Latest / Chrisman Commentary - Daily Mortgage News / 8.31.26 Fair Housing Update; Two Dots’ Henson Orser on Underwriting; Warsh's Birdsong

Transcript

8.31.26 Fair Housing Update; Two Dots’ Henson Orser on Underwriting; Warsh's Birdsong · Chrisman Commentary - Daily Mortgage News

• 0:02
  Speaker 1: Welcome to the Chrisman Commentary, Daily Mortgage News Podcast.
• 0:05
  Speaker 1: I'm your host, Robbie Chrisman.
• 0:07
  Speaker 1: Topics on today's episode include some fair lending updates.
• 0:13
  Speaker 1: Why the Fed may sit on its hands until 2027.
• 0:16
  Speaker 1: In my interview with Two Dots Henson Orser on the wave of innovative tech products and vendors flooding the mortgage market.
• 0:23
  Speaker 1: Here.
• 0:24
  Speaker 1: In your estimation, what does mortgage underwriting nirvana look like?
• 0:33
  Speaker: Well, I think everybody can imagine a world where you know you could sort of get approved for a mortgage, um, even if it's a complex non-QM mortgage nearly instantly rather than it ha it being a six-day process.
• 0:48
  Speaker: Now, obviously, it's not going to be like getting a credit card and you just click a button on an app and wham bam, you're approved.
• 0:55
  Speaker: It'll be slightly more complicated than that.
• 0:57
  Speaker: Um, but I think you know, you could just imagine that like the experience of you know, you're buying a Tesla that's instant, or you know, even the like happy path on a rocket or something where it actually is relatively fast and easy.
• 1:10
  Speaker: And you know, SoFi has these for some types of mortgages.
• 1:14
  Speaker: It is relatively fast and easy to get approved in a digitally native way.
• 1:17
  Speaker: You know, expanding that to all the more complex mortgages and non-QM and people who maybe don't fit into the normal underwriting box where you have this conversational interface that just makes the whole process feel modern, easy, and fast as opposed to a six-week slog.
• 1:36
  Speaker 1: Can you walk us through how that actually works, getting to a decision faster?
• 1:42
  Speaker: Yeah, so like you could imagine you're maybe applying for a bank statement loan, which is one of the most document-intensive non-QM loans.
• 1:52
  Speaker: And you know, your broker helps get you the initial 12 months of bank statements together, but there's a large transaction in one of the bank statements, which the underwriter is going to want a letter of explanation for.
• 2:03
  Speaker: And you submit all the documents.
• 2:05
  Speaker: No one realizes that that large transaction is going to need a letter of explanation up front.
• 2:09
  Speaker: It gets submitted to the underwriter, maybe the underwriter is Deep Haven or something.
• 2:13
  Speaker: And then, you know, Deep Haven a day or two later is sending a conditional back to the broker.
• 2:17
  Speaker: It's like, hey, you've got to go get a letter from Joey to explain, like, you know, why did this person wire him $40,000 five months ago?
• 2:23
  Speaker: And all of a sudden we're on like day three or four.
• 2:26
  Speaker: And you could imagine, you know, if Joey's the borrower, he uploads his 12 bank statements.
• 2:32
  Speaker: Instead, like a chatbot is doing a pre-qualification where it already knows that like large deposits need a letter of explanation.
• 2:38
  Speaker: And it just asks him for that in real time and natural language and like generates that letter in natural language.
• 2:44
  Speaker: And you can extend that to any of the conditions that come along with a mortgage, any extra documentation that you need, any issues that come up in the appraisal process and on down the line.
• 2:55
  Speaker: Um so instead of having this kind of multi-party, multi-day email thing, you're just kind of getting real-time feedback as the borrower to get everything you need up front.
• 3:08
  Speaker 1: Thanks to Zillow Home Loans, Zillow's in-house mortgage lender, for sponsoring this week's podcasts.
• 3:13
  Speaker 1: By integrating Zillow's real estate platform with financing, Zillow Home Loans helps buyers move from dreaming about a home to holding the keys.
• 3:20
  Speaker 1: With tools built for modern lending, Zillow Home Loans loan officers can focus on guiding buyers with care and confidence.
• 3:26
  Speaker 1: Zillow Home Loans is an equal housing lender.
• 3:29
  Speaker 1: More than a hundred community-based fair housing organizations facing potential funding losses received a reprieve after a federal judge in Massachusetts blocked a U.S.
• 3:41
  Speaker 1: Department of Housing and Urban Development funding overhaul that advocates said threatened to close or advocates said threatened to close organizations or sharply reduced services nationwide.
• 3:52
  Speaker 1: The U.S.
• 3:52
  Speaker 1: District Judge granted relief sought by the National Fair Housing Alliance and Massachusetts Fair Housing Center, setting aside HUD's fiscal year 2025 funding notices and directing the agency to distribute the money under the structure used the previous year.
• 4:06
  Speaker 1: Housing statistics are all over the place, but in general housing is slow.
• 4:12
  Speaker 1: Single family housing starts keep falling, and July's 808,000 annualized are at their lowest or second lowest level since July of 2020.
• 4:21
  Speaker 1: July new home sales this year at 607,000 annualized are down 6.3% year over year, and haven't been consistently this low since 2017.
• 4:30
  Speaker 1: Moreover, median prices are at a five-year low.
• 4:33
  Speaker 1: As a percentage of GDP residential fixed investment, that's home construction, renovation, repairs, etc., continues falling and is at a level associated with recessions.
• 4:44
  Speaker 1: Student housing starts blew the proverbial roof off.
• 4:47
  Speaker 1: Headline starts surge 19% month over month to 1.427 million annualized units.
• 4:52
  Speaker 1: Regrettably, single family starts slid 0.2%, continuing their slow, steady two and a half year deterioration, while the lumpy, noisy, and volatile multifamily sector saw starts roar to 76.2% month over month, two or three-year higher with 532,000 annualized units.
• 5:09
  Speaker 1: This will add to the rental supply overhang, which will reinforce the downtrend in shelter components of CPI.
• 5:15
  Speaker 1: Softening home prices, rising rates, and increasing fraud are creating a growing risk management challenge for lenders and the GSEs.
• 5:22
  Speaker 1: Repurchase claims are creeping higher, with missing or ineffective private mortgage insurance reportedly accounting for the majority of claims, while occupancy and income fraud add further exposure.
• 5:32
  Speaker 1: Falling home values could push more loans into PMI required territory and increase losses when loans default.
• 5:39
  Speaker 1: Fraud is becoming more prevalent in DSCR and investor property lending, including fabricated leases, inflated appraisals, altered financial documents, and borrowers misrepresenting investment properties as rentals while actually occupying them.
• 5:52
  Speaker 1: Lenders should pay less attention to politically appealing wins and more attention to repurchase, underwriting, collateral, and fraud risks that could materially affect mortgage profitability.
• 6:03
  Speaker 1: The Federal Reserve doesn't set mortgage rates, but we still want to know what the group is thinking.
• 6:10
  Speaker 1: Fed Chair Warsh's Jackson Hole speech was overtly hawkish, emphasizing inflation and rate hikes as the remedy.
• 6:16
  Speaker 1: He explicitly said this summer's encouraging inflation ratings don't indicate meaningful improvement on inflation.
• 6:22
  Speaker 1: A rate hike at the September policy meeting is now being priced in as slightly better than a coin toss, but you should recall that Walsh loves to screech like a hawk and vote like a dub, so his remarks should be viewed as more philosophical than forward guidance.
• 6:34
  Speaker 1: The base case is still for the Fed to remain on hold this year amid continued disinflation, with Walsh hoping his comments get long bond yields to fall a little.
• 6:43
  Speaker 1: For today's interview, I wanted to welcome to the show Two Dots hence an Orsear to talk about the wave of innovative tech products and vendors flooding the mortgage market.
• 6:53
  Speaker 1: Co-founded Two Dots with the intent of unlocking unstructured financial documents, which unlock smarter financial decisions.
• 7:00
  Speaker 1: Whether it's applying to rent an apartment or take out a loan, today's approval process is full of blind spots and inefficiencies.
• 7:07
  Speaker 1: Critical data sits inside locked documents, leaving companies with an incomplete picture that causes delays, increased risk, and inconsistent decisions.
• 7:15
  Speaker 1: His ambition was to build a better system where underwriting and screening is automated, not manual, and applications happen in real time within a dynamic and contextual conversation.
• 7:26
  Speaker 1: There's this whole new wave of young tech forward, either founders or companies, why combinator-backed things.
• 7:38
  Speaker 1: Yeah, they're entering the mortgage industry.
• 7:41
  Speaker 1: And it's exciting.
• 7:43
  Speaker 1: It's going to push the industry forward.
• 7:45
  Speaker 1: There's also a whole bunch of legacy people, some of which I'm sure are listeners to this podcast of this podcast that would say, We we've done it this way for so many years, and these people think they can come in and change it, and there's a reason that it hasn't been done.
• 7:58
  Speaker 1: Frequent listeners of this podcast, now I worked at SoFi in the early 2010s, and we got a billion dollar round from SoftBank, and we thought we were the B's and E's and we were going to change mortgage, and that didn't quite come to fruition.
• 8:11
  Speaker 1: So there's a lot to unpack, but maybe to ask a direct question, it would be Can you talk about the opportunity that mortgage presents and why that's so alluring to so many young, intelligent people out there?
• 8:25
  Speaker: Yeah.
• 8:26
  Speaker: Well, why is mortgage alluring?
• 8:28
  Speaker: I mean, there's probably two big reasons.
• 8:31
  Speaker: One, of course, is the American dream and homeownership is like one of the most important things in our society, and financing that for people is also one of the most important parts of making that a reality.
• 8:44
  Speaker: But I think the other thing from a more tech standpoint is um yes, there are reasons all these processes exist, but you know, it's hard to apply for a mortgage.
• 8:56
  Speaker: Most people, you know, talk about getting approved for a mortgage as a very difficult process that uh could be better.
• 9:04
  Speaker: And I think that's evidenced by the fact that you know people like Rocket Mortgage and so far have been trying to innovate on this stuff for a long time.
• 9:11
  Speaker: Everybody knows that there's a problem.
• 9:14
  Speaker: You know, even if you solve parts of it, it's still like a really, really, really big opportunity to help people achieve the American dream, achieve you know financial independence.
• 9:24
  Speaker: And it's just like anybody who lives in this space knows that like there's a lot of tough, annoying paperwork that like actually isn't necessarily that valuable or doesn't need to be done the same way it's always been done.
• 9:38
  Speaker 1: Knowing that there are problems with mortgage, it's ripe for disruption.
• 9:43
  Speaker 1: There's been a lot of attempts.
• 9:45
  Speaker 1: Why are we still talking about this?
• 9:47
  Speaker 1: Why is that still kind of exist out there as this North Star, like it's doable?
• 9:51
  Speaker 1: I could somebody's gonna do it, somebody's gonna crack mortgage at some point.
• 9:53
  Speaker 1: Why haven't we gotten there yet?
• 9:55
  Speaker: That's a great question.
• 9:56
  Speaker: So, you know, my uh my co-founder was um early at Blend when we were founding this company.
• 10:02
  Speaker: We talked a lot about that experience and why Blend really failed to kind of fully automate mortgage underwriting and just became really an easier front end for people to apply for a mortgage at a large bank.
• 10:14
  Speaker: I mean, I think the core thing, which has just changed now, is traditionally like you're making a lot of decisions in mortgage based on like unstructured data and documents, and like that sort of document understanding stuff in the past was never really good enough to get to full automation.
• 10:30
  Speaker: It's finally at like a really high, good accuracy level where you could maybe do a lot more automation.
• 10:36
  Speaker: But that's even less of a big thing than the part that was never really solved, which is actually very important, is when you're underwriting a mortgage, you're doing a lot of back and forth with the borrower.
• 10:48
  Speaker: Um, and you're trying to get additional information and get clarification, or you're doing back and forth to get the appraisal or whatever it is.
• 10:54
  Speaker: So there's actually this coordination communication aspect that has been kind of a lot of the hidden work within underwriting and processing and brokering mortgages.
• 11:05
  Speaker: And like, you know, what solves all of that kind of coordination and natural language communication to collect things and get real-time feedback on the documents you need?
• 11:13
  Speaker: Well, like LLMs, LLMs are new.
• 11:16
  Speaker: Um, and to build the compliance infrastructure and workflows around them that make them and integrations that make them functional in the mortgage industry are are just fresh and coming out of the oven right now.
• 11:27
  Speaker 1: Of all the different parts of the mortgage process to disrupt, why did you choose underwriting?
• 11:36
  Speaker: Um I think it goes back to those two things we were just talking about.
• 11:40
  Speaker: What are the things where like the fundamental technology shift like makes it easier to do something you know different and new now?
• 11:48
  Speaker: One of those things is like kind of better document understanding.
• 11:50
  Speaker: But the big thing is you can have an interactive process with the borrower in order to get all the information and get them through the process faster.
• 12:01
  Speaker: So it just seemed clear that like that was the place where the technology could like and this new technology, this new AI LLMs thing could be implemented well to like solve a new real problem.
• 12:16
  Speaker 1: So either this is the $2,000 question or $12,000 question or $2 trillion question or $35 trillion question.
• 12:24
  Speaker 1: At what point does technology actually reduce the cost to originate and start saving companies money to their bottom lines, or are you know some of that be able to be passed on to borrowers in the form of lower rates?
• 12:37
  Speaker: Maybe it'll take three to five years.
• 12:39
  Speaker: It probably won't be tomorrow, but we are able to do a lot of these like very labor-intensive tasks much cheaper.
• 12:46
  Speaker: So I think the end state is probably milestones.
• 12:51
  Speaker 1: What is it going to take to actually see that?
• 12:52
  Speaker 1: Because a lot of the technology exists out there and companies have implemented it, but we haven't seen reductions in headcount or some of these efficiencies actually translate to cost savings yet.
• 13:01
  Speaker 1: Like what are the steps for us to actually get there?
• 13:04
  Speaker: That's a great question.
• 13:05
  Speaker: I mean it it just uh it depends on exactly what sector of the mortgage industry we're talking about and you know what part of the process.
• 13:13
  Speaker: Non-QM, there's a lot of fees and labor in the distribution.
• 13:19
  Speaker: Um, so that's probably like one of the places that you can attack up front where there's like actually, you know, probably you know a point plus of commission in some of these more complex mortgages.
• 13:29
  Speaker: But it's a slow process where you have to begin with the first steps and make it easier and faster, and uh that eventually translates into lower cost.
• 13:40
  Speaker 1: When we think about going from manual to automated, you obviously have a lot of experience in working with companies to talk about the adoption and implementation.
• 13:52
  Speaker 1: Best practices for moving from manual to automated, getting teams to buy in, maximizing ROI of some of the this new technology out there?
• 14:04
  Speaker: Yeah, at least the design choice we usually make, which we think is also best for the organizations that we sell to, is rather than coming in and saying, hey, on day one, you're just gonna hand the keys to the car to us and we're gonna drive it faster and better than you've ever driven it.
• 14:19
  Speaker: We try and build things that are more akin to you know sidekicks or assistance.
• 14:24
  Speaker: So you can just imagine that somebody who is a processor or an underwriter uh just gets these sort of AI power tools and assistance to use where they can direct the agent that's communicating with the borrower about exactly what's needed so that you're taking the folks who are kind of working on the ground doing the processing and just kind of giving them superpowers to start.
• 14:50
  Speaker: And then over time, once you've given them superpowers, they learn to trust it more and they become much more efficient.
• 14:56
  Speaker: So you kind of walk before you run, and then you know eventually there are parts of the process where they're not double checking anymore, right?
• 15:02
  Speaker: And then it's like, okay, well, maybe we can stop looking at that entirely.
• 15:06
  Speaker 1: You founded the company four years ago or so.
• 15:09
  Speaker 1: What do you wish you knew back then about the industry?
• 15:12
  Speaker 1: What have been some fun or interesting or tough lessons to learn along the way?
• 15:17
  Speaker: I mean, I've learned a lot and you know, uh learned a lot the hard way.
• 15:22
  Speaker: One thing that I wish I kind of knew about the industry.
• 15:26
  Speaker: I used to work at uh Goldman, so I kind of lived on a different side of the industry for a long time and then was involved in another tech company before starting this one.
• 15:36
  Speaker: I think like I probably would just tell myself, like, um, hey, like it's gonna be really, really, really, really, really hard, but like that's that's the part that makes it worth it.
• 15:46
  Speaker: It's not specifically about mortgages and it's more about entrepreneurship, but um I think it rings true.
• 15:52
  Speaker 1: So we've focused on the underwriting side of things for this interview, but there's also a whole fraud component that two dots helps companies solve.
• 16:01
  Speaker 1: And obviously with AI, the proliferation of fraud is certainly out there.
• 16:07
  Speaker 1: The ability of bad actors to utilize AI for fraudulent purposes is out there.
• 16:11
  Speaker 1: You talk about what you're doing on the the fraud prevention side of things, and the overall I mean the overall fraud space and mortgage in general, uh how you how you see that evolving?
• 16:20
  Speaker: Yeah, you know, we kind of view if you're gonna be doing automation, then the anti-fraud is sort of table stakes.
• 16:26
  Speaker: If you can't trust the data you're doing automated decisioning on, um, maybe because somebody's committing fraud, you're just opening up this like big vector for people to potentially take advantage of you.
• 16:37
  Speaker: So in addition to the kind of core technology of the borrower-facing chatbot that helps processors and underwriters get information up front and do pre-calls, um, we have document anti-fraud built into the system that makes sure, hey, does this you know, is this bank statement edited?
• 16:55
  Speaker: Is it real?
• 16:56
  Speaker: Is the metadata correct?
• 16:58
  Speaker: And there are you know parts of mortgage like non-QM that have a surprising amount of fraud where you have borrowers that can almost meet the requirements but don't quite meet the requirements, and they want to modify their documentation in order to like get around to a certain condition or requirement.
• 17:15
  Speaker: And we're building in the detection of all that stuff up front.
• 17:19
  Speaker: So it's not something that you know underwriting has to try and catch or you know, sniff out down the line or you know, even worse stuff that eventually leads to default.
• 17:28
  Speaker 1: A lot of your work started out in verification for apartment rentals.
• 17:32
  Speaker 1: And so much like that, we've seen with the the home equity side of things these these kind of niche products be able to influence mainstream lending.
• 17:42
  Speaker 1: And maybe I I kind of blubbered my way through that.
• 17:45
  Speaker 1: But it's neat to take some of these fringe approaches and say that could actually work really well for these companies that have said we're doing it for forever.
• 17:53
  Speaker 1: All that to say, what have you been working on recently?
• 17:55
  Speaker 1: What's your roadmap?
• 17:56
  Speaker 1: What excites you about what's going on at two dots?
• 17:59
  Speaker: I think, yeah, some of the history of the company.
• 18:01
  Speaker: We were originally doing this kind of like consumer-facing underwriting automation for uh big institutional landlords of rentals.
• 18:10
  Speaker: So we were doing multifamily, then we're doing big institutional single-family rentals.
• 18:14
  Speaker: So, you know, big P firms uh like a Cerberus and maybe own 40 or 50 or 60,000 single-family rentals.
• 18:22
  Speaker: And uh, you know, we're processing almost 100,000 applicants through that kind of rental underwriting product today.
• 18:31
  Speaker: And then we've recently taken that and expanded it into residential mortgage.
• 18:36
  Speaker: And uh the first product which is in people's hands is in non-QM across the whole gamut of non-QM.
• 18:43
  Speaker: So the core technology is this chatbot that gives feedback to the borrower's base and the documents they upload and gets people to pre-quall faster, and then also helps you know underwriting and processing, uh, handle conditions faster and just overall get to the close and funding of the loan in a couple days as opposed to a couple weeks.
• 19:03
  Speaker: And then we have that anti-fraud integrated as well.
• 19:06
  Speaker: We also allow the processors and the underwriters to help steer the agent.
• 19:10
  Speaker: And we're primarily distributing that product to non-QM brokers today, but also underwriters and processors.
• 19:16
  Speaker: On a go forward, we're really just kind of continuing to focus on you know, how can we get the quality of the conversation, the quality of the underwriting that folks are having with our bot to such a high place that like the you know, speed continues to get better.
• 19:32
  Speaker: It requires a significant amount of work.
• 19:35
  Speaker: Like I am reading the conversations myself in the morning along with most people in the company, so we could like continue to improve the bot and have it say more reasonable things to the borrowers and get them through the process faster.
• 19:50
  Speaker 1: You preempted my final question here slightly, but it's it's good that you brought it up because I want to know at what point AI underwriting stops being a workflow automation tool and becomes a source of credit policy.
• 20:02
  Speaker 1: Or put another way, where do you draw the line between automating execution and actually delegating lending judgment?
• 20:09
  Speaker: Yeah, I mean, I think at least for us and our company today, you know, we are not changing people's credit boxes.
• 20:15
  Speaker: We're just taking all of the ops that it takes to get the data to get someone into the credit box and making that faster, easier, and better.
• 20:23
  Speaker: But we do think that in the future, if you can have this interactive session with the borrower, it makes it a lot easier to like get additional information from them when you need it.
• 20:35
  Speaker: Uh so you could imagine more sophisticated uh credit boxes or underwriting in the future, where you know, depending on who the applicant is and what stage they're in, you know, you might change the requirements or get more information from them.
• 20:48
  Speaker: I think the thing that is maybe even more exciting, it doesn't involve changing the credit box or changing credit underwriting rules, is in real time just offering people different products that they might qualify for.
• 21:02
  Speaker: So you could imagine you're applying maybe in the non-QM space for a DSCR loan because you're a property investor, and then you actually don't qualify for that based on the property's income, but you could qualify for a bank statement loan based on your kind of self-employment personal income.
• 21:18
  Speaker: And we kind of already know that from the information you've given to us, and we you know in real time move you to that program.
• 21:24
  Speaker: Those are the types of things that like brokers are doing today or trying to do today that one could do in like real time with AI that would you know increase the amount of loans you're actually issuing and increase win rates.
• 21:36
  Speaker 1: I'm excited for kind of the inception of the mortgage industry.
• 21:40
  Speaker 1: I mean, inception is in the the movie out there when there's going to be AI supervising dozens of AI agents sort of thing, and it's like this meta meta.
• 21:52
  Speaker 1: We're we're at least three to five years from now.
• 21:55
  Speaker 1: We'll see what happens in the 2030s.
• 21:57
  Speaker 1: Henson, I really enjoyed this, man.
• 21:59
  Speaker 1: We waited way too long.
• 22:00
  Speaker 1: Pleasure was all mine, and uh hopefully we'll do this again soon.
• 22:02
  Speaker: Yeah, thanks, Robbie.
• 22:03
  Speaker: It was good to be here.
• 22:06
  Speaker 1: The treasury market is increasingly reflecting a combination of weaker demand and rising term premium pressures pressures with mid-curve yields near multi-year highs and long-term rates pushing toward the upper end of the recent ranges.
• 22:18
  Speaker 1: At the same time, deteriorating consumer fundamentals, such as real disposable income, lagging spending for 25 straight months, and income growth slowing to just 0.5% year over year, suggests that elevated borrowing and energy costs could increasingly constrain consumption.
• 22:34
  Speaker 1: The third-year treasury yields, 37-day stretch above 5%, alongside rising 100 and 200-day moving averages, points to a shift toward a higher rate regime that would be materially bearish for long-duration bonds.
• 22:47
  Speaker 1: This week's focus will be on the August U.S.
• 22:49
  Speaker 1: Employment Report, where non-farm payroll growth is expected to rebound to 80,000.
• 22:53
  Speaker 1: While job growth remains subdued, labor market conditions continue to look broadly balanced, with low layoffs and steady wage growth consistent with a gradual cooling rather than a sharp deterioration.
• 23:03
  Speaker 1: With nothing of note on today's economic calendar, things pick back up tomorrow with manufacturing figures and July construction spending.
• 23:11
  Speaker 1: Outside of payrolls, highlights from the rest of the week include August ADP employment and the September Fed page book.
• 23:17
  Speaker 1: After the US and Iran traded strikes for the first time in about a month, we begin the week with agent CMBS prices.
• 23:23
  Speaker 1: Little change from Friday's close, the two-year yielding 4.32, and the 10-year unchanged from Friday and yielding 4.72%, down to two basis points over the course of last week.
• 23:34
  Speaker 1: Let's wrap up with a joke and some housekeeping.
• 23:40
  Speaker 1: When you listen to rock backwards, you hear satanic messages.
• 23:43
  Speaker 1: What do you get when you listen to country music backwards?
• 23:47
  Speaker 1: Well, your wife comes back, your life comes back, you stop drinking, and your dog comes back.
• 23:52
  Speaker 1: Thanks again to this week's podcast sponsor, Zillow Home Loans, Zillow's in-house mortgage lender.
• 24:01
  Speaker 1: With tools built for modern mortgage lending, Zillow Home Loans loan officers can focus on guiding buyers with care and confidence.
• 24:08
  Speaker 1: To learn more, visit Zillow.com/slash home loans.