“Damn, there is so much great knowledge out there. Did you know that “BOOKS” are full of smart?? No, I mean like life changing, I-wish-I-knew-that-years-ago type stuff.

I know that I was waaaayyy late to the game figuring it out. And I know that a lot of you are too busy to read as much as you ‘should’. And that is why you need me.

I still remember how it started for me. It started in June of 2008. After 11 years …..Click to continue

Your AI Tools Work for You — But They Answer to Someone Else

{{catlist}}
April 23, 2026

Your AI Tools Work for You — But They Answer to Someone Else

Every loan officer using an AI-powered CRM. Every real estate agent running automated market analysis. Every mortgage company leaning on machine learning to speed up underwriting. You're all building your business on a foundation you don't own, can't see, and probably don't think about. That foundation is controlled by five companies spending half a trillion dollars a year to make sure it stays that way. Here's what that actually means for your business — in plain English.

The AI You Use Is Built on Rented Land

When you use an AI tool in your mortgage or real estate practice, you're interacting with layers of technology stacked on top of each other. Your vendor's software is the layer you see. But beneath it? Servers, chips, and networking hardware running inside a data center owned by Amazon, Google, Microsoft, Meta, or Oracle. Those companies charge your vendor for that computing power. Your vendor wraps it in a product, adds their own features and support, and charges you. It's like leasing an office in a building you've never visited, managed by a landlord you've never met, who answers to five corporate real estate giants who just happen to own most of downtown. As long as pricing is competitive and the infrastructure runs reliably, this arrangement works fine. But it creates dependencies that run deep — and those dependencies have real consequences when things shift.

The Cost Cliff Is Real

Here's a number worth sitting with: the top three hyperscalers alone plan to invest more than $500 billion in AI infrastructure in fiscal year 2026. That's an almost unimaginable amount of capital being deployed to build the biggest, most expensive computing infrastructure in human history. That capital has to come from somewhere, and eventually it has to be paid back. Right now, hyperscalers are investing aggressively to capture market share. Pricing is relatively competitive. But history suggests that once a platform achieves dominant market position, pricing tends to follow. We've seen this pattern in cloud storage, in advertising technology, in payment processing. The early days are cheap and competitive. Then consolidation happens. Then pricing power shifts toward the platform. This is not cynicism — it's just how concentrated markets work.

What This Means for Smaller Lenders and Independent Brokers

Large banks and national lenders can absorb rising AI tool costs — they spread them across millions of transactions. They also have the leverage to negotiate directly with vendors and, in some cases, to build their own AI capabilities. Smaller mortgage companies, independent brokers, and boutique real estate firms don't have that luxury. They are almost entirely dependent on the vendor ecosystem. And that vendor ecosystem is almost entirely dependent on the hyperscalers. A study of how lender size affects AI adoption found meaningful differences in what tools are available and affordable at different scales of operation. As the infrastructure layer becomes more concentrated and potentially more expensive, those differences could widen significantly.

There's a Silver Lining — If You Act Like It

This is not a doomsday scenario. The AI tools available to mortgage and real estate professionals today are genuinely powerful, genuinely affordable, and genuinely transformative. The landscape is better than it has ever been for practitioners willing to learn how to use these tools effectively. But being a smart user of AI in this environment means understanding what you're depending on. It means asking vendors the right questions: Who hosts your infrastructure? What happens to our data if your pricing changes? What's your contingency if your primary cloud provider raises rates or changes terms? It means diversifying where you can — not putting your entire client communication strategy, your entire lead pipeline, and your entire marketing engine inside a single AI platform from a single vendor on a single hyperscaler's infrastructure. And it means staying informed, because the decisions being made at the top of this technology stack — in board rooms at Google, Microsoft, and Amazon — will eventually ripple all the way down to the independent loan officer in Scottsdale or the real estate team in Nashville.

The Data Center Is the New Real Estate

Here's a thought worth leaving you with: data centers are becoming one of the most valuable classes of physical real estate in the world. Hyperscalers operated 1,360 major data centers globally by the end of 2025 — nearly triple the number from 2018. They are consuming enormous amounts of land, power, and water to house the infrastructure that AI depends on. In a very real sense, the companies winning the AI race are the ones winning a real estate race first. They're locking up the physical and digital land that everything else depends on. You spend your career helping people navigate the most important real estate decisions of their lives. The same principles apply here: location matters, ownership matters, and understanding who controls the supply is everything. The hyperscalers understand this. The question is whether the rest of us are paying close enough attention.

Mortgage Today (AM) - 04/22/26

{{catlist}}
April 22, 2026

--- WTMS Blog Today = What's up in Mortgage Today (AM) - 04/22/2026 Treasuries fell sharply overnight as the 10-year yield dropped 3.5 basis points to close at 4.264 percent, with the broader curve following suit across all maturities. Market uncertainty surrounding the Iran ceasefire deadline and Middle East tensions drove bond strength, though oil briefly climbed above $100 per barrel amid failed peace talks in Islamabad. UMBS securities held relatively stable with modest intraday gains, while strong economic data released this morning—including better-than-expected retail sales at 1.7 percent and ADP employment of 54.75K—now pose a headwind for further rate declines. Pending home sales surpassed forecasts with 1.5 percent growth, signaling continued resilience in real estate activity despite elevated rates. The geopolitical risk premium remains the dominant driver of fixed-income sentiment today. Corporate earnings season continues to support equity futures, with major chip names and AI-beneficiary stocks driving broad market participation. Tesla, Texas Instruments, and Lam Research are reporting today, while semiconductor index momentum extends to fifteen consecutive days of gains. Boeing recovered 3.8 percent in premarket trading on strong first-quarter aircraft deliveries, demonstrating manufacturing strength despite global headwinds. Data-center electrification plays like GE Vernova surged 8 percent after booking more orders in Q1 than all of 2025 combined. However, airlines including United cut full-year guidance due to elevated fuel costs from ongoing Middle East conflict. GNMA 30-year securities posted slight intraday gains with the 5.0 coupon up 0.2 points to 99.78 and the 6.0 coupon at 101.93. UMBS performance was similarly modest, with 5.0 coupons gaining 0.16 points to 99.34 and the 6.0 coupon at 102.37. These pricing signals suggest investors are cautiously positioning ahead of potential weekend developments in ceasefire negotiations. Bond volatility remains elevated due to the fluid geopolitical situation, which could shift market sentiment without warning. Mortgage originators should remain vigilant on intraday Treasury moves as they directly impact secondary market execution. The dollar eased 0.1 percent while Bitcoin climbed 3.4 percent to $78,300, reflecting modest risk-on sentiment tempered by uncertainty. Treasury curve positioning shows the 2-year yield at 3.755 percent, up slightly from overnight movement, while longer-dated maturities compressed downward. This flattening pattern typically occurs when near-term uncertainty creates flight-to-quality demand in extended duration bonds. European yields also declined, with Germany's 10-year falling one basis point to 2.99 percent and Britain's sliding two basis points to 4.87 percent. Global bond markets are clearly pricing in recession risk and geopolitical premium simultaneously. Mortgage originators facing intense rate volatility should recognize that two competing forces now influence pricing: strong domestic economic data supporting rates higher, and geopolitical risk supporting rates lower. This tug-of-war creates both opportunity and danger in execution strategy over the coming sessions. Borrowers sitting on rate locks face improved pricing if ceasefire news deteriorates further, while those in float status risk deterioration if equity markets and oil prices stabilize. The economic data flow remains supportive of Fed patience, which argues for sticky elevated rates once geopolitical concerns ease. Lock/float decisions today demand careful attention to both weekend news cycles and early-week market repricing. Locking vs Floating Volatility is intensifying ahead of Wednesday's potential ceasefire deadline, with bonds coping relatively calmly so far despite the obvious risks. Strong economic surprises in retail sales and employment suggest underlying rate pressure remains upward, but Middle East tensions are temporarily suppressing yields. Mortgage originators should use MBS price intraday moves as a real-time gauge of risk appetite, since these shifts often precede broader Treasury repricing. Both locking and floating borrowers should brace for sharp moves depending on whether geopolitical headlines improve or deteriorate over the next forty-eight hours. Today's Events ADP Employment Change Weekly: 54.75K versus forecast unspecified, previous 39K Retail Sales (Mar): 1.7% versus 1.4% forecast, 0.6% previous Retail Sales Control Group MoM (Mar): 0.7% versus 0.2% forecast, 0.5% previous Pending Home Sales (Mar): 1.5% versus 0.1% forecast, 1.8% previous Bond Pricing UMBS 30 yr | Coupon | Price | Intra-Day Change | | 5.0 | 99.34 | 0.16 | | 5.5 | 100.97 | 0.09 | | 6.0 | 102.37 | 0.11 | GNMA 30 yr | Coupon | Price | Intra-Day Change | | 5.0 | 99.78 | 0.2 | | 5.5 | 100.94 | 0.06 | | 6.0 | 101.93 | 0.09 | Treasuries | Term | Yield | Price | Intra-Day Yield Change | | 2 yr | 3.755 | 100.23 | -0.015 | | 3 yr | 3.769 | 99.243 | -0.025 | | 5 yr | 3.878 | 99.986 | -0.03 | | 7 yr | 4.059 | 101.152 | -0.032 | | 10 yr | 4.264 | 98.879 | -0.035 | | 30 yr | 4.878 | 98.002 | -0.029 | Market Data