Latest / Chrisman Commentary - Daily Mortgage News / 9.2.26 Nuances of Affordability; Lender Toolkit's Brett Brumley on Technology; Rates Not Friendly

Transcript

9.2.26 Nuances of Affordability; Lender Toolkit's Brett Brumley on Technology; Rates Not Friendly · Chrisman Commentary - Daily Mortgage News

• 0:02
  Speaker: Welcome to the Chrisman Commentary, Daily Mortgage News Podcast.
• 0:05
  Speaker: I'm your host, Robbie Chrisman.
• 0:08
  Speaker: Topics on today's episode include some of the nuances of home affordability.
• 0:13
  Speaker: Rates are not our friends right now.
• 0:15
  Speaker: I'll tell you why.
• 0:17
  Speaker: And my interview with Lender Toolkits, Brett Brumley, on the proper due diligence necessary with counterparties in the artificial intelligence era.
• 0:23
  Speaker: Here, take a listen, do a little preview.
• 0:26
  Speaker: There's all these artificial intelligence companies popping up in the mortgage industry.
• 0:31
  Speaker: And you'll see them online, you'll see them at booths, at some of the conferences.
• 0:36
  Speaker: If I'm an originator out there, a lender, and I want to kick the tires on these companies and actually figure out real competitive differentiators beyond what people say the product can do.
• 0:45
  Speaker: What are some of the best ways to figure out what company is going to work best with my organization?
• 0:50
  Speaker 1: I don't envy lenders today.
• 0:51
  Speaker 1: They're getting hit up by new companies that have come out of the woodworks on a regular basis.
• 0:56
  Speaker 1: A story, if you'll indulge me.
• 0:58
  Speaker 1: You know, we sponsored a local university hackathon.
• 1:01
  Speaker 1: Um, and I put a bounty on a mortgage-related product to see what kind of college students can do.
• 1:06
  Speaker 1: And in about 30 days, some of the winner actually only took two weeks.
• 1:10
  Speaker 1: They put together what appears to be a very fully agentic system that if a loan officer uses it, they're like, wow, this is powerful.
• 1:18
  Speaker 1: But when you peer behind the covers, even a little bit, it is not scalable, it is not SOC certified.
• 1:24
  Speaker 1: Um, it doesn't have an industry standard like ISO 42001, the new AI governance standard.
• 1:30
  Speaker 1: You know, there's a couple of college students that built something in three weeks, and you as a lender might not know who's supporting you or backing you.
• 1:37
  Speaker 1: So it's okay to use things, it's okay to test things.
• 1:40
  Speaker 1: But when you talk about enterprise change, you also need to understand who you're working with, what their experience level is, how many customers can they have?
• 1:48
  Speaker 1: 10 customers can overwhelm a company, and the next thing you know, they're not supporting your product.
• 1:52
  Speaker 1: So I think it's really about who do you want to work with?
• 1:55
  Speaker 1: What's their credit credentials in the mortgage space, not just Silicon Valley?
• 1:59
  Speaker 1: How well do they understand mortgage compared to how well the AI is just understanding mortgage?
• 2:05
  Speaker 1: Um, AI misses a ton of things.
• 2:07
  Speaker 1: Um, it's powerful, but it needs to be backed with some more deterministic outcomes that are programmed in by mortgage professionals, not just letting the AI do it.
• 2:15
  Speaker 1: Um, but you know, I would say test it.
• 2:17
  Speaker 1: Be a little bit experimental.
• 2:18
  Speaker 1: It's okay.
• 2:19
  Speaker 1: You can change things in 30 days.
• 2:21
  Speaker 1: It's not the end of the world if something doesn't work out, have it.
• 2:25
  Speaker 1: So a company that's willing to let you have that conversation and test something a little bit and work directly with you, I think you're going to be far better off than just jumping with the latest, greatest thing that might be totally different in six months.
• 2:39
  Speaker: Thanks to Zillow Home Loans, Zillow's in-house mortgage lender, for sponsoring this week's podcasts.
• 2:44
  Speaker: By integrating Zillow's real estate platform with financing, Zillow Home Loans helps buyers move from dreaming about a home to holding the keys.
• 2:52
  Speaker: With tools built for modern lending, Zillow Home Loans loan officers can focus on guiding buyers with care and confidence.
• 2:58
  Speaker: Zillow Home Loans is an equal housing lender.
• 3:01
  Speaker: If you're bored at work, print someone else's resignation letter and leave it in the printer.
• 3:08
  Speaker: This rate and affordability environment is tough, and any lender who finds these higher rates boring, well, don't hold your breath waiting for them to come down.
• 3:17
  Speaker: Recently, President Trump addressed intervening in the bond market and brought up the military.
• 3:23
  Speaker: Principal interest taxes and insurance is only part of the affordability equation, but market forces determine long-term rates, not the government or the Fed.
• 3:33
  Speaker: Some estimates have property taxes and insurance up twenty to thirty percent over the last year.
• 3:38
  Speaker: Condo special assessments.
• 3:40
  Speaker: I saw a story about some condo owners being hit with a five million dollar emergency roof bill.
• 3:46
  Speaker: What does that do to an owner's financial situation?
• 3:49
  Speaker: Anyways, the United States is not alone here in this global sell-off in government bonds.
• 3:56
  Speaker: It's pushed borrowing costs around the globe and some of the world's largest economies to the highest levels in decades.
• 4:02
  Speaker: Mortgages and other debt are being impacted, which in turn hits consumers who borrow.
• 4:07
  Speaker: A combination of factors is prompting investors to demand higher returns to hold government debt.
• 4:12
  Speaker: A flood of borrowing by the world's richest nations, expanding budget deficits, persistent inflation, and a few signs that countries are able or willing to take steps to improve these conditions.
• 4:22
  Speaker: The yield on 10-year U.S.
• 4:24
  Speaker: Treasury notes, perhaps the world's most influential interest rate, reached its highest level since January 2025 of 4.8%, and the yield on 30-year bonds continued to hover around a two-decade high.
• 4:37
  Speaker: The increase in rates has set off a battle between Treasury Secretary Scott Besson and investors.
• 4:42
  Speaker: But the factors pushing up bond yields in the United States are also issues in other big markets.
• 4:48
  Speaker: And in my experience, the markets will always win.
• 4:51
  Speaker: In terms of supply and demand, investors have options.
• 4:54
  Speaker: There are lots of companies that don't have a deficit of forty trillion dollars like the US.
• 4:59
  Speaker: A borrowing binge by technology companies to build artificial intelligence systems is another factor pushing up the costs of all types of debt.
• 5:08
  Speaker: Companies have issued billions of dollars in bonds, swamping markets and pulling investors away from government debt.
• 5:14
  Speaker: These companies, known as hyperscalers, are also increasingly turning to Euro denominated bonds.
• 5:20
  Speaker: One of the most pressing and unpredictable drivers of bond market turmoil is the protracted war in Iran.
• 5:26
  Speaker: As the United States and Iran renewed attacks recently, the price of oil and natural gas began to climb again.
• 5:32
  Speaker: Brent crude, the international oil benchmark, rose on Tuesday above $94 a barrel, about 30% higher than pre-war levels.
• 5:41
  Speaker: We are reminded of this every time we fill up our gas tanks.
• 5:44
  Speaker: The prices of refined fuels like gasoline and diesel have risen even faster, increasing expectations of higher inflation that could prompt central banks to raise short-term interest rates.
• 5:54
  Speaker: And yes, the US's gross national debt topped forty trillion dollars for the first time last month or more than 120% the size of our economy.
• 6:03
  Speaker: France's public debt is 117% the size of its economy.
• 6:07
  Speaker: In Japan, it's twice the size of its economy.
• 6:10
  Speaker: And many politicians on both sides of the aisle don't appear worried enough about these debt levels to do anything.
• 6:16
  Speaker: Instead, investors see government plans that are not likely to shrink budget deficits.
• 6:21
  Speaker: With expectations of more borrowing to come, investors are demanding higher returns to hold government bonds.
• 6:27
  Speaker: And there's little reason for rates to go down.
• 6:30
  Speaker: For today's interview, I wanted to welcome back to the show Lender Toolkit's Brett Brumley to talk about the proper due diligence necessary with counterparties in the artificial intelligence era, as well as designing technology for specific use cases.
• 6:45
  Speaker: He's the founder and CEO of Lender Toolkit, a trusted technology used by over 400 mortgage lenders.
• 6:53
  Speaker: He's been a pioneer in the field of artificial intelligence and automation and mortgage lending long before the topic became du jour in the mortgage industry.
• 7:03
  Speaker: He conceived and built a company that seamlessly integrates AI into the underwriting process to help lenders save time and money and has created tools for effortlessly automating and streamlining many mortgage tasks from origination to underwriting to the secondary market.
• 7:16
  Speaker: When we hear that a there's a mortgage AI company and they built some AI model, does that just mean, and I'm so sorry to pull the lid back on much of the industry here, but does that just mean that they've gone on anthropic or open AI and trained something and now they're presenting it to clients and ooh, it's a multi-million dollar idea because of the model we've trained, or are companies actually building their own proprietary AI systems?
• 7:42
  Speaker: Can you bring some clarity to this for me?
• 7:45
  Speaker 1: Sure.
• 7:45
  Speaker 1: Yeah, I mean, obviously a really good question and does pull the lid back a little bit on some folks.
• 7:50
  Speaker 1: No doubt there are those companies in the market that are training specific models for a specific purpose.
• 7:55
  Speaker 1: But I think the training the model concept kind of gets conflated with what's really happening compared to the need for training.
• 8:02
  Speaker 1: When you train a model, you're you're really trying to teach it very, very specific items and have the model it surface those items up in a more cohesive way or give you more definitive answers.
• 8:12
  Speaker 1: And you've really kind of perfected that answer.
• 8:15
  Speaker 1: Does that always require a train model?
• 8:18
  Speaker 1: No.
• 8:18
  Speaker 1: Um the generalized models from the large providers are getting better every single time.
• 8:23
  Speaker 1: So, you know, well, you can train it.
• 8:25
  Speaker 1: Uh, there's not always a need to train them.
• 8:28
  Speaker 1: So I would say there's plenty of companies that are out there that are just putting a wrapper around a generalized enthropic model or GPT model or otherwise.
• 8:36
  Speaker 1: That's not the value, right?
• 8:37
  Speaker 1: You can get a lot of that information out of just Chat GPT directly.
• 8:40
  Speaker 1: You can you can feed it information.
• 8:42
  Speaker 1: The value is what you do with that information once you get it, um, and how you're doing tracing, durability, observability.
• 8:49
  Speaker 1: You know, can you actually work through what that model is doing and make sure it's doing it correctly?
• 8:54
  Speaker 1: That's the hard part.
• 8:56
  Speaker 1: That adding a model is being done by plenty of people, and there's a lot of rappers out there.
• 9:00
  Speaker 1: So, you know, it's definitely something to be aware of, how the company's using AI beyond just what model they're doing and if they've trained it or not.
• 9:08
  Speaker: When we look at the tech landscape here in the third quarter of 2026 in the mortgage industry, how would you categorize it broadly?
• 9:17
  Speaker: And at least from my perspective, frequent listeners of this podcast will have heard me say this, but we've moved from here's what the technology can do to here's what it actually is doing.
• 9:27
  Speaker: And I would say that that's no longer necessarily the leading edge anymore.
• 9:32
  Speaker: The leading edge is new and innovative ways of problem solving.
• 9:38
  Speaker: Maybe it's maybe it's still in the agentic space.
• 9:40
  Speaker: But from your perspective, where is the mortgage industry in the third quarter when it comes to AI technology or technology and and what is pushing the envelope forward?
• 9:48
  Speaker 1: Yeah well, the velocity is real and it's definitely accelerating.
• 9:51
  Speaker 1: This year has been um probably more innovation from AI than all the years combined.
• 9:58
  Speaker 1: And I think we're gonna continue to see that trend.
• 10:00
  Speaker 1: A lot of startups are shipping kind of good enough solutions that work maybe directly within you know the largest LOS platforms, maybe just work with the data.
• 10:08
  Speaker 1: Um, and they can ship them pretty quick.
• 10:10
  Speaker 1: Uh, the gap between early and late adopters is widening fast.
• 10:14
  Speaker 1: Stratmore came out with a study recently uh that said something like 38% of lenders are already using 80 uh uh AI, and over 88% are planning to adopt it right now.
• 10:25
  Speaker 1: Nobody's pulling back from it.
• 10:27
  Speaker 1: So the models could do, or the AI that's available today can probably do 80% of it really well.
• 10:32
  Speaker 1: And that's great, you know, for the mortgage industry.
• 10:34
  Speaker 1: But it's that 20% the mortgage runs on.
• 10:36
  Speaker 1: Very, very specific how mortgage lending works, and these models just don't know all the edge cases.
• 10:41
  Speaker 1: So are you prepared to have multiple workflows?
• 10:44
  Speaker 1: Maybe.
• 10:45
  Speaker 1: But no matter what, the uh answer to your question is it's accelerating and accelerating fast.
• 10:50
  Speaker: Well, I was gonna say, how do we close that 20% gap?
• 10:52
  Speaker: Before you answer that, I was gonna say, what's the roadmap for where things go from here?
• 10:56
  Speaker: But maybe the better question is how do we get it from 80 to 95 or 99 or 100?
• 11:02
  Speaker 1: Yeah.
• 11:03
  Speaker 1: Well, most of what people call AI and mortgage today is completely generic.
• 11:06
  Speaker 1: You know, the really specialized stuff, that 20% you're talking about, you know, that's really built around the workflow of a given lender.
• 11:12
  Speaker 1: What works for one lender might not work for the other lender.
• 11:15
  Speaker 1: Um, what is the same as data?
• 11:17
  Speaker 1: So if you've got really detailed, documented processes, procedures, workflow for those exceptions, then AI is going to be better suited to consume that information than if you just hope a model can figure it out.
• 11:29
  Speaker 1: So really building in those use cases into a given product and having the AI solve the 20% is what everybody's really focusing in their energy on.
• 11:38
  Speaker 1: Um, you know, our company handles a lot of the 80 uh 20%, the last 5%, really challenging from an ROI perspective.
• 11:45
  Speaker 1: And uh, you know, do you even have enough use cases to demonstrate what it should probably be doing?
• 11:50
  Speaker 1: So, you know, just documentation.
• 11:51
  Speaker 1: The more you document, the more data you have available, the better AI is going to be able to handle that.
• 11:58
  Speaker: I consider you one of the foremost technologists in the mortgage industry.
• 12:02
  Speaker: And I want to ask about designing technology.
• 12:06
  Speaker: And this is such a broad question that I apologize.
• 12:09
  Speaker: But when you think about designing technology for practical use cases versus a utopian vision of what things can be versus what will actually get implemented and adopted and used uh throughout an organization, how do you come to kind of that middle ground of all of that?
• 12:27
  Speaker 1: Well, first, thanks for the compliment.
• 12:28
  Speaker 1: I appreciate that.
• 12:29
  Speaker 1: I've always respected you and your dad and what you guys are doing is is influential in the mortgage space in big, big ways.
• 12:36
  Speaker 1: So I appreciate the compliment from you guys, you especially.
• 12:39
  Speaker 1: Um, you know, agentic AI and mortgage, though, is really early.
• 12:42
  Speaker 1: Uh the pattern's clear though.
• 12:43
  Speaker 1: If you want AI to take like a multi-step workflow, you know, like get docs, validate the income, check the guidelines, flag the exceptions, run it through without a human clicking through every single step.
• 12:53
  Speaker 1: That's gonna really change how you build that compared to if you have that human in the loop.
• 12:58
  Speaker 1: Humans are where the creativity comes from.
• 13:02
  Speaker 1: And in fact, we're seeing something at AI where everybody's using the same models, everybody's building the same thing.
• 13:07
  Speaker 1: Eventually, in a little couple more cycles, it there is gonna be no more human input.
• 13:11
  Speaker 1: So the human is what brings that creativity, that exceptions, that highlights really what needs to build.
• 13:17
  Speaker 1: So if you want to get adoption, solve one challenge, one workflow.
• 13:21
  Speaker 1: Uh, don't try to boil the ocean and definitely don't focus all your energy on the exceptions.
• 13:26
  Speaker 1: If you want to get adoption, use it for the 80% and make sure your company is using it consistently and you can measure some form of ROI.
• 13:33
  Speaker 1: And I don't mean ROI from uh like how much time spent.
• 13:36
  Speaker 1: I mean like, are you using it?
• 13:38
  Speaker 1: Are you getting real world benefits from using the product?
• 13:41
  Speaker 1: And sometimes that's anecdotal and sometimes that's highly measurable.
• 13:44
  Speaker 1: Um, but the trick really is just break it down into some workflows and use it, just use it.
• 13:49
  Speaker 1: Um, and I think you'll get more and more adoption as people see the the value that it brings, and then they'll want more, and then you keep building on that uh that that momentum.
• 13:58
  Speaker: Yeah, you hit on an interesting point there, and that's the the humanization of things.
• 14:02
  Speaker: Because if we go back 20 or 30 years when you could call United Airlines and you get an automated thing, cool, my wait time has been cut down by 98%.
• 14:12
  Speaker: But after a while, you're like, I just want to talk to a human.
• 14:16
  Speaker: And so there's these pendulum swings that go on out there.
• 14:18
  Speaker: I see these people in the mortgage industry that are just trying to throw AI at everything, thinking, well, even if some of it doesn't really work out, I'm still an early adopter and I'm well ahead of the curve.
• 14:28
  Speaker: At the same time, you have people that I don't want to say they're in denial of it, but there's people are saying, I can't lose the humanity here.
• 14:36
  Speaker: And I just I want your thoughts on trying to have AI solve for everything versus where humans, and obviously you said the the creative side of it, but where people are going a little overboard with trying to throw AI at things and and where you think kind of that middle ground will be as we as we move into the future, considering things are moving so rapidly.
• 14:56
  Speaker 1: Oh, you know, it's an interesting question.
• 14:58
  Speaker 1: When you try to solve every single workflow, it feels like a daunting task.
• 15:02
  Speaker 1: And so, you know, you might buy a product from a given vendor or you might develop something yourself.
• 15:08
  Speaker 1: If you take the scope and you try to solve all of mortgage or all of conventional or all of the FHA, think of all the different steps and use cases you need to solve in that world.
• 15:18
  Speaker 1: So you don't need to solve everything.
• 15:21
  Speaker 1: There are things that it can do really, really well.
• 15:23
  Speaker 1: Like take a credit report.
• 15:24
  Speaker 1: A credit report is a standard document.
• 15:27
  Speaker 1: It looks the same every single time, other than the data, right?
• 15:30
  Speaker 1: It's very structured.
• 15:32
  Speaker 1: And you can have AI analyze a credit report.
• 15:34
  Speaker 1: You don't need your people looking at the credit report.
• 15:37
  Speaker 1: And the, you know, the expression checking checkers, checking the checkers, checking the checkers.
• 15:41
  Speaker 1: We kind of need, to some extent, as an industry, throw that model out the door.
• 15:45
  Speaker 1: I don't need an AI to check the credit report and then a human to double check the visual document of the credit report and then analyze the data on whether it matched.
• 15:53
  Speaker 1: You need to go through that exercise during implementation to make sure you trust the system.
• 15:58
  Speaker 1: But you don't do that in your day-to-day workflow.
• 16:00
  Speaker 1: That's just noise.
• 16:01
  Speaker 1: And it also doesn't give you any value in the product.
• 16:04
  Speaker 1: If you have AI analyzing a credit report and then you have an underwriter do the exact same thing, you added no value at all.
• 16:10
  Speaker 1: In fact, you have two steps instead of one.
• 16:12
  Speaker 1: Uh, multiple points of failure as well.
• 16:15
  Speaker 1: So I think really embrace what it can do and don't try to make it do things that it's not yet capable of doing.
• 16:21
  Speaker 1: And in that case, you're going to have the humans kind of guide that direction.
• 16:24
  Speaker 1: Things that are the same every time that are standard, that you can provide a very defined workflow, those are excellent for AI.
• 16:31
  Speaker 1: Things that have, you know, happened this one time on this one loan.
• 16:34
  Speaker 1: We want to make sure we never see again.
• 16:36
  Speaker 1: You might not even have enough use cases to tell AI what to do properly.
• 16:40
  Speaker 1: So don't focus on that.
• 16:41
  Speaker 1: Focus on the thing that's standard every time and then just automate it.
• 16:45
  Speaker: Well, I'm glad you brought up the word implementation, because implementation is where the rubber meets the road, the make or break for a lot of thoughts on best practices for what you've seen when it comes to successful implementations, when it comes to companies getting buy-in and adoption and all that, and then maybe a couple common pitfalls after that.
• 17:06
  Speaker 1: Yeah, implementation is a challenging thing.
• 17:08
  Speaker 1: It's change management, it's people change management.
• 17:10
  Speaker 1: I mean, there's been books and countless articles written on how to get people to change, and that's not the easiest thing in the world.
• 17:16
  Speaker 1: That is the hardest part of an implementation.
• 17:18
  Speaker 1: So it really does start out with leadership.
• 17:21
  Speaker 1: You know, I know we say that a lot, and everybody knows it starts from the executive suite down, but I can't tell you how actually true that is.
• 17:28
  Speaker 1: You know, we work with a lot of companies where the executive endorsed the product, but then as soon as they signed the contract, kind of disappeared from the day-to-day.
• 17:37
  Speaker 1: Well, they never really properly communicated to their downstream employees what the impact of this is, what they're expecting, how it'll impact their day-to-day, their jobs, how even to approach the problem.
• 17:47
  Speaker 1: They just said go implement the software.
• 17:49
  Speaker 1: Well, from that, you're gonna get very, very different outcomes.
• 17:54
  Speaker 1: You also have to assume that AI on the surface appears like it's trying to eliminate jobs.
• 17:59
  Speaker 1: And, you know, sometimes that's gonna happen.
• 18:01
  Speaker 1: And I don't think we should hide from it.
• 18:02
  Speaker 1: But if you're clear about your intention with your employees and they don't have fear going into implementation for their jobs, they're gonna be far more willing to embrace the technology and help the technology.
• 18:14
  Speaker 1: So that's one thing is adoption and communication.
• 18:17
  Speaker 1: Make sure you understand and people know what you're gonna do with it and how it can benefit them.
• 18:21
  Speaker 1: It's an organization, not just leave it out there for them to interpret.
• 18:24
  Speaker 1: A lot of people sometimes just go to the worst case scenario.
• 18:27
  Speaker 1: So communication is absolute key.
• 18:30
  Speaker 1: The second part I would say is don't try to build your existing process into new technology.
• 18:37
  Speaker 1: There was a reason you bought the technology in the first place.
• 18:40
  Speaker 1: If you try to force a software company, a vendor, or even a manual workflow to do it like it used to be, you're gonna have friction.
• 18:49
  Speaker 1: It's not always gonna work.
• 18:50
  Speaker 1: Frankly, use it the way it's supposed to be used for a little while, gain feedback on whether it's working well or whether it's not working, and then iterate from there.
• 18:59
  Speaker 1: Change it after you've used it.
• 19:01
  Speaker 1: Don't do that as a barrier that we won't use it until it's perfect, because then it'll never get done.
• 19:06
  Speaker 1: You know, mortgages, you know.
• 19:07
  Speaker 1: I mean, if there's one problem this week and another problem comes up next week, they kind of pivot sometimes to the next problem.
• 19:13
  Speaker 1: Same thing happens in software.
• 19:15
  Speaker 1: If you really need something to be fixed in software, put your energy behind it until it's resolved.
• 19:20
  Speaker 1: Don't just move on to the next topic.
• 19:22
  Speaker 1: So implementation is really about communication, focus, and change management, but also how to adopt and adapt to new software.
• 19:29
  Speaker 1: And all of that's needed for a successful implementation.
• 19:33
  Speaker: And I'd be remiss if I didn't ask about lender toolkit.
• 19:36
  Speaker: What is the latest from y'all?
• 19:38
  Speaker: What have you been working on?
• 19:39
  Speaker: What are you excited about and where are you headed from here?
• 19:42
  Speaker 1: Yeah.
• 19:42
  Speaker 1: Well, you know, last year we really focused on, you know, getting our um Prism underwriting product out, um, extending it into the origination service where loan officers, processors, assistants can use it.
• 19:53
  Speaker 1: Um, and it's been going really, really well.
• 19:55
  Speaker 1: Um does everything from income, assets, credit, underwriting, you know, even some fringe cases and has a lot of customization potential.
• 20:02
  Speaker 1: So we're thrilled about that.
• 20:03
  Speaker 1: We obviously want to uh get more adoption in the marketplace for technology that way.
• 20:07
  Speaker 1: But what we're working on next, I think, is is really about mortgage intelligence.
• 20:11
  Speaker 1: You know, how can we leverage all of the data we have to do better training, how to reach consumers in a more productive way, how to make how to report on automation, and then of course, how to meet all the regulatory explainability things that are coming down from Fannie Mae and the CFPB.
• 20:26
  Speaker 1: So it's a lot of work, but we're excited about it.
• 20:28
  Speaker 1: Um, to be honest, I don't think I've been as passionate about building mortgage-related solutions than I have um in years.
• 20:34
  Speaker 1: Uh, this has been really exciting.
• 20:37
  Speaker: Wise words.
• 20:37
  Speaker: You know I always love talking to you.
• 20:40
  Speaker: Hopefully, I'll see you on the road here somewhere soon.
• 20:43
  Speaker: And uh it had been too long since we did one of these, so I'm I'm glad we were able.
• 20:47
  Speaker 1: Yeah, thanks.
• 20:47
  Speaker 1: I mean, I'll be in digital mortgage in Vegas in uh September.
• 20:50
  Speaker 1: I think uh MISMO's coming up here in August.
• 20:52
  Speaker 1: Um, and then of course we got MBA Annual uh coming up in uh Chicago.
• 20:56
  Speaker 1: So I'll be at all of those, and I have no doubt you will as well.
• 20:59
  Speaker: Looking forward to it.
• 20:59
  Speaker: Thanks, Brett.
• 21:00
  Speaker 1: Thanks.
• 21:03
  Speaker: Today's economic calendar kicked off with mortgage applications from MBA, which rose 0.8% week over week for the weekending August 28th, driven by a 2% increase in seasonally adjusted purchase applications, while refinancing activity fell 1% and remained 19% below year ago levels.
• 21:20
  Speaker: Other releases today include August ADP employment, July factory orders, weekly crude oil inventories, and the September Fed beige book.
• 21:30
  Speaker: We begin the day with agent CMBS prices, little change from Tuesday's close, the two-year yielding 4.39, and the 10-year yielding 4.79 after closing yesterday at 4.80%.
• 21:39
  Speaker: Let's wrap up with a joke and some housekeeping.
• 21:45
  Speaker: At one point during a game, the coach called one of his seven-year-old football players aside and asked, Do you understand what cooperation is?
• 21:52
  Speaker: What a team is?
• 21:53
  Speaker: The little boy nodded in the affirmative.
• 21:55
  Speaker: Do you understand what matters is not whether we win or lose, but how we play together as a team?
• 22:00
  Speaker: The little boy nodded yes.
• 22:02
  Speaker: So, the coach continued, I'm sure you know when a penalty is called, you shouldn't argue, curse, attack the referee, or call him a peckerhead.
• 22:09
  Speaker: Do you understand all that?
• 22:10
  Speaker: Again, the little boy nodded.
• 22:13
  Speaker: And when I call you off the field so that another boy gets a chance to play, it's not good sportsmanship to call your coach a dumb a-hole, is it?
• 22:20
  Speaker: Again, the little boy nodded.
• 22:22
  Speaker: Good.
• 22:24
  Speaker: Now go over there and explain all that to your mother.
• 22:27
  Speaker: Thanks again to this week's podcast sponsor, Zillow Home Loans.
• 22:34
  Speaker: Zillow's in-house mortgage lender.
• 22:36
  Speaker: With tools built for modern mortgage lending, Zillow Home Loans loan officers can focus on guiding buyers with care and confidence.
• 22:43
  Speaker: To learn more, visit Zillow.com/slash home loans.