“Hey, it’s either network or no work.” For lenders, networking is an important part of their business. In addition, generally speaking, renters are prime “feeding grounds” for loan originators searching for clients. But there’s some disturbing signs out there: what if renters can’t even afford their rent? I was talking to a successful LO recently who uttered, “50 calls, 5 leads, 2 applications, 1 closing. The next day, 50 calls, 5 leads, 2 applications, 1 closing. Rinse and repeat, every day, it’s a pyramid.” Being an originator is a numbers game, as is running a branch. A manager recently told me, “Manage to the numbers. It doesn’t matter how long your team has been with you.” You can be cheap, easy, or fast. Pick one. You’re not going to be all three consistently. Meanwhile, LOs faced with keeping up with technology. Recently I received, “I stopped asking, ‘Will AI replace me’ and started asking, ‘How can AI make me better?’” And I regularly receive emails to the effect of saying, “I am not competing against other LOs. I am competing against Google, YouTube, TikTok, AI search, Reddit, ChatGPT, and (Today’s podcast can be found here. This week’s ‘casts are presented by Spring EQ, the clear choice in home equity and non-QM solutions. Since 2016, Spring EQ has helped more than 160,000 homeowners access over $16 billion in equity. Today’s has an interview with Spring EQ’s Reno Heine on actively navigating a volatile mortgage market by expanding into non-QM offerings, evolving its third-party origination strategies to meet modern broker needs, and advancing a new digital platform to capture future growth opportunities.)

AI Is Coming. But What Will It Learn From?

Cotality CEO Pat Dodd likes to draw the distinction between speed and velocity. Velocity, he says, “Is the combination of speed and intentional direction. Our industry’s focus on speed alone is the reason why we are in a paradox today. Everything (credit pulls, verifications, AVMs, etc.) can be done in seconds, but originating a mortgage still takes ~45 days. That’s because the mortgage industry has created islands of instant in a sea of disconnected manual and complex processes.

Cotality unlocks real velocity by partnering with clients and key tech providers, like Google, Databricks, and ICE, to reimagine how insights are delivered. Cotality is developing solutions that will make our industry faster, smarter, and more people centric.

“The 4 Cs of lending (credit, capital, collateral, and capacity) aren’t going away. So where will AI go to learn them? Today, we are the industry’s go-to source for fiduciary-grade lending data and solutions. Our clients rely on us for credit reports (Credco) home price performance (Cotality Home Price Index), appraisal QC (CMS and Mercury Network), income verification, flood certs, and fraud prevention (LoanSafe). We also collect information and pay property taxes on 49 million homes. Cotality is where the mortgage industry goes for the data behind the 4 Cs and where AI agents will go to find the most current, accurate, and defensible data. When they get there, they’ll find our data is AI-ready and seamlessly accessible on advanced MCP servers.

“Prospecting, underwriting & quality control are good places to start. The average lender spends between $1,200 and $2,000 to land one loan, adding up to billions industry wide. Today, Cotality leverages advanced machine learning and predictive data to power sophisticated Propensity Score models and Market Intelligence on Araya. These active solutions monitor critical behavioral and property signals, such as rate spreads, equity positions, and market trends, to help originators, servicers, and MSR holders pinpoint high-potential opportunities. Instead of blindly chasing leads, your loan officers can act with precision and certainty ahead of the competition, driving superior acquisition, retention, and portfolio growth.

“The clock starts ticking as soon as you have a deal, and for the average loan it ticks 3,888,000 times (or 45 days) before the loan closes. Why? Because some information is instantaneous and in a lender’s control, like credit and flood certs, and some isn’t. There are delays waiting for the borrower to send and resend information, then there’s the appraisal (4 or 5 days), an appraisal review, and possibly more days if a rework is needed. What if you could connect all of the islands of instant to accelerate the key processes that comprise residential lending? One way to do this is to work with a provider like Cotality that has a complete, integrated data and analytics ecosystem.

“Cotality solutions are accelerating the 4 Cs of lending: credit, capital, collateral, and capacity. Our Credco platform is integrated with more than 80 leading LOS and POS solutions. Cotality's AutomatIQ Borrower lets lenders validate income, assets, and employment in minutes. For complex borrowers (self-employed, multiple income streams, rental income) it lets lenders analyze income, assets, liabilities, and cash flow according to their own lender business rules. Lenders using the platform eliminated nearly eight hours of manual work per file and cut cycle times by four days. At the same time, they were able to take on an additional 40 percent volume without adding staff.

“Our CMS and Mercury Network Solutions are the industry standard for appraisal review and QC… and are UAD 3.6 ready. This autumn, we are introducing Collateral Investigate, a new AI-powered solution that reduces rework and exceptions, surfacing insights lenders can query in plain language. It also moves rules and quality alerts right onto the appraiser’s desktop. Similarly, Cotality’s new Virtual Inspection accelerates collateral reviews for home equity and investment lending. Every image is captured live, validated at the source, and returned in a tamper-proof package that shows exactly where, when, and how it was collected.

“Think about servicing risk, retention opportunities, and tax forecasting. Say ‘servicing’ and the C-suite used to think cost center. But not anymore. Today, servicing plays a strategic role in risk management, profitability, and brand reputation. Forward-looking lenders and investors understand that every portfolio is always changing, creating both opportunity and risk. Cotality’s Portfolio Intelligence and Monitoring provides both sides of that picture in real time. On the opportunity side, it flags equity-rich homeowners who may be interested in HELOCs and cross-sell opportunities. On the risk side, it watches for what’s easy to miss, but costly “gotchas,” like unpaid or hidden HOA, municipal, mechanics, and tax liens associated with properties, as well as the risks that have always been there, like escrow shortfalls and delinquent property taxes.

“Tax increases are one of the biggest sources of friction for servicers, and it will only get worse. In fact, Cotality estimates that in 2026 roughly one in three borrowers could have escrow shortages of more than $100/month. As the nation’s largest tax servicer, Cotality has developed the industry’s first forecasting tool, DigitalTax Forecasting, which helps servicers predict how taxes will trend across their portfolio, allowing them to plan for escrow shortages before payment shock hits. In addition, our DigitalTax Portal is the self-service platform used to centralize tax management and give servicing teams real-time visibility into loan statuses, deadlines, and delinquency trends. Paired with growing IVR, text, and digital portal connections, homeowners get proactive answers before they ever pick up the phone.

“Attrition… Retention… The luck you make. It’s the industry’s dirty little secret: 80 percent of your clients will do their next deal elsewhere. That’s not bad luck, that’s bad retention. Cotality's OneHomeowner closes that retention gap. Post-closing, every homeowner gets two jobs they haven’t signed up for: property manager and asset manager for what is likely their biggest investment. OneHomeowner supports them in both, surfacing refinance opportunities, helping them track wealth through home equity, connecting them to trusted service providers, and keeping them informed on their local market, all under the lender’s brand, year in year out. The early results speak for themselves. One Movement Mortgage loan officer using OneHomeowner saw applications from past customers climb 63%. Stop giving your hard-won customers away, defend your database with OneHomeowner.” Thank you, Pat!

NEXA Interview

As NEXA has grown it has been under a microscope by those outside the company. Questions range from everything, from its compensation model and recruiting practices to its approach to servicing. Mike Kortas isn't running from those questions. In fact, he thinks some of the criticism comes from people trying to explain a model they haven't actually taken the time to understand. In a candid new essay that is an exclusive to Chrisman Commentary, Kortas addresses the biggest questions head-on, including where the money comes from, why he rejects the MLM comparison, and how NEXA thinks about regulation and compliance. He also gets unusually personal about the criticism he's received, the mistakes he's made, and why he believes competition, not the failure of competitors, is what makes NEXA better.

In response to everyone trying to figure out NEXA's economics from the outside: “Call me. I’ll explain it.” In his new essay, he tackles the questions people are actually arguing about: Why does NEXA pay people to recruit? Where does that money come from? Why does he say the borrower isn't funding it? And why does he think the industry's “MLM” comparison misses the point? Then he gets into how some of the company's most unconventional programs exist because he believes traditional mortgage companies leave too much value on the table. (If you’d like to listen to the actual interview between Robbie and Mike K., it starts at the 8:20 mark.)

Thoughts on AI, Experimentation, and Opportunities

A lender may have an agent working on pricing while an investor has another agent operating on its side, and both may be acting in real time based on their own instructions and objectives. The capability is exciting, but it also creates a fundamental governance question: when something goes wrong, who ultimately owns the decision? We need people who understand both the technology and the messy reality of mortgage execution, because simply having a risk manager on one side and an AI specialist on the other does not answer that question.

That becomes even more important as AI moves into negotiations, servicing, and the long-term borrower relationship. Agentic systems could potentially negotiate pricing or exceptions, while servicing agents can maintain relationships with borrowers over years rather than relying on occasional outreach when rates change. Those capabilities create real opportunities, but they also make governance something that has to operate in real time rather than simply as an annual certification exercise.

We need to understand what these systems are doing, where their instructions come from, when human intervention is required, and how responsibility is assigned when multiple agents are interacting. The path forward depends on understanding what we are actually deploying rather than getting distracted by the latest buzzword. The cost of experimentation is lower than it has ever been, so lenders do not need to wait for the industry to have every answer. They do, however, need to experiment thoughtfully, bring in people who understand the technology beyond the highlights, and build the governance necessary to know exactly what they are getting into.

Capital Markets

Morgan Stanley knows a thing or two about the markets. Morgan’s economists expect the Fed to stop after hikes in December and March, about one hike short of market pricing for 2027. “We forecast 10y UST yields about 30bp below forwards by the end of 2027, believing that in a year rates will be somewhat lower than they are now. The 10y UST yield again tracks market pricing of the Fed path closely, so higher yields would require a more hawkish Fed path than markets already price… A moderate oil rise should lift UST yields by adding to expected hikes, but a spike to $140-160 per barrel would likely push the Fed toward cuts and yields lower.”

Continuing the lesson that “you can’t talk rates up, or down,” U.S. Treasury yields climbed inexorably yesterday, with the 30-year yield nearing 5.5 percent, its highest level since 2004. Investors are grappling with robust domestic economic data including a solid labor market, persistent inflation pressures, and the enormous supply of U.S. government debt. Fixed-income traders have rapidly adjusted their projections, driven less by inflation expectations than by markets pricing a much higher-for-longer Fed path, with the terminal rate approaching 4.86 percent, nearly 100-basis points above the Fed's September projections. Treasury and MBS demand are showing signs of strain, with persistent cheapening across the coupon stack and another disappointing long-end buyback, while equities have remained surprisingly resilient despite the rise in long-term yields.

The bond market is asking, “Who is going to step in as the buyer of size?”” With increasingly expensive auctions struggling to attract traditional buyers (yesterday's $44 billion 7-year note auction saw relatively soft dollar demand and foreign interest), and no obvious source of incremental demand emerging, the market is testing just how much higher yields need to go before buyers of size finally step in. This action has pushed the market-implied odds of an October 28 rate hike above 75 percent, while easy broader financial conditions and resilient equity markets near record highs suggest the benchmark curve has structural room to sell off further before triggering defensive flight-to-quality capital flows.

Falling for-sale inventory, driven by millions of homeowners locked into sub-4 percent mortgage rates, has triggered an unprecedented pricing anomaly, pushing median existing home prices to a record nearly $49k premium over new construction. Agency mortgage dollar volume has surged 50 percent over the last decade due to rising home values, even as overall loan counts dropped by 10 percent, illustrating how high prices are crimping broader consumer demand. With 30-year lending rates hovering near 7 percent and a rising share of new homes exceeding $600k, affordability headwinds persist, leaving mortgage-backed securities (MBS) issuers facing muted supply that remains highly sensitive to fluctuations in annual home sales volume. The housing market is shifting leverage toward buyers as affordability pressures force sellers to adjust: nearly one in five homes for sale received a price cut in August, the highest share for that month in Redfin’s data since 2020. Inventory is also up 46 percent from 2023, and homes are taking longer to sell, giving buyers more negotiating power as elevated borrowing costs continue to weigh on demand.

Today’s economic calendar is underway with August Durable Goods Orders (normally volatile, it was unchanged). Later today brings Final September University of Michigan Consumer Sentiment, and remarks from New York Fed President Williams and Kansas City Fed President Schmid. We begin the day with Agency MBS prices little changed from Thursday’s close, the 2-year yielding 4.90, and the 10-year yielding 5.18 after closing yesterday at 5.18 percent. The 2-yr note yield has risen 57-basis points this month, while the 10-yr note yield has jumped 44-basis points... and mortgage rates along with them.