Latest / Chrisman Commentary - Daily Mortgage News / 7.23.26 Industry Consolidation; JazzX’s Varant Herculian on Change Management; Oil to Bonds

Transcript

7.23.26 Industry Consolidation; JazzX’s Varant Herculian on Change Management; Oil to Bonds · 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 more consolidation in the mortgage industry, why oil controls bonds in these dog days of summer, and my interview with Jazz X's, Varant Herculian on an effective change management and a clear strategy for integrating AI into workflows in ways that empower employees and deliver measurable business outcomes.
• 0:27
  Speaker 1: Here, take a listen to a little preview.
• 0:30
  Speaker 1: When we think about adopting technology in the mortgage industry or organizational change, best practices out there, common pitfalls, just things you you think companies do do well or do poorly that might be worth bringing to people's attention today.
• 0:48
  Speaker: One of the biggest pitfalls in just executing and getting it out into production is that getting stuck in a design phase.
• 0:55
  Speaker: Just to walk you through it in sequence of how it typically happens is uh first you got to like align and mobilize your change team, right?
• 1:03
  Speaker: These are the people that are gonna infect that change.
• 1:04
  Speaker: And then they're gonna go and sit in a conference room with tons of mortgage experience and really bright and smart people and whiteboard every single permutation on how they're gonna deploy this technology before they even intimately understand what it is, what it can do, what its capabilities are, and how it's gonna fit into their organization.
• 1:21
  Speaker: That's an impossible task to begin with, right?
• 1:24
  Speaker: Plus, you're gonna get stuck in that room for five to six months uh trying to get that perfect outcome really dialed in.
• 1:30
  Speaker: Whereas those same people can probably get you 80% of that future state operating model in like 30 minutes to an hour because they know the business so well.
• 1:37
  Speaker: It's that last mile that's so hard to get to.
• 1:41
  Speaker: And so, you know, to avoid getting stuck in that pitfall uh of just never executing and constantly being in design mode trying to figure out how to stay safe and safeguard everything to the umpteenth degree, getting a real live loan in production with real people holding the hand all the way from cradle is gonna accomplish much more in the first five loans than anything you can do in five months in a conference room setting, just by the nature of our industry and how mortgages work and operate.
• 2:14
  Speaker 1: Looking for the first true end-to-end AI platform built for mortgage?
• 2:17
  Speaker 1: Look no further than Jazz X.
• 2:19
  Speaker 1: From application to closing, Jazz X is a new operating model that helps you scale growth, boost productivity, and transform how your team performs.
• 2:28
  Speaker 1: Learn more at jazzx.ai.
• 2:30
  Speaker 1: Deals continue to happen as the big get bigger.
• 2:37
  Speaker 1: The latest example being Union Home buying Ameritrust to shoot for $20 billion a year.
• 2:42
  Speaker 1: Rocket closed on a multi-billion dollar credit agreement with JPMorgan Chase that will replace the facility it took out while two of its large acquisitions were pending last year.
• 2:52
  Speaker 1: And lawsuits are filed and are resolved, with the latest example being Nexa lending announcing the successful resolution of all litigation between CEO Mike Cortis and former business partner Matt Grella, bringing multi-year legal matters to a close and establishing Cortis as the sole owner of Nexa lending.
• 3:08
  Speaker 1: Shifting to rates, price down, rates up.
• 3:12
  Speaker 1: MBS and U.S.
• 3:13
  Speaker 1: Treasuries extended their price sell-off yesterday as rising oil prices fueled concerns that inflation could remain stubbornly persistent, pushing yields on two, three, five, and seven-year maturities to fresh yearly highs and leaving longer-term yields within striking distance of their peaks from earlier this year.
• 3:29
  Speaker 1: With little economic data to guide trading, a tepid $13 billion 20-year treasury reopening reinforced the cautious tone and kept pressure on the long end of the curve.
• 3:39
  Speaker 1: And speaking of that long end of the yield curve, the 30-year treasury yield has remained above 5% for its longest stretch since 2007, largely due to growing investor concerns over persistent inflation, worsening fiscal deficits, and the expanding supply of long-term debt.
• 3:55
  Speaker 1: Those elevated long bond yields come despite a Federal Reserve policy rate that is 150 basis points lower, suggesting investors are demanding a significantly larger premium to finance a treasury market that is nearly six times the size it was pre-great financial crisis.
• 4:10
  Speaker 1: Although core inflation measures continue to indicate a gradual moderation in underlying price pressures, policymakers are unlikely to interpret a single favorable inflation report as conclusive evidence that inflation has been sustainably contained.
• 4:23
  Speaker 1: The Federal Reserve is widely expected to leave its policy rate unchanged at next week's meeting while maintaining a data-dependent stance, with the July and August inflation reports likely to play a decisive role in determining the policy outlook for September.
• 4:37
  Speaker 1: Renewed tariff threats against Canada and persistent geopolitical uncertainty have had only a limited impact on treasury yields, suggesting that financial markets remain characterized by modest bearish sentiment, but limited conviction.
• 4:51
  Speaker 1: With the Federal Reserve and its pre-meeting communications blackout period, subdued summer trading volumes, and a sparse economic calendar this week increase the likelihood that technical positioning and market flows are exerting a greater influence on near-term price action.
• 5:05
  Speaker 1: Mortgage production has rebounded meaningfully from the post-QE refinancing drought, with first half 2026 gross agency MBS issuance up 28% year over year.
• 5:16
  Speaker 1: The increase has been driven largely by a sharp recovery in refinance activity, although higher mortgage rates following the US-Iran conflict are expected to slow prepayments and temper issuance in the months ahead.
• 5:27
  Speaker 1: Today's market differs significantly from 2023 because a meaningful share of borrowers still retain refinancing incentive, keeping activity well above the cycle lows despite remaining below pre-pandemic norms.
• 5:41
  Speaker 1: As a result, agency supply is expected to stabilize rather than revisit prior troughs, though prolonged geopolitical uncertainty and higher inflation weigh on both home purchases and refinancing, with every 25 basis point increase in mortgage rates estimated to reduce monthly mortgage production by roughly $10 billion.
• 5:59
  Speaker 1: For today's interview, we wanted to welcome to the show Jazz X's Veront Herculean to talk about effective change management and the clear strategy for integrating AI into workflows in ways that empower employees and deliver measurable business outcomes.
• 6:15
  Speaker 1: He has nearly 30 years of experience in the mortgage industry, spanning lending operations, product leadership, and customer success.
• 6:22
  Speaker 1: His focus has been on leading customer success organizations for leading mortgage fintech companies, helping lenders accelerate technology adoption, navigate organizational change, and achieve measurable business outcomes across the full customer life cycle.
• 6:37
  Speaker 1: A lot of people out there are talking about AI adoption, but I had heard you say that the biggest challenge isn't the technology itself.
• 6:48
  Speaker 1: What's actually causing so many AI initiatives to stall?
• 6:53
  Speaker: Yeah, no, that's a great question.
• 6:54
  Speaker: Like the technology itself isn't the problem, right?
• 6:57
  Speaker: There's a ton of AI solutions out there that can skin the cat a thousand different ways.
• 7:01
  Speaker: But it's really the way the organizations plan and govern and operationalize that change.
• 7:08
  Speaker: Too many lenders are really trying to like remove every uncertainty before they begin as part of the design phase and that future state operating model that they can't even begin to conceive or imagine quite yet.
• 7:20
  Speaker: But AI is something that you really understand by using it in a controlled production environment type of launch.
• 7:27
  Speaker 1: You've said AI requires a completely different approach to change management than previous technology rollouts.
• 7:33
  Speaker 1: What makes it so different?
• 7:36
  Speaker: So, I mean, just by definition, it's different, right?
• 7:38
  Speaker: It's not a deterministic solution and uh a technology that you're gonna go ahead and um, you know, test, deploy, train, uh, and rinse and repeat over and over again, right?
• 7:48
  Speaker: AI is gonna learn and adjust based on context, whereas the traditional software really just changes how people interact with a system.
• 7:55
  Speaker: Uh, the AI changes how the decisions work and distributed, or the work itself is actually distributed amongst people and technology, right?
• 8:02
  Speaker: It's kind of bifurcated between the two, and it can um really not just change what we do, but how we duck how we do it, who does it, and who owns those outcomes.
• 8:11
  Speaker: So, you know, by determining that, interpreting that context, working through the ambiguity and making some recommendations and through improved feedback, that's what change management, you know, can't end at training.
• 8:22
  Speaker: That's just really the beginning of it.
• 8:23
  Speaker: It goes on to perpetuity, and you have to think of it more as a continuous part of operating the business.
• 8:29
  Speaker 1: One idea Jazz X has been talking about is an agile production launch.
• 8:35
  Speaker 1: What does that mean?
• 8:36
  Speaker 1: And what does it actually look like inside a mortgage lender?
• 8:39
  Speaker: An agile production launch is kind of what I was describing there, really.
• 8:42
  Speaker: Is you know, instead of trying to design the future state, the training plan, the communication plans in day one, uh, ahead of having any kind of learning happening in production, um, that's that's really the whole preface of an agile production launch, is getting that first loan and that first user in a narrow scope use case, right?
• 9:03
  Speaker: Don't try to boil the ocean on day one.
• 9:05
  Speaker: Put a vanilla deal in there because no deal is truly vanilla.
• 9:07
  Speaker: And the high fidelity learning that's gonna happen with that human oversight and control on that production learning opportunity is gonna yield so much more benefit than anything else that you can possibly do.
• 9:20
  Speaker: So the idea is get it into a production launch, one loan, one small team, the smallest team possible, and apply some agile methodologies to it so that you have a weekly cadence to observe, um, make changes, improve, enhance your processes, identify where the handoffs are, where the gaps are.
• 9:37
  Speaker: Uh, there could be additional options that you didn't conceive during your initial design phase that you've learned.
• 9:44
  Speaker: Document those behaviors, those testimonials of how that user's or that person's life has changed, it improved, any friction points that you need to address, and rinse and repeat and make those small changes on a fast, iterative approach is going to yield much larger incomes incrementally over the next 30, 60, 90 days than any kind of whiteboard planning session is going to do.
• 10:07
  Speaker: So that's really the premise behind it, is similar to like a pilot that you might consider that's a little bit more rigid and more of a one and done and then a big bang scale, is to be able to do this incrementally, add additional use cases, add additional users, expand and scale based on proof that you're actually hitting your value metrics and improving business outcomes instead of just theory at the end of the day.
• 10:30
  Speaker: And that's going to get you to where you need to be.
• 10:32
  Speaker: Well, it does kind of beg the question.
• 10:34
  Speaker 1: What's it like working at Jazz X?
• 10:36
  Speaker 1: What do you enjoy about your role, trials and tribulations?
• 10:39
  Speaker 1: Take us, take us inside uh the company a little bit here.
• 10:43
  Speaker: Well, I'll tell you, we eat our own dog food.
• 10:45
  Speaker: We we live and die by technology, we leverage AI, we deploy agile methodologies, we we think fast, learn fast, we execute.
• 10:54
  Speaker: And I I gotta say, I've just been impressed with the pedigree and the level of talent in this organization has been unmatched in anything I've been part of in the past 30 years of of my career.
• 11:05
  Speaker 1: AI naturally raises concerns among operation teams.
• 11:09
  Speaker 1: I'm sure that's something that you've dealt with on the implementation side here.
• 11:13
  Speaker 1: How do lenders build trust with employees who worry AI will change or even replace their jobs?
• 11:19
  Speaker 1: And I would add to that, fortunately, it seems like we're we're past some of the discussion, just kind of the blind binary worries that AI is going to take over and I'm going to be out of a job.
• 11:27
  Speaker 1: It's it's meant to uh enhance productivity for a lot of people.
• 11:30
  Speaker 1: But uh, how how do you build trust with the with those employees who who are still worried?
• 11:36
  Speaker: Yeah, the answer is simple.
• 11:37
  Speaker: It's just hard to have the discipline to execute on it, right?
• 11:40
  Speaker: It's just like um, but what I need to do to lose weight, I exercise, sleep well, don't binge drink on the weekends.
• 11:45
  Speaker: Um, same thing with building trust with your employees when deploying AI is you you have to you have to communicate and you have to be transparent in your communications.
• 11:55
  Speaker: Uh, you don't need to placate people and give vague and ambiguous uh answers that, hey, nothing's gonna change.
• 12:00
  Speaker: This is all for the greater good.
• 12:02
  Speaker: They're gonna read between the lines and then they know that you know the cost comes through for right now.
• 12:07
  Speaker: And if we want to survive and compete in this market, we have to reduce that friction.
• 12:11
  Speaker: And so you have to think through what your vision, mission, statement is and what you're looking to accomplish.
• 12:16
  Speaker: Be honest and truthful with the employees that look, life is gonna change on the other side of this, our work is gonna shift.
• 12:23
  Speaker: There might be um, you know, less of a specific role in existence in the future state operating model, but new roles are gonna evolve as a result of it.
• 12:30
  Speaker: And the idea is to do more with the resources that we have to scale without having to add linear headcount to provide a better experience for our borrowers, our customers, and our employees alike.
• 12:43
  Speaker: And you gotta be, you know, able to draft a compelling message.
• 12:48
  Speaker 1: So we touched the tip of the iceberg a little bit in that preview, but I want to ask what separates lenders who are succeeding with AI today from those still stuck in pilot mode.
• 13:00
  Speaker: You know, a lot of that comes down to uh execution uh at the end of the day.
• 13:05
  Speaker: It's we found that the lenders that have, you know, treat this with the urgency and care that deserves.
• 13:11
  Speaker: They have a true executive sponsor that owns the business results, right?
• 13:15
  Speaker: And they're not just measuring to adoption and checking a box.
• 13:18
  Speaker: Um, they're really held accountable for improving something important, the business outcomes, you know, reducing cycle times, cost per loan, capacity, pull-through, borrower experience.
• 13:26
  Speaker: Um track those metrics, have the teletry in pace in place, um, get your baselines there and be truly invested in every part of the deployment of this, down to the people, the communication, the training, the deployment, and part of this um agile deployment philosophy.
• 13:43
  Speaker: Got to roll your sleeves up and make sure that you're willing to be agile in your thought process as well and pivot as you see more opportunity throughout the deployment and capitalize on that and stay steadfast.
• 13:55
  Speaker: Have your North Star metrics in place and follow through on it.
• 13:58
  Speaker 1: And finally, before I let you go, what is the roadmap for JazzX?
• 14:04
  Speaker 1: Any any cool product releases upcoming?
• 14:07
  Speaker 1: Can it give us give us a little preview of what's to come from the company?
• 14:12
  Speaker: We got a lot of great things in the works right now at Jazx AI.
• 14:16
  Speaker: And uh, you know, it's unlike anything that you've seen in the mortgage industry before.
• 14:19
  Speaker: It's truly an end-to-end enterprise AI solution, taking you all the way from the point of sale uh through fulfillment, post-closing, underwriting, and all the way down to servicing.
• 14:30
  Speaker: And what's truly amazing about it is just how you can institutionalize a knowledge and a knowledge fabric, that these digital workers and agents can have tunable autonomy so that it can have a level of oversight uh depending on your preference and your level of trust as you kind of validate and deploy and scale based on evidence.
• 14:50
  Speaker: Uh, and then what all our you know, compliance and secondary and post-closing people love to hear about is that uh it's that auditable, traceable decisions fabric that's being done there that ties everything back to a policy source document uh and the explainability of made the judgment call that it is.
• 15:10
  Speaker 1: Ron, I really appreciate the time, man.
• 15:11
  Speaker 1: Ton of cool stuff here for listeners.
• 15:13
  Speaker 1: And uh I'm personally rooting for JazzX to do some really revolutionary things in the industry.
• 15:17
  Speaker 1: It sounds like y'all are off to a great start, and hopefully there's more to come.
• 15:20
  Speaker 1: So thank you.
• 15:21
  Speaker: Awesome.
• 15:22
  Speaker: Thank you, Robbie.
• 15:22
  Speaker: Appreciate you.
• 15:26
  Speaker 1: Today's economic calendar kicked off with weekly jobless claims, which came in, at least for initial claims, at the very low number of 187,000.
• 15:34
  Speaker 1: Continuing claims were 1.796 million.
• 15:38
  Speaker 1: Later today brings July, Casey Fed manufacturing and composite indices, as well as a 10-year tips auction.
• 15:44
  Speaker 1: After overnight Iranian war news and those jobless claims figures, agency MBS prices are worse about an eighth, the two years yielding 4.32, and the 10 years yielding 4.69 after closing yesterday at 4.66%.
• 15:59
  Speaker 1: Let's wrap up with a joke and some housekeeping.
• 16:04
  Speaker 1: What do you call an unemployed Rastafarian?
• 16:07
  Speaker 1: Job less.
• 16:09
  Speaker 1: Looking for the first true end-to-end AI platform built for mortgage?
• 16:18
  Speaker 1: Look no further than JazzX.
• 16:20
  Speaker 1: From application to closing, Jazz X is a new operating model that helps you scale growth, boost productivity, and transform how your team performs.
• 16:28
  Speaker 1: Learn more at jazzx.ai.