funding might occur across 3, 4, 5, 6 months as the cohort matures, since certain people aren't happy with the precise offer, aren't quite ready yet, come back later, or need their spouse on board. All of these dynamics show up in what is a pretty meaningful life event financing deal for consumers. Keep in mind that, on average, 32% of their income flows to us. That makes it a big transaction. A lot of back-and-forth happens between the moment we approve someone and the moment we can genuinely record that funding event; still, on a cohort basis, it matures to 5%. With certain partners it lands higher, since those partners carry stronger brand, tighter matching, or more deeply integrated systems. With other partners it lands somewhat lower; we'll watch how that unfolds.

Kartik Mehta

analyst
#13

And Vishal, how far along are you when it comes to using the stablecoin ecosystem for funding? Clearly, you mentioned that — bringing down funding costs — and it sounds quite compelling; so I'm curious, what stage are you at with that?

Vishal Garg

executive
#14

My sense is that we are 6 months out from the point where it begins to show up in the bottom line.

Brendan McCarthy

analyst
#15

Hello, Loveen. I'd like to begin with the Credit Karma partnership. As things stand, does it cover every one of your mortgage offerings? Or is it only focused on refi?

Vishal Garg

executive
#16

At this moment, we've kicked things off with refi, and our expectation is that we'll next roll out HELOC, and after that, purchase.

Brendan McCarthy

analyst
#17

Got it. That makes sense, and considering the addressable market of $140 million, clearly that's far bigger than what was initially anticipated. Perhaps, just thinking longer term, what penetration rate would you view as reasonable for driving volume?

Vishal Garg

executive
#18

Over the long term, our expectation is that Credit Karma Home Loans, powered by Better, will become the biggest mortgage originator nationwide in the United States.

Brendan McCarthy

analyst
#19

Understood, and that's excellent. Moving on to what you expect for breakeven adjusted EBITDA by the close of Q3. My assumption is that this will roughly line up with the $1 billion in monthly funded loan volume. Could you lay out your expectations there regarding the volume contribution from D2C, NEO, and then Credit Karma too?

Loveen Advani

executive
#20

Yes. As I mentioned in my script, the Tinman AI platform's contribution stood at 0% in 2024. In 2025 it was roughly 35%, and we anticipate around 60% of total volume coming from that platform, which covers Credit Karma, NEO, and other partnerships.

Brendan McCarthy

analyst
#21

Got it. Now, regarding what we saw in Q4, if we examine the gain on sale margin, my sense is it went down a little compared with the prior quarter. I'm guessing the main reason was simply more refinance D2C volume?

Vishal Garg

executive
#22

That's right.

Brendan McCarthy

analyst
#23

Then my final question. In the slide deck, I noticed what appears to be a top three personal lending fintech in the pipeline. You said, I believe, that it's in the pilot phase right now. Could you share any details? Will it lean more toward the Tinman mortgage software partnership angle? Or do you expect it to resemble NEO or Credit Karma, where you handle the originating?

Vishal Garg

executive
#24

Our view is that at the start, it will resemble Credit Karma, with us handling the originating. Beyond that, this fintech has a fairly prominent bank attached, so they might decide to onboard onto that bank's balance sheet. I believe I've stated this openly before: bank capital regulation requirements are set to dramatically reshape the mortgage landscape. Regarding the volume of calls we've received from banks since the ChatGPT app launched, we now have more than 45 financial institutions across the United States and beyond its borders expressing interest in using that platform. My expectation had been mortgage brokers. My expectation had been retail mortgage lenders — yet look at how many banks have reached out, driven by anticipation of what's coming. Let me go deeper on this. Suppose you're a midsized bank right now, and stablecoins are disintermediating you, correct? Grabbing Internet deposits on the cheap is no longer possible, right? You have to go — your deposit cost of capital keeps climbing. You have to go locate assets; and once you need to locate assets yielding more, parking money in treasuries and short duration instruments is no longer an option, because if that's your approach, your cost structure simply won't let you compete against stablecoin. So what can you do, with economic growth somewhat <<inaudible>> and the American consumer a bit stretched — will you go long credit cards? Will you go long personal loans? Will you go long those assets people have pursued over the past 5 years? Or, now that mortgage reg reform and especially bank capital levels have shifted, will you go long credit risk? Or will you go long duration? Banks are built for going long duration; as a result, we're going to see the bank bid for mortgage explode. The bank bid for HELOCs explode. And compared with our rivals in HELOC land, we are uniquely positioned to accommodate that bank bid. Our rivals in HELOC land have built a one-size fits all offering, and they take pride in it like a box for securitization. To any bank we say: bring your guidelines, bring your regional preferences, bring any of those, and we will accommodate them instantly, to whatever level of detail you want. So I believe that's something you'll witness quite a bit of, plus additional alliances along those same lines. Another development taking place is that every one of these fintech companies is applying for bank charters, and they're all recognizing it as well. Consequently, I expect the distinction between fintech and fintech bank will become kind of fuzzy. Still, I figured it could be useful to offer some background on this, especially given what was announced today regarding rules for bank capital.

Eric Hagen

analyst
#25

This was a genuinely useful discussion. I really valued what you shared a moment ago regarding bank capital. According to you, underwriting costs sit well below what the industry typically sees. If I caught it right, that figure is roughly $800. So my question is, why do you believe those savings aren't making their way to borrowers? What's the limiting factor — the thing that, in your view, essentially chokes off the ability to hand those savings over? And after that...

Vishal Garg

executive
#26

On our end, the savings do get handed to borrowers. That's right — we hand borrowers those savings on our side, and at the same time we keep working to lift our contribution margin along the road to profitability, since our goal is to keep handing those savings to borrowers for a great many years ahead; so I'd say our rates come in 30 basis points below the typical mortgage rate. Compared with Rocket and loanDepot, our rates are more than 50 basis points lower. And in a purchase market, I'd say the customer really takes their cue from the local realtor and the local LO. So my sense is that — you haven't witnessed that yet. But once refi returns, Eric, you'll recall that between 2016 and 2021, our volume climbed from $500 million to $58 billion during the refi period — roughly the rate environment of 2019, 2020, 2021 — our volume went from $4.5 billion up to $58 billion; and once rates decline and refi returns, I believe, there's a real possibility for serious scale, because when it comes to refi, our proposition clearly wins. It may be hard for the American consumer to tell the difference between 5 and 5/8, and 5 and 7/8 — stop someone on the street and ask, "Hey, what's 7 divided by 8." That kind of math isn't taught in American schools these days. But say to them, "Hey, would you rather save $422 a month instead of $375 a month?" That math is simple to do. So I think that's roughly where you'll see those savings truly show up for the consumer; on the B2B side, meanwhile, the savings are immediate and plain to see. And a typical bank's cost to produce comes to about $14,500. So when we approach these banks and tell them, we can handle this for you — suppose you'd like option A at $4,000 per loan, option B at $5,000 per loan, option C at $6,000 per loan — that's an extremely disruptive kind of sale. It does take time, though, since it's a CEO and CFO level sale. If you consider AI deployment anywhere, pitching that to someone mid-level at the bank, or pitching it to the head of origination, or the head of sales for mortgages at a bank, that's quite disruptive, because what Tinman does and how Tinman reaches that point is genuinely cutting down the volume of RoTE work and RoTE calculation work that exists in mortgage today. And 99% of the mortgage industry still relies on stare and compare underwriting. Phone any of my competitors, ask the loan officers how those loans get underwritten, and they'll tell you. And I believe that's simply a fact.

Rohit Kulkarni

analyst
#27

A few questions have come in regarding the Tinman AI platform. Could you help us understand how the ramp is shaping up based on what you know right now — 40% of volume already sits in Q4. Where do you think that share will end up as the year progresses? And with so many recent developments, what are the gating factors that would let you scale distribution for Tinman? Would it be technical integration, regulatory compliance approvals, or partner training? Please walk us through what converting all your Tinman leads would require, and then how that flows into the overall proportion of funded volume.

Loveen Advani

executive
#28

Rohit, I appreciate that. The first part's mine to answer, then Vishal will handle the second. Here's the picture: back in 2024, the Tinman AI platform contributed nothing — 0% of revenue. When we look at 2025 across all four quarters, that figure sits near 35%. For the current year, our expectation is roughly 60%. Now, we aren't providing loan volume guidance for the entire year, correct? Still, if you connect the dots and follow where things are heading, the number we've given for Q1 loan volumes works out to approximately 77% growth. If you take that same pattern and project it forward — and to be clear, I'm not offering guidance — plus our piece of the Tinman AI platform rising from 35% up to roughly 60%, it becomes clear that this segment is expanding very quickly.

Vishal Garg

executive
#29

Okay, so look at our big institutional partnerships — the ones where those companies are 10x to 100x bigger than we are. Getting from the first demo to a signed term sheet typically takes 3 months. Then from a signed term sheet to launching the platform, that's usually another 2 months. After the platform goes live, finishing a pilot runs about 90 days from there. Then once that pilot wraps up, achieving complete institutional buy-in and reaching their whole customer base takes somewhere around 9 to 12 months, simply because we're stripping out so much cost. On top of that, this is the most complex financial product that gets sold to consumers. So there are a ton of wires that need connecting. The nice part is once everything's hooked up, it's only one. It's a single system. And now, thanks to what we've built with ChatGPT, we're genuinely working to shorten both that sales cycle and the connective cycle, since it's an interface their own internal folks are already familiar with. That significantly reduces what used to be roughly a 9-month timeline — from first demo all the way to full implementation — probably bringing it to 6 months, or even 3 months if they're eager to move quickly and carry no legacy systems. That's exactly why you're watching plenty of players who aren't in the mortgage business come into mortgage through us. For those already in mortgage, with all that enormous incumbent infrastructure they're carrying, it takes them longer. And the bottleneck is this: our Biz Dev team had two people as of last quarter, and now there are about five on that team. We simply want to ensure revenue lines up with cost, so we don't run off and bring in 100 go-to-market salespeople, get them out in the field, and then watch the revenues fail to materialize. We recognize we've got to push this business to profitability, and that we hold something a full generation ahead of the incumbents, two generations ahead of the tech stack sitting at the banks. Meanwhile, the nation's largest bank is right in the middle of migrating to incumbents. Okay, right? That system where one person can only use it once at a time. So they're essentially in the middle of moving to SharePoint. So I'd say we have — we have a lead and — but our aim ought to be to monetize that lead, while doing it in a manner that keeps expenses and revenue aligned with each other.

Rohit Kulkarni

analyst
#30

Alright, that's helpful. And when it comes to Credit Karma in particular, could you discuss the ways Credit Karma helps magnify the advantages and maybe boosts how visible the distribution is across their membership? Is there some kind of two-way arrangement — if you become tightly woven into a fintech partner such as Credit Karma, in what way does that shift how they market or give greater prominence to what you provide?

Vishal Garg

executive
#31

In my view, Credit Karma is a highly sophisticated organization. They've got — once more, should you look at anything they've published publicly, they've got a platform named Lightbox. And into Lightbox we have woven ourselves. And at this point that platform is deciding which offers get shown. As of March 13, we currently sit below 1% penetration among their membership; therefore the future has us quite enthusiastic.

Rohit Kulkarni

analyst
#32

All right. Great. So perhaps one final question from my side — and maybe you've already addressed this — is about the contribution margin, or the marginal margin, on D2C compared with what you get per dollar earned through partnerships. How do those two stack up at present? And as time goes on, where do you expect that to go? And is that, in a sense, an assumption baked into your EBITDA breakeven for the second half?

Vishal Garg

executive
#33

I'd say we're not — since our partnership volume comes in unevenly, I think we are — and for reasons related to competition, we aren't disclosing that degree of fine-grained detail at this point. That said, you're right that contribution profit for each loan isn't a fixed figure, just like partner profit — as I mentioned before, it shifts according to which system resources and which personnel resources get used, and how much of each. Still, our adjusted EBITDA breakeven does depend on hitting those penetration rates with the partners we've already signed.

Ryan Tomasello

analyst
#34

One more thing about the Tinman AI platform. Clearly, many different models exist in the market that also offer this wide tech infrastructure supporting origination and funding within mortgages. Some of those competitors are building on blockchain rails. Vishal, you brought up certain legacy LOS providers and POS incumbents. So broadly, could you discuss what you believe sets this third-party infrastructure category apart from those peers? And later on, do you think this platform might be extensible into other consumer credit categories beyond mortgage and HELOC?

Vishal Garg

executive
#35

Sure. As far as I know, Tinman stands alone as a platform where loans can be offloaded to a wide-ranging group of investors, each bringing their own criteria and rate structures to bear. With Figure, my understanding is they built something that allows integration, yet the terms attached to that offering are fixed by the offering itself. Their first lien option doesn't qualify for Fannie, Freddie, FHA, or VA backing—it's structured as a first lien HELOC—and its pricing runs considerably above what a conforming mortgage would cost. It also depends on the very HELOC framework that Figure already has in place. That framework is, without question, extremely well built, but a lot of its pieces are proprietary and simply cannot be stripped out. You're free to pick any title company anywhere in the United States. Any appraisal firm in the country is fine with me. Same goes for whichever home insurance provider you prefer. Say you belong to an HOA, or the appraisal is complicated, or you're looking at jumbos — maybe non-QM is what you want, or bank statement loans, or DSCR loans. Whatever combination of those you're after, if the goal is reaching the deepest possible penetration across your customer base, it's sort of necessary — keep in mind, when you come on as a partner and sign up, think of me as bank A with 100 customers. Am I looking to — right now, my main reason for offering mortgages is really just to help out customers, you know? Like, I want to take care of someone who's got money deposited with me, correct? So should I team up with someone whose standards are designed around securitization, focused on a specific set of loans, and who only approves about 15% of applicants? Or should I be helping the person where we'll accept FHA loans down to a 580 FICO score for lower-income borrowers, since that individual still keeps a deposit account with the bank, and the bank wants to take care of them? To me, that's really the key distinction between what we do. The way we structured our approach was around AI-driven mortgages and AI-driven HELOCs, while theirs was organized around securitizing HELOCs first and then securitizing mortgages. That really reflects a fundamental difference in how each business is designed. I'd put it this way: they are genuinely anchored in blockchain, absolutely, and all the things that go along with it. We, on the other hand, are truly focused on artificial intelligence and personalization — delivering personalization at scale to the broadest and biggest set of potential partners, all enabled by AI. Beyond that, we also apply blockchain in places where it makes sense, specifically to lower our cost of capital. So that's sort of — but I have tremendous respect for their team, and they've accomplished something remarkable, right, in creating an excellent platform and genuinely bringing the home equity offering back to the American consumer. So we're quite pleased to trail their example when it comes to the home equity offering and to keep being the frontrunner in mortgage innovation. As for the remaining competitors, I mean, that's simply extremely outdated technology infrastructure, correct? A lot of them are still completely charging per seat. Our charges are based on the outcome. Others charge based on their, like they resemble, we charge based on the outcome. It's simply a fundamentally disruptive approach. Now a few of them are joining forces and saying, "Hey, sure, you can purchase, like, the three of us as a package deal." But that resembles selling Microsoft Office Word alongside Windows 95 and, like, offering them as one thing, correct? But, like, you recall what happened with copy paste during your Windows 95 era, correct? It isn't the same; it doesn't refresh in real time. It isn't a smooth workflow. It isn't any of those qualities. The customer experience is fractured. And even more crucially, you still require all the staff, which explains why, if you consider the whole idea of digital mortgage, the mortgage sector, and if you speak with any CEO in mortgage, they say digital mortgage, it's 2015, it previously cost me $9,000 per loan to originate a loan. And now it's 2025, it costs me $11,700 to originate a loan, like I've moved backward since 2015, correct, as a mortgage company CEO due to digital mortgage, because it's 8 separate digital systems, 8 separate, like, middleware components, 8 separate sets of consultants I must bring on and keep on payroll, all the staff I must educate to become specialists in these 8 separate systems who can't, like, perform various functions across various systems. The entire — that's the disruptive strength of the AI Agentic architecture. You don't have to educate staff to perform this. A machine simply handles it. And on our machine at Tinman, the staff have been performing this work. And so we simply — when we wish to shift a role to Agentic, we literally just have the machine and the AI observe what the humans engaged in that specific task have been doing; and I believe there remain, like, certain tasks that will require, from a regulatory perspective, the necessity for someone to render the decision, a human to render the decision. And that's perfectly okay, since a person could then handle a hundred such choices each day instead of just a couple daily. So that, I believe, is the future we're genuinely working toward. And in that respect, I'd say we're fairly distinctive.

Ryan Tomasello

analyst
#36

Ryan, thanks for those kind words about the commentary. Let me take the Sky stablecoin partnership question. Could you describe, or perhaps put a figure on, how much cheaper the capital from that funding source is compared with what we get through our usual facilities? Beyond that, how might this collaboration grow past the warehouse stage toward something more like permanent financing? And stepping back, Vishal, what sort of value do you think DeFi could eventually unlock for the mortgage industry?

Vishal Garg

executive
#37

Alright. Wow — I could talk about this for hours, but let me try to keep it very straightforward. First, I'd say the upfront cost of funds benefit comes to 100 basis points, and that's really significant, correct? Simply from the start, we earn an additional $500 on each loan, right — per loan. Second, looking ahead, we believe mortgage is genuinely an underpenetrated asset class when it comes to stablecoin issuers. As stablecoins grow more widespread, I expect issuers seeking broader yield will search out DeFi mortgage assets to put money into; and we feel the loans we originate — 95% of them backed by some type of GSE or agency guarantee — are optimal on a Sharpe ratio basis when you weigh yield pickup against risk. I like to joke that in a technical sense a Fannie Mae mortgage beats a treasury, since you get the government guarantee plus an actual house and an actual person. That gives you three forms of collateral; so I'd say the spread premium attached to prepayment risk isn't something institutional investors worry about, yet when it gets passed along to consumers, it is something consumers do worry about. So I find that really, really fascinating. And I believe the lasting benefit DeFi delivers to the American consumer mortgage market amounts to a 100 basis point drop in rates. At present, the premium for holding a fixed rate GSE mortgage rather than a 10-year treasury sits at roughly 200 basis points. I think over time we can bring that down to approximately 100 basis points.

Owen Rickert

analyst
#38

Starting off for me, to move from $1.5 billion of volume up to $3 billion each quarter, what has to take place?

Vishal Garg

executive
#39

We have to go deeper with the partners we already work with, and we have to keep expanding NEO along with D2C, in the places where that makes sense, where those D2C loans earn us money. Yet getting from $1.5 billion to $3 billion simply means going deeper with the partners we've already signed and brought on board.

Owen Rickert

analyst
#40

All right, that's helpful. Now regarding those 4 partnerships that are ramping up, could you put them in order by how big the opportunity is? Obviously Credit Karma sits at number one, but where would you place FOA, the top 5 nonbank originator, and that bank partner?

Vishal Garg

executive
#41

My ranking would be Credit Karma Home Loans powered by Better first, then the top 5 nonbank originator, and after that FOA along with the top three leading fintech.

Owen Rickert

analyst
#42

All right. Then my final question, building somewhat on that point — apart from those 4 partners, do you folks actually possess the capacity to bring possible partners 5, 6, and 7 online during 2026? Or does 2026 mostly concern scaling up those 4?

Vishal Garg

executive
#43

No, my expectation is that you'll witness a single marquee partner go live roughly each quarter, plus you'll see a number of smaller partners go live each quarter.

Vishal Garg

executive
#44

We appreciate all of you being here. To repeat, Q4 '25 represents a quarter of major change and recovery for our company, since we are shifting away from operating as a direct-to-consumer originator at Better.com toward functioning as a platform that enables every originator across the mortgage sector. We're grateful for your attention, and we appreciate your involvement and collaboration as we work toward achieving that goal — and by doing so, bringing down the cost, increasing the speed, and simplifying home finance, while delivering an improved experience for every American. Thanks.

Operator

operator
#45

That brings us to the end of today's conference call, everyone. We appreciate you taking part, and at this point you're free to hang up.

Read the full transcript via the API

You're viewing the first half of this call. Get the complete Better Home & Finance Holding Company transcript — plus 132,000+ transcripts from 7,000+ companies, speaker segments, AI summaries and full-text search — through the EarningsCalls.dev API.

Get the API View API docs →

This call discussed

For developers and AI pipelines

Programmatic access to Better Home & Finance Holding Company earnings transcripts and 132,000+ others is available through the EarningsCalls.dev REST API. Plans from $39.99/month — full transcripts, speaker segments, full-text search, and the recently-added /api/v1/transcripts/recent polling endpoint for ETL pipelines.

More tools for stock research

OptionsBell: an unusual options activity scanner that tracks options flow across all US stocks and emails you when large trades appear in Better Home & Finance Holding Company or any ticker on your watchlist.

StockMarketScan: a stock screener with alerts: 24 pre-built screens refreshed every 5 minutes, watchlists and email notifications. Both offer API and MCP access and fair monthly pricing.