Latest / Chrisman Commentary - Daily Mortgage News / 10.7.26 Headlines Into Annual; Ralo's Arjun Lalwani on Anthrotech Intersections; Fundamentals and Valuations

Transcript

10.7.26 Headlines Into Annual; Ralo's Arjun Lalwani on Anthrotech Intersections; Fundamentals and Valuations · 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 big headlines going into MBA annual, how rates are impacting lenders' futures, and my interview with Ralo's Arjun Lalwani on using AI to reduce mortgage origination costs and operational work while preserving human support for advice in emotionally significant moments.
• 0:29
  Speaker 1: Here, take a listen, do a little preview.
• 0:32
  Speaker 2: When we think about time savings, cost savings.
• 0:36
  Speaker 2: You you know as well as I do that it costs $10,000, 11, $12,000, $13,000 to originate alone.
• 0:41
  Speaker 2: Where do you see the biggest time saving potential out there?
• 0:45
  Speaker 2: Where do you see the biggest cost saving potential out there?
• 0:48
  Speaker: The biggest time saving is actually very straightforward.
• 0:50
  Speaker: It's all the operational work, right?
• 0:51
  Speaker: It's like, oh, you have to follow up, you have to ask them about this document, um, they want a pre-approval letter, like like these routine tasks that AI is very good at handling, understands like very clear like uh data that is structured, it knows what's missing.
• 1:04
  Speaker: It's all text.
• 1:04
  Speaker: Like AI is doing like image and video and all these things.
• 1:07
  Speaker: Text is like the easiest thing for it to understand, right?
• 1:09
  Speaker: And numbers.
• 1:10
  Speaker: And so I think that stuff will be like easily automated, and people also feel more comfortable asking like an AI agent dumb questions, right?
• 1:17
  Speaker: Instead of bothering their loan officer at 11 p.m.
• 1:20
  Speaker: In those cases, we'll do really well.
• 1:22
  Speaker: In terms of like cost savings, we here think that the loan officer getting paid on like a commission basis is going to change.
• 1:30
  Speaker: That model will change.
• 1:31
  Speaker: The incentives will be help provide the best like service and the advice.
• 1:35
  Speaker: And in there are cases where you know a borrower has got a great deal in a credit union.
• 1:40
  Speaker: We're like, you should go for it.
• 1:41
  Speaker: I have no product to sell.
• 1:42
  Speaker: I make nothing on commission.
• 1:43
  Speaker: I want to do what's best for you.
• 1:45
  Speaker: And that borrower sent me like his friend the next day, being like, Oh, you should talk to them.
• 1:49
  Speaker: They give good advice.
• 1:50
  Speaker: I was like, Great, of course, let's connect, right?
• 1:52
  Speaker: So I think the cost model where you remove like a loan officer commission and pass that all as savings to the end consumer and you pay them on a different model, maybe in a base or maybe on some performance-based stuff, um, but not on like a commission where they're incentivized to push product.
• 2:06
  Speaker: Um, I think that is going to be a big cost saving.
• 2:08
  Speaker: And then obviously, behind the scenes, like underwriting and like all the other stuff, like we're a two-broker shop at this moment, and we have done almost like 10 million in like loans funded.
• 2:19
  Speaker: And so we have really learned and analyzed, and I think we can do a lot more, right?
• 2:24
  Speaker: So it's like we're trying to now scale.
• 2:26
  Speaker: Um, and we don't think we need like more loan officers.
• 2:29
  Speaker: I think we need a little maybe a little bit of engineering help, maybe a little bit of like marketing help, but I don't think we need another loan officer.
• 2:39
  Speaker 1: Thanks for today's podcast sponsor, FlowFi, the mortgage industry's leading point of sale platform.
• 2:44
  Speaker 1: Dynamic apps, which can be seen at Booth 600 during MBA annual next week, lets lenders create fully customizable loan applications for any type, including HELOCs, construction, agricultural lending, non-QM, and more without custom development.
• 2:58
  Speaker 1: To learn more, visit Flowify.com.
• 3:01
  Speaker 1: There's certainly a flurry of activity this week ahead of next week's National MBA conference, not all necessarily good.
• 3:11
  Speaker 1: Anyone impacted by the FICO layoffs this week or anyone looking for employment can post a resume or look at a job listings on the Crisman Job Board.
• 3:20
  Speaker 1: For decades, the market share split between Fanny and Freddie was historically about 6040.
• 3:27
  Speaker 1: Rumor has it that now the percentages have reversed and Freddie Mac is buying the lion's share.
• 3:33
  Speaker 1: Rumor also has it that employees at both agencies are anxious about their jobs in future.
• 3:38
  Speaker 1: But it would be best to ask them about morale and not take it from me.
• 3:42
  Speaker 1: For fans of the builder business, Lenar's stock is down 22% this year.
• 3:47
  Speaker 1: For the fans of the wholesale lending model, UWM holding stock fell 34% after hedging strategy issues were disclosed, and a securities fraud class action was filed.
• 3:57
  Speaker 1: When you're number one, you have a target on your back, and in this case, UWM has been sued for securities fraud after its stock plummeted 34.78% because UWM allegedly misrepresented its mortgage servicing rights hedging strategy and the risks created by hedging connected to the two harbors transaction.
• 4:16
  Speaker 1: Matt Ishbia, ex-MBA player Isaiah Thomas, and other UWM Holdings Corp directors may have shortchanged shareholders by engineering a $1.5 billion bailout from Oak Tree Capital Management after a bad bet on interest rates.
• 4:30
  Speaker 1: U.S.
• 4:30
  Speaker 1: Treasuries and agency mortgage backed securities rallied yesterday, with the long bond recovering most of Monday's losses and tenure and shorter maturities turning positive for the week, supported by broad overnight strength and global sovereign debt and a pullback in oil prices.
• 4:44
  Speaker 1: The $58 billion three-year treasury auction was well absorbed, although foreign demand was below average, with markets now looking ahead to today's $39 billion tenure reopening.
• 4:55
  Speaker 1: Treasury demand was also evident in the buyback operation, where the government accepted just $1.33 billion of the $14.76 billion offered, underscoring solid demand for shorter dated securities, even as the market remains attentive to rate volatility and upcoming supply.
• 5:11
  Speaker 1: Elevated yields reflect expectations for a higher neutral policy rate, persistent fiscal concerns, and rising term premiums, with bearish sentiment spreading from Europe, the UK, and Japan into U.S.
• 5:23
  Speaker 1: treasuries.
• 5:24
  Speaker 1: Much of the inflation pressure appears to be supply-driven, especially from energy, which could allow the Fed to tighten policy more gradually and keep markets sensitive to economic data and forward guidance.
• 5:35
  Speaker 1: Will the Fed pause in October and potentially resume tightening in December?
• 5:39
  Speaker 1: The September move could prove to be a rare one-off hike, but that will depend on upcoming inflation data.
• 5:44
  Speaker 1: Higher rates could pressure gain on sale margins and make pipeline hedging more difficult.
• 5:49
  Speaker 1: Earning estimates for Rocket and United Wholesale Mortgage are being reduced with UWM downgraded and its price target cut as lower earnings expectations and a higher discount rate more than offset the company's continued ability to redeem its oak tree preferred by 2030.
• 6:04
  Speaker 1: KBW's Bose George points out that title insurers face more modest earning pressure because refinances represent only about 7% of premiums, although purchase activity is expected to remain roughly flat in 2027, and higher rates could constrain commercial activity.
• 6:21
  Speaker 1: Estimates and price targets are therefore reduced while rates remain unchanged.
• 6:26
  Speaker 1: In contrast, mortgage insurers remain fundamentally attractive, with earnings largely insulated from origination volumes.
• 6:33
  Speaker 1: Solid expected third-quarter results, activity supported by low unemployment and stable home prices, and only modest pressure from rising rates and higher delinquencies.
• 6:42
  Speaker 1: Recent weakness in mortgage insurer shares appears driven largely by political and headline concerns surrounding FHFA director Bill Poulty, pushing valuations below historical averages.
• 6:55
  Speaker 1: For today's interview, I wanted to welcome to the show Rallo's Arjun Lalwani to talk about using AI to reduce mortgage origination costs and operational work while preserving human support for advice and emotionally significant moments.
• 7:10
  Speaker 1: He's co-founder and CEO of Rallo, the AI native mortgage brokerage, which was started in 2025 as approval AI and relaunched as a licensed brokerage in June of 2026.
• 7:21
  Speaker 1: He's a former Google product leader with expertise in marketing systems.
• 7:27
  Speaker 2: There's a whole wave of new tech entering the mortgage ecosystem.
• 7:32
  Speaker 2: There's a lot of really young, smart entrepreneurs that are bringing this tech and this AI to market.
• 7:38
  Speaker 2: Your thoughts on the opportunity you saw when it came to mortgage and specifically for Allo, kind of why you chose the product that you did.
• 7:49
  Speaker: Yeah, we're very excited by the amount of like AI unlock that can happen in this space here.
• 7:54
  Speaker: We were kind of, I'm sure you've read the stat, everyone's read this stat, the 13.12 to $13,000 cost of origination.
• 8:02
  Speaker: It's insane and it does not make any sense to us.
• 8:04
  Speaker: The more we dug into it, the more we learned about it.
• 8:07
  Speaker: We were like, wow, there are so many inefficiencies here.
• 8:10
  Speaker: And we totally think that we can automate a lot of this stuff.
• 8:14
  Speaker: The question is, why has it not been done?
• 8:16
  Speaker: Um and there are a bunch of these tools that are emerging right now, and you're seeing this on the B2B side, where a bunch of companies are like launching these tools to make loan officers a lot more efficient.
• 8:25
  Speaker: We're going direct to consumer.
• 8:26
  Speaker: We're like, it will take time for consumers to feel the impact of those savings because businesses have to adopt it, businesses have to learn how to use it.
• 8:36
  Speaker: We're like, let's just go direct.
• 8:38
  Speaker: Let's build an entire market shop with AI from the ground up from day zero.
• 8:43
  Speaker: It is going to take a lot more work.
• 8:44
  Speaker: We'll have to learn the industry inside out, which means we have to become loan officers ourselves.
• 8:49
  Speaker: We are doing all of this work.
• 8:51
  Speaker: I am on calls with like borrowers trying to understand what they want.
• 8:54
  Speaker: And we're training our AI to like get really, really good at that so they can automate it the next time.
• 8:59
  Speaker: Um, so we are excited to get the savings and push the industry forward and really give people the benefits of AI today instead of like three years from now when the entire industry catches up.
• 9:10
  Speaker 2: This is where we're going to anger a lot of legacy mortgage people.
• 9:14
  Speaker 2: Probably your goal is to fully automate the loan officer.
• 9:18
  Speaker 2: What made you think that was even possible?
• 9:20
  Speaker 2: And I will add the context of for a couple decades now, it was, oh, well, we can remove the real estate agent from the equation, or oh, we can remove the loan officer from the equation.
• 9:28
  Speaker 2: Neither of those things have quite happened yet.
• 9:31
  Speaker 2: And maybe, maybe you're the Elon Musk of the mortgage industry where you people have been trying electric cars for 100 years, but he finally made it work.
• 9:39
  Speaker 2: What makes you optimistic that you can actually make this work?
• 9:42
  Speaker: Robbie, I think I've updated my thinking on this a little bit.
• 9:45
  Speaker: I don't think I I'm going to look, there was a world where I thought uh loan officers could be completely eliminated and AI could do everything.
• 9:53
  Speaker: That was like a naive understanding of like the world.
• 9:55
  Speaker: Right now, I think it's very much where like the loan officer is available when the AI agent can't help you.
• 10:01
  Speaker: Um and it's exactly at those emotional moments.
• 10:04
  Speaker: It's exactly when like I put an offer down, I'm in contract help.
• 10:08
  Speaker: They don't want to talk to an AI agent, they want to talk to a human.
• 10:11
  Speaker: So there are these particular touch points where humans are very, very important when you have to build trust and let them know who you are, educate them about like this is what the service is.
• 10:20
  Speaker: By the way, I'm also looping in my AI agent in email and when my phone number, and there's a three-way group chat between me, my AI agent, and the home buyer.
• 10:29
  Speaker: All the customer questions, like I want a pre-approval letter, I'm putting an offer down at 10 p.m., my AI agent can handle that.
• 10:35
  Speaker: I don't need to be there for that.
• 10:37
  Speaker: Operational tasks go to the AI agent.
• 10:39
  Speaker: Anything that is advice related, that is, I'm having like an emotional moment and I need support, that's where I think the human will step in.
• 10:47
  Speaker: Which means over time, each human loan officer can probably do hundreds of mortgages.
• 10:53
  Speaker: And if there are a small sliver of people that do not want to talk to a human, we have dealt with those customers.
• 10:58
  Speaker: They're very tech forward.
• 10:59
  Speaker: Um, and for them, that AI agent is perfect.
• 11:02
  Speaker: So I think it's we're gonna see like a mix of people where over time they'll be comfortable, but in those, like in those precise moments, we will need that human touch point and we will offer that.
• 11:12
  Speaker 2: Let's talk about the time to close alone.
• 11:15
  Speaker 2: Because many will say, Oh, it takes most lenders 30 to 45 days to close alone.
• 11:21
  Speaker 2: Correct.
• 11:22
  Speaker 2: How are we measuring?
• 11:23
  Speaker 2: And I'll get to why I'm asking this.
• 11:26
  Speaker 2: How are we measuring that?
• 11:27
  Speaker 2: Because I would say what is the fastest the lenders that are closing on average in 45 days, what's the fastest they could close alone?
• 11:32
  Speaker 2: Maybe it's 15 or 20 days.
• 11:34
  Speaker 2: Yeah.
• 11:35
  Speaker 2: Rallo says we close in an average of 17 days.
• 11:38
  Speaker 2: Well, I guess in that same sense, what's the fastest that you could close alone?
• 11:43
  Speaker 2: Like, how let's let's standardize some of the figures of closing times we're seeing out there.
• 11:48
  Speaker: Yeah.
• 11:48
  Speaker: Um, so this all varies, right?
• 11:50
  Speaker: Like, actually, unfortunately, a lot of this stuff is sometimes out of our hands too.
• 11:55
  Speaker: It's like appraisal doesn't get back to us in time.
• 11:57
  Speaker: Oh, the property has like an appraisal waiver, which means we can close much faster now.
• 12:01
  Speaker: Or the title person's out of office and not responding back to our emails.
• 12:04
  Speaker: Our AI agent keeps following up with them and it's like, hey, hey, hey, please give me this talk.
• 12:09
  Speaker: So I think a lot of the variance really comes from things outside our control, or the borrower is not providing the document we exactly want, and we are like asking them over and over again.
• 12:17
  Speaker: So there are these inefficiencies that I think can dramatically change how quickly we can actually close the low one.
• 12:25
  Speaker: The fastest I think we've been able to do it is like 12 days.
• 12:27
  Speaker: It's a refinance, straightforward, the appraisal is waived.
• 12:31
  Speaker: You know, they just upload their documents, underwriter proves it, cool, like we're closed.
• 12:35
  Speaker: But the longest can be like 20, 25 days as well for us because we are dependent on these external factors.
• 12:41
  Speaker: So I would say on our side, things move instantly.
• 12:44
  Speaker: Borrower uploads, goes to underwriting, underwriter gives us like feedback.
• 12:47
  Speaker: We're able to communicate with the borrower.
• 12:49
  Speaker: When the borrower uploads something, we guide them through like, hey, this will not qualify for what the underwriting needs.
• 12:54
  Speaker: You probably need to adjust this, they're just that, right?
• 12:56
  Speaker: Like we can do all these tweaks in our platform.
• 12:58
  Speaker: And these are these small efficiency gains we're making today, which has brought down like our speed to 17 days.
• 13:05
  Speaker: But I think there's a lot more room to innovate on that side of things.
• 13:08
  Speaker 2: Yeah, the mortgage industry is notorious for there's there's kind of two camps.
• 13:14
  Speaker 2: There's the old codgers that say this is how it's always been done, this is how it's gonna be done.
• 13:17
  Speaker 2: There's these new tech people that say, Well, we can we can completely disrupt this.
• 13:20
  Speaker 2: And I worked at SoFi in the early 2010s, and we had a billion dollars for soft bank, and we thought we were gonna change mortgage, we didn't.
• 13:27
  Speaker 2: And so I'm I'm wondering kind of hindsight 2020 for you, what you know now, having been in it, that you you should have told maybe your more naive self, or is it always good to just dream and and that's going to get the industry to where it needs to be?
• 13:42
  Speaker: Yeah, um, it's a good question, Robbie.
• 13:44
  Speaker: I I think about this a lot, and every day I like update my thinking with every customer I work with.
• 13:50
  Speaker: I do think right now, what's really interesting is you can really understand unstructured data, like borrowers texting you about something, and your AI agent like reading that, understanding, like, wait, they're stressed.
• 14:05
  Speaker: That capability did not exist before.
• 14:08
  Speaker: And the AI can then tell me as a loan officer being like, this person's really stressed.
• 14:12
  Speaker: You need to call them and talk to them, otherwise, you're gonna lose the deal.
• 14:15
  Speaker: Cool.
• 14:16
  Speaker: Okay, I'm gonna call them.
• 14:17
  Speaker: Um, or this, uh, or the AI agent understanding like unstructured data, where they write a note explaining, I want to get a redify, I want to cash out, I want it to be the max amount.
• 14:26
  Speaker: Usually you require like a human to step in and be like, okay, what is the max amount?
• 14:30
  Speaker: How do I structure this deal?
• 14:31
  Speaker: Now you can have AI actually understand that note and be like, oh, this is why they need it.
• 14:36
  Speaker: And here's how I would structure this.
• 14:38
  Speaker: And our portal will not support this one-to-one, but I know exactly what to give them at this moment in time.
• 14:43
  Speaker: That efficiency gain is why I think we can really like change things.
• 14:48
  Speaker: And I do think like the more we lean into that, what we have noticed is people care a lot more about outcomes, not so much about like the process.
• 14:56
  Speaker: They want their pre-approval or they want to know today on a five-year ARM, like, how much will I get?
• 15:01
  Speaker: Can I close in 17 days?
• 15:03
  Speaker: Is that a yes or a no?
• 15:04
  Speaker: And what will what will make it go to 25?
• 15:06
  Speaker: The AI can answer all this stuff, right?
• 15:08
  Speaker: We just have to like give it the right context and do it in a language that the consumer kind of can match their language.
• 15:14
  Speaker: If they want detailed answers, you can give detailed answers.
• 15:16
  Speaker: If they want a short answer, we can give them a short answer.
• 15:18
  Speaker: And I think that ability did not exist before.
• 15:21
  Speaker: And I'm very excited about that coming into mortgages now.
• 15:23
  Speaker: And we're seeing people do this.
• 15:25
  Speaker: Like you're seeing Chat GPT, you're seeing people like use Claude, people are leaning into like personal finance apps like Clio.
• 15:31
  Speaker: A lot of like people are getting comfortable asking AI.
• 15:34
  Speaker: On ChatGPT, a lot of customers come to us and they're like, I asked ChatGPT, this is a good loan estimate.
• 15:39
  Speaker: And I was like, okay, cool.
• 15:40
  Speaker: I was like, if you're going to ask about it, here's a prompt, here's what you should ask, and ask to check for like points, ask to check for like the APR and check if there are like any hidden fees, and now compare our loan estimate with whoever else's loan estimate.
• 15:54
  Speaker: And we won't try to sell you, you ask ChatGPT, and they're like, Cool, I will go do that.
• 15:58
  Speaker: And then they come back to us being like, wow, yours is much better.
• 16:00
  Speaker: And I was like, Cool, all right, let's work together.
• 16:02
  Speaker 2: I spoke to a realtor the other day who said, Look, borrowers are using Chat GPT.
• 16:07
  Speaker 2: Now, if we're going to be using AI, we need to have gone 10 steps further already and be like, I am so far ahead of you, and whatever you think you've discovered on there.
• 16:14
  Speaker 2: Like, trust me, like, I'm the expert still, despite all the information that you can gather.
• 16:20
  Speaker 2: Let's talk your roadmap a little bit.
• 16:23
  Speaker 2: Where do you feel like the company is now in terms of practical use and application?
• 16:27
  Speaker 2: And where are you headed?
• 16:29
  Speaker: Yeah, I think we're in our early days.
• 16:31
  Speaker: Every day we discover new edge cases to solve, and we teach our AI, like, oh, you can now do this.
• 16:36
  Speaker: And if this kind of question comes in, here's how you answer it.
• 16:38
  Speaker: And they're getting much better at it.
• 16:40
  Speaker: And so we're very excited about that capability.
• 16:42
  Speaker: We literally had like one customer text our AI agent being like, I love Ralo.
• 16:46
  Speaker: And I was like, cool, it's working, right?
• 16:48
  Speaker: Like I see that.
• 16:49
  Speaker: I see that potential that they're like, they're talking and they're happy.
• 16:52
  Speaker: And it was like a 2 a.m.
• 16:54
  Speaker: text conversation.
• 16:55
  Speaker: I was like, great.
• 16:55
  Speaker: So I think in the next like six months, we're gonna be doing a lot of these loans.
• 17:02
  Speaker: I'm gonna be doing them.
• 17:02
  Speaker: My co-founder is gonna be doing them.
• 17:04
  Speaker: We're gonna learn this really inside out, teach it to the AI.
• 17:06
  Speaker: And over time, right now it's like 50% of it's handled by AI.
• 17:09
  Speaker: I would want like 80, 90% of it to be handled by AI.
• 17:12
  Speaker: And I think we can see that within within 10 million in funded volume, we've already gotten to 50.
• 17:17
  Speaker: I don't see a world where like I come across something where I'm like, hmm, I don't think the AI can do that.
• 17:22
  Speaker: Other than that emotional moment where I need to like talk to them.
• 17:25
  Speaker: I haven't come across a single situation so far where the AI can't handle this.
• 17:30
  Speaker: I'm excited to like give our AI that that capability, and that is going to come with a lot of experience.
• 17:36
  Speaker: And we just have to get our hands dirty and do this job in and out because we understand the tech and we are understanding mortgages now and getting really good at it.
• 17:43
  Speaker: And that's that's when we marry those two, where you will see like a beautiful user experience.
• 17:50
  Speaker 2: It's interesting.
• 17:51
  Speaker 2: That's how you view it because a lot of people say, well, AI is good for the conventional conforming, down the middle, vanilla, fair way type stuff.
• 17:58
  Speaker 2: And you're going, well, I think it can actually handle a lot of these fringe cases.
• 18:02
  Speaker 2: And so that's a refreshing kind of demarcation point for the industry to see somebody still saying, Well, I'm going to solve these edge one, and that's actually what's going to work well.
• 18:12
  Speaker: Totally, totally.
• 18:13
  Speaker: I mean, just I think we did our first DSCR recently, and we were just shocked at how complex the file was.
• 18:19
  Speaker: And we're like, you know, there's again, nothing here is something that the AI can't handle.
• 18:24
  Speaker: It's like we have to teach it.
• 18:25
  Speaker: And for that, we need to know what to teach it.
• 18:27
  Speaker: Um, so we had to do that DSCR by hand, figure out everything, learn about it, and then teach our AI to do it.
• 18:33
  Speaker: And now, like this morning, uh, I'm helping a customer with a DSCR.
• 18:36
  Speaker: And guess what?
• 18:36
  Speaker: Like, my AI agent is doing most of it.
• 18:39
  Speaker 2: Right.
• 18:39
  Speaker 2: So for people that want more information, where should they go?
• 18:43
  Speaker 2: How can they connect with you?
• 18:45
  Speaker: Yeah, um, ralo.com is our site.
• 18:47
  Speaker: Um, you can go there, learn more about us, sign up, get a free rate code as well.
• 18:52
  Speaker: And uh, if you want to connect with me directly, it's uh LinkedIn is the best way to reach me, uh linkedin.com slash uh Arjun dash Lalamani or Arjun at Ralo.com.
• 19:02
  Speaker 2: Perfect.
• 19:02
  Speaker 2: Arjun, I really appreciate the time, man.
• 19:04
  Speaker 2: I wish you the best of luck.
• 19:06
  Speaker 2: As I was saying to you offline, there's some very smart people I know in the industry that are uh looking at what you're doing and are and are impressed.
• 19:12
  Speaker 2: So keep up the good work and hopefully we'll talk again soon.
• 19:15
  Speaker 2: That's awesome.
• 19:16
  Speaker: Thank you, Robbie.
• 19:16
  Speaker: Do you want to have a couple of moments to ask you a couple of questions too?
• 19:20
  Speaker: Sure.
• 19:20
  Speaker: Do you want to be part of the podcast or um up to you?
• 19:23
  Speaker: I I'm just curious.
• 19:24
  Speaker 2: All right, go ahead.
• 19:25
  Speaker 2: Yeah, go ahead.
• 19:25
  Speaker 2: We'll we'll see here.
• 19:27
  Speaker: Um I mean, from from your experience, right?
• 19:29
  Speaker: Like obviously you worked at SoFi, you saw like they were going to change mortgage.
• 19:33
  Speaker: Change mortgages and like add digitization.
• 19:37
  Speaker: What do you think?
• 19:38
  Speaker: I wouldn't say it went wrong, but what do you think kind of failed in their thesis where I just applied like I do this all the time.
• 19:45
  Speaker: I go and like apply to a bunch of different banks and see what their like sales motion looks like, right?
• 19:49
  Speaker: And where they're using AI, where they're not.
• 19:51
  Speaker: And just like I've been getting a loan officer from SoFi following up with me for the past like 10 days, every day.
• 19:58
  Speaker 2: Where do you think Yeah, yeah, I understand the question.
• 20:00
  Speaker 2: So I I think that what SoFi got right before its time.
• 20:04
  Speaker 2: So SoFi was before its time in in a variety of ways.
• 20:07
  Speaker 2: What they got right before its time was being a non-depository that was able to engender customer loyalty.
• 20:17
  Speaker 2: Hey, you get out of school, you refinance with a student loan.
• 20:20
  Speaker 2: Hey, we offer personal loans.
• 20:21
  Speaker 2: Now we're going to offer a mortgage, now we're going to get into wealth management products.
• 20:24
  Speaker 2: And so we have that customer for life strategy.
• 20:28
  Speaker 2: That's that's what they got.
• 20:29
  Speaker 2: And now the mortgage industry is very focused on recapture and retention and buying servicing, and that's so SoFi was well ahead of its time.
• 20:35
  Speaker 2: SoFi also was a little ahead of its time with thinking that either the there was technology that would make for the end-to-end mortgage, or customers were ready for a fully digitized mortgage.
• 20:49
  Speaker 2: Even if cut and even if customers were, a lot of the process still wasn't.
• 20:55
  Speaker 2: You know, whatever part of the origination process you look at, appraisals or inspections or underwriting and clearing, all those things have made leaps and bounds in the last 10 years.
• 21:04
  Speaker 2: And so for SoFi, I say we want a fully digital mortgage.
• 21:07
  Speaker 2: Yeah, you might be able to design something, but it still has all these different manual steps that takes time, manual stuff that takes time.
• 21:14
  Speaker 2: And so the entire industry now has become more automated, where I think that a lot of what SoFi was trying to do is now much more possible.
• 21:23
  Speaker 2: And I would add to that, though, the cautionary tale of remote online notarization, which yeah, the pandemic helped remote online notarization, but it's still not accepted in all 50 states.
• 21:36
  Speaker 2: And we've been talking about this for nine or 10 years at this point.
• 21:39
  Speaker 2: Something that seems so simple of like, yep, a digital notary, I go on, I sign, I don't have to go to the closing.
• 21:44
  Speaker 2: It's taken that so the mortgage industry has just been so long to actually get stuff in, even if they want it.
• 21:52
  Speaker 2: So, like I said, I'm hopeful that now with AI disrupting every single part of the process, if you build something from Scratch with AI, in theory, that'll work.
• 22:04
  Speaker 2: Right.
• 22:04
  Speaker 2: A lot of really brilliant ideas in the mortgage industry have faltered because of regulations and just you know the ways that people have done it and people that are have uh gate kept and and various reasons.
• 22:17
  Speaker: Got it, got it.
• 22:18
  Speaker: Yeah, that's that's interesting.
• 22:19
  Speaker: I'm also curious.
• 22:20
  Speaker: Like, you talk to a bunch of people in the industry, like where do you see like when you when you look at all the AI innovation happening, what do you look at and be like, I've seen that movie before and that's not gonna work, versus, oh, I'm really excited about this one.
• 22:34
  Speaker: Like, what parts of like the AI application get you excited versus others not so much?
• 22:39
  Speaker 2: I get the sense that there's a borrower that's always going to want a phone call and a sit down with somebody in person, and there's somebody that's going to want to, and so it's almost like I don't think there can be a some a company that's going to design for the full spectrum of borrower.
• 22:55
  Speaker 2: I almost think that it's different pieces within that you're going to be able to optimize.
• 23:00
  Speaker 2: What gets me excited?
• 23:02
  Speaker 2: I think underwriting in real time gets me excited.
• 23:09
  Speaker 2: Being able to clear conditions like that versus having back and forth.
• 23:13
  Speaker 2: I think that's that's a big part of the process.
• 23:15
  Speaker 2: I mean, I could say that you know, Fannie and Freddie accepting automated valuations, and so we don't have to wait six days for an appraise.
• 23:21
  Speaker 2: That doesn't really get me excited.
• 23:23
  Speaker 2: Oh, I can go scan a house with my phone and I get an appraised value.
• 23:26
  Speaker 2: That doesn't really get me excited.
• 23:27
  Speaker 2: You know, clearing underwriting conditions is a big deal.
• 23:30
  Speaker 2: Uh then my mind almost wants to say the voice calling, but my experience, I'm I'm withholding judgment until it feels fully real.
• 23:42
  Speaker 2: Because I go, I go back to the example of the the airlines introduced you know an automated help assistant for you 20 years ago.
• 23:52
  Speaker 2: But I just want to I'm like zero, zero, zero until I talk to a person.
• 23:56
  Speaker 2: When AI truly gets to the point where I cannot tell, and and for boomers, I think they're there where like they can't tell they're talking to AI.
• 24:04
  Speaker 2: I think younger people still can some of the time.
• 24:06
  Speaker 2: Yeah, when it gets to the point of of I can't tell, and it can give me everything, then I think that's that's exciting too.
• 24:14
  Speaker 2: But I and I I said it to you kindly during the podcast, I'll say it a little more bluntly now.
• 24:19
  Speaker 2: You know, this has been the realtor or the the LO.
• 24:22
  Speaker 2: People have been trying to take them out for years and years, and it hasn't happened.
• 24:26
  Speaker 2: And and obviously, when we talk cost to originate alone, duh, the biggest cost, six thousand bucks of the 12,000, is is sales comp.
• 24:35
  Speaker 2: Yeah.
• 24:36
  Speaker 2: But it's this it's a weird sacred cow where like the really good innate originators are worth that in certain ways, bad originators are not.
• 24:45
  Speaker 2: And so I'd almost think that we'd move toward people almost choosing your choose your game, choose your option, choose your path of like, do you want a fully digitized talk to nobody?
• 24:55
  Speaker 2: Like, yeah, you're gonna pay a lot less.
• 24:57
  Speaker 2: If you want to pay less, you're probably not gonna get as great of service.
• 25:00
  Speaker 2: Maybe technology will get it there eventually.
• 25:02
  Speaker 2: If you want really good white glove concierge service, this is our best human loan officer.
• 25:07
  Speaker 2: They have AI backing, but they're also going to give you the best experience because they've been doing it so long and they understand so well.
• 25:13
  Speaker 2: You're gonna pay a little more for that.
• 25:14
  Speaker 2: What people want, choose your choose your pal.
• 25:16
  Speaker: That's fair.
• 25:17
  Speaker: I think we're seeing that too.
• 25:18
  Speaker: Like when we read our own reviews, if you go to our Google reviews and you see what people are saying, you'll notice our name a lot, Arjun and Heli, Arjun and Heli, Arjun and Heli.
• 25:27
  Speaker: Because like we are with them.
• 25:28
  Speaker: Or like, you want to talk, I'm here.
• 25:30
  Speaker: You want to text me, I'm available.
• 25:31
  Speaker: My AI agent is there to help you, but they they appreciate, they're like, wow, like I got a great rate.
• 25:36
  Speaker: I got like really quick service.
• 25:38
  Speaker: I was able to talk to someone if I really want to talk to someone, which happens once or twice in the whole loan process.
• 25:42
  Speaker: And that to them is like awesome.
• 25:44
  Speaker: And so we're like, great, we're designing that kind of experience because we see the product is not just the rate, right?
• 25:50
  Speaker: The rate is one component of it.
• 25:51
  Speaker: Um, it is a lot of other things as well.
• 25:54
  Speaker: It is like the education, it's the advice, it's the trust, it's the the speed to when you communicate with them and get back to them over text or phone.
• 26:01
  Speaker 2: Like those I spoke, I spoke with someone recently who who his view of the future loan officer is well, because you're freed up to do all this tedious stuff, yeah.
• 26:11
  Speaker 2: Step one is obviously building relationships and and doing more meaningful tasks.
• 26:15
  Speaker 2: But the even further step is the loan officer as the full home concierge of let me set up the moving van, let me help you with logistics terms, registering for schools, let me do all these things to add that much more value to the transaction to justify my sales cost because I don't have to do a lot of these tasks that I had to do before.
• 26:36
  Speaker 2: They can now be automated.
• 26:38
  Speaker 2: I can do more things to help my client to make them feel like holy smokes, I don't know, I didn't have to worry about anything.
• 26:43
  Speaker 2: And I think the future of mortgage, from my perspective, is like in one word, certainty.
• 26:49
  Speaker 2: How soon can you get to certainty?
• 26:51
  Speaker 2: Can I be certain of funding while I'm searching?
• 26:54
  Speaker 2: There's some really cool companies that are working on that.
• 26:56
  Speaker 2: You know, I want I want to be fully qualified for a mortgage while I'm searching on whatever platform, homes.com or Zillow or wherever.
• 27:02
  Speaker 2: I want to be certain of closing, and then I don't have to worry.
• 27:06
  Speaker 2: At what point can you bring some title and insurance and all other things forward in the origination process where I'm just good?
• 27:14
  Speaker 2: And I don't, it's not later, and oh, well, this issue came up and now it's back to the drawing board.
• 27:19
  Speaker 2: Like getting everything further forward in the process is where I think we're heading.
• 27:23
  Speaker 2: Yep.
• 27:23
  Speaker 2: That's a good point.
• 27:24
  Speaker 2: Great.
• 27:24
  Speaker 2: Robbie, it was great chatting with you.
• 27:26
  Speaker 2: Thank you for taking the time.
• 27:26
  Speaker 2: You too, Arjun.
• 27:27
  Speaker 2: We'll talk soon.
• 27:28
  Speaker 2: Bye.
• 27:28
  Speaker 2: Bye.
• 27:31
  Speaker 1: Today's economic calendar kicked off with mortgage applications from MBA, which fell 4.2% week over week for the weekending October 2nd, driven by an 8% decline in refinancing activity and a 2% drop in purchase applications, with both categories also down sharply year over year.
• 27:47
  Speaker 1: The pullback reflects renewed pressure from mortgage rates as the 30-year fixed rate climbed to 7.49%, its highest level in nearly three years amid rising treasure yields, wider spreads, and greater rate volatility.
• 28:01
  Speaker 1: Later today brings weekly crude oil inventories, the September FOMC minutes, August consumer credit, and $39 billion of 10-year Treasury note reopening results.
• 28:12
  Speaker 1: We begin the day with agency MBS prices worse than Tuesday's close by an eighth to a quarter, the two-year yielding 4.80, and the 10-year yielding 5.32 after closing yesterday at 5.27%.
• 28:22
  Speaker 1: Let's wrap up with a joke and some housekeeping.
• 28:28
  Speaker 1: This joke can be one gender about the other.
• 28:31
  Speaker 1: Don't take it as sexist depending on the way I say it.
• 28:33
  Speaker 1: So why do guys gain weight after marriage?
• 28:36
  Speaker 1: And see, that's better than if I said why do women gain weight after marriage?
• 28:39
  Speaker 1: Because when they're single, they come home, see what's in the fridge, and go to bed.
• 28:43
  Speaker 1: When they're married, they come home, see what's in the bed, and go to the fridge.
• 28:46
  Speaker 1: Thanks again to FlowFi for sponsoring this week's podcasts.
• 28:54
  Speaker 1: FlowFi is the mortgage industry's leading point of sale platform and dynamic apps, which can be seen at Booth 600 during MBA annual next week.
• 29:02
  Speaker 1: Let lenders create fully customizable loan applications for any loan type, including HELOCs, construction, agricultural lending, non-QM, and more without custom development.
• 29:13
  Speaker 1: To learn more, visit Flowify.com.