Better Home & Finance Holding Co. Q3 FY2024 earnings call

Key Achievements

• Funded loan volume increased to $1.035 billion, beating Q3 guidance, with 42% year-over-year and 8% quarter-over-quarter growth.
• Revenue was $29 million, with excluding non-recurring benefits, revenue up ~8% quarter-over-quarter.
• Year-over-year gain on sale margin improved from 1.58% in Q3 2023 to 2.08% in Q3 2024.

Strategic Priorities

• Growth: Continued progress with funded loan volume growth across product categories, including 13% purchase, 493% home equity, and 177% refinance year-over-year. Launched streamlined refinance product for FHA and VA borrowers.
• Operational Efficiency: Increased expenses due to growth initiatives like marketing and loan origination, but offset by lower general and administrative and corporate compensation expenses. Utilizing Tinman, reduced mortgage process cost by over 35% of industry average. Launched Betsy, the voice-based AI loan assistant.
• Distribution Channels: Partnered with NEO Home Loans to expand into the distributed retail channel, combining NEO's customer service excellence with Better's AI-driven Tinman technology.

Financials

• Generated funded loan volume of $1.035 billion, revenue of ~$29 million, adjusted EBITDA loss of ~$39 million, and GAAP net loss of ~$54 million.

In the third quarter, funded loan volume was $1.035 billion, representing a 42% year-over-year increase and an 8% quarter-over-quarter increase. Revenue was $29 million. Adjusted EBITDA loss was approximately $39 million, and total GAAP net loss was approximately $54 million. Funded loan volume was 75% generated through the D2C channel and 25% through the B2B partner channel. It was 71% purchase, 16% HELOC (including home equity lines of credit and closed end second lien loans), and the remainder was refinanced.

• Fourth quarter funded loan volume expected to be in-line with third quarter volume.
• Focused on driving operating leverage through efficiency investments, corporate cost management, and distribution channel diversification, targeting medium-term profitability.

Q: Hey, guys. Thanks very much for taking the question. Vishal, I wanted to ask about the launch of your recent AI loan assistant Betsy. And I mean it sounds like it certainly has potential, the demonstration is very impressive. But in your view, how is Betsy differentiated from other AI technology in the lending space?

A: Totally. Thanks. That's a great question. I think Betsy is very advanced relative to some of the things that we've seen in the AI stuff space that we've seen with mortgage companies. A lot of the AI that mortgage companies that are our public peers have done is really related to a lot of the back office part of the process, processing or underwriting where there's been OCR, which is effectively reading data off a page. So looking at income data off a page stub and coming up with what's the closest approximation to what the income might be. And we've had that in our Tinman system for years. A lot of the other AI that people talk about is decisioning based AI, where effectively the machine is taking the decision rather than a human taking the decision to move the loan file forward. Again, we've had that for years inside Tinman. We've built that one day mortgage where the bulk of the process of getting the mortgage from a lock to a commitment letter is automated. And then the last bit that we've seen in other mortgage companies too is chatbots, so doing chat based customer support online. And if you've ever actually used those chatbots, you'll see there's a huge lag and the chatbots are not actually able to move the customer along other than basically being an interactive FAQ or frequently asked questions list. What we've done with Betsy is actively integrate Betsy into our platform, Tinman. And as you could tell from the call, Betsy was able to, one, understand what the items were missing from the consumer's loan file, was able to take that data in, and then append it into Tinman itself and move the process along and determine that the next step would be a rate quote, which would require a human based loan officer and then effectively transfer that to a human based loan officer. So you're talking about fundamentally replacing the work that a traditional loan assistant or SDA, sales development assistant, would have done it better here. Now to give you a context, back in 2021 when we were doing $58 billion of volume, we had over 1,500 SDAs and loan assistants. And I would say that the total spend on them was over $100 million a year. And Betsy is basically able to functionally do the task of these loan assistants and allow us to scale up. Betsy is now taking in 100% of our inbound phone calls that our loan officers can focus on helping consumers when they're ready to actually lock a rate, have a discussion across products or things like that that are licensable activities. So we think that the technology we've used on Betsy, the way we've developed Tinman, which really is fundamentally a decisioning engine on a tree based data structure inside what is called a JSON blob, which is something that is really easy for LLMs to understand and then take through at a speed that is really unmatched. We haven't seen anyone be able to bring the speed of response time, which, as you saw, was almost natural, about 800 milliseconds relative to the 3 second kind of response time that you see in chatbots or voicebots out in the industry. Nobody's been really able to engineer that in the way that we have, and that's why we felt comfortable launching it commercially and now have scaled it to 100% of inbound calls being actively taken by Betsy. And so I think that we've done a lot of work on this and we're just getting started. Today, Betsy is the worst that it will ever be. It gets better week on week and we're very, very excited about what we're just going to be able to do for our cost structure going forward.

Q: Hi. This is [indiscernible] on for Rayna Kumar. Thank you for letting us ask questions. So firstly, can you just discuss your outlook for 2025 and any initial views on the operating environment you expect for next year?

A: Sure, sure. So it's Kevin. I'll start. Vishal may want to supplement. Look, I mean, I think we've kind of all watched the market together, right? We sit around [4.40] on the 10 year. We just posted rates this morning, like we do every morning. We're just north of 7%. So I think, and the MBA and others have kind of softened their forecast versus for 25 versus where they may have been a couple of months ago. So I think we expect continued improvement in the market environment. I mean, our refinance volume was up off a small base, albeit, but up massively in Q3. We've seen that others were as well just for the companies that are public and already reported. And so we assume we're going to get a slow grind lower in mortgage rates over the course of 2025 is our base case and we're planning accordingly. You've seen us take up marketing expense. You've seen us lean into growth. We pull back when it doesn't work. We lean in when it does. And so I think our view on the operating environment is much of the same for next year, but modest improvements throughout the course of 2025. Unlikely on a linear basis, this is generally not the way capital markets work. But we expect on this call November of 2025, we'll be in a marginally better operating environment and we're running the company accordingly.

A: Yeah. I mean, look, three years ago this time, when the rate started first picking up, we were doing 5% of our volume was purchased and 95% of our volume was refinanced. We've had a tough couple of years, but we have pivoted the company super hard. And now 71% of our volume, excuse me, is purchase mortgage, which I think is bigger than the bulk of our direct-to-consumer online peers. And so we've figured out how to do purchase mortgage online. We're cracking the code further and learning further from some of the partnerships we're going to be doing with the local retail operations out there. And I think that's just going to make us better. And so if the market environment stays difficult, I think you'll just continue to see us improve our execution on purchase mortgage and continue to drive growth through superior customer acquisition costs, superior cost of manufacturing, and lower cost of production and greater efficiency and customer support through what we're doing with Betsy.

A: Yeah. I mean, it's an important point, right? Neo Specialty is purchase. We think we're getting stronger in purchase. There's a lot of pent-up demand for purchase, home purchase in this country, right? And we know all the reasons why it's been slow to get going around affordability, both rates and housing prices. But we feel like we're putting the best technology in the industry into the hands of these local experts and that should further drive our ability to penetrate purchase in a low refi volume environment.