🏠 AI BOOM, MORTGAGE SHOCK?! Is Big Tech Pricing You Out of Your First Home?

Well Google is borrowing billions to build AI. Jake and Emily want to borrow 400,000 to buy their first home. So what do those two things have in common? You know, they may be competing for the same investors. The AI boom isn't just creating demand for chips and electricity, it's creating enormous demand for money. Anthropic, the company behind Claude, just filed to go public. And Meta Meta wants $97 billion, Amazon borrowed $54 billion in just two days. You know, everyone's asking, who's going to get rich? I think that's the wrong question. The better one is, who's going to pay? Stick around. By the end, I will show you how this AI boom could make Jake and Emily's mortgage more expensive. So let's start with the Anthropic's IPO filing. On Monday, Anthropic filed to go public, meaning it's preparing to sell shares on the stock market. It's proposed the ticker ANT, ant. I mean, that's very cute. But this ant eats like an elephant. In the first half of last year, Anthropic brought in about $1.2 billion in revenue. And in the first half of this year, $16.3 billion. Well, that's roughly 14 times as much revenue in a year. I mean, if your paycheck did that to you, you could have just quit and buy an island. Now, revenue isn't profit. Anthropic still reported an operating loss for the first half of this year, and it has promised to spend more than $500 billion on computing power in the years ahead. Well, that same Monday, Meta announced plans to raise up to $97 billion by selling stock. Well, that's billion with a B. And NVIDIA, AMD, and Broadcom, you know, those chip makers that sell to Meta, indicated interest in buying up to $5 billion of Meta shares between them. So it's a little bit like the pizza place chipping in for your party so you can order more pizza. But stock isn't a loan, right? Shareholders get a piece of the company and not a promise to be paid back. The loans are where it gets interesting. Because here's the thing. Anthropic doesn't own most of the computers it runs on. Amazon and Google do. And to build those data centers, Amazon and Google are doing something they rarely did before. They are borrowing. For years, companies like Amazon, Google, and Meta funded much of their investment with cash generated by their own businesses. But their AI ambitions are just growing faster than their ability or, frankly, willingness to fund everything internally. In 2024, borrowing paid for about a dime of every dollar these tech giants spent building things like data centers. And now, about 33 cents. So they're issuing bonds, while a bond is basically an IOU. You lend Amazon $1,000, for example, Amazon promises to pay you interest and return your principal, the $1,000, on a specified date. And the scale is staggering. Goldman Sachs estimated that AI-related debt issuance had reached about $489 billion by late July this year. For All of last year, it's estimated it was $322 billion. So in other words, the amount of AI-related debt issued this year had already surpassed last year's total with months still to go. So how much money is that? It's enough to cover more than 1,400,000 home loans. Well, in about seven months and that's Jake and Emily's loan a million times over. And this isn't just a bunch of smaller companies scrambling for cash. In March, Amazon borrowed about $54 billion in just two days through bond sales in the US and Europe. Google's parent company, Alphabet, even issued a bond that won't mature for a hundred years. You know, Google borrowed the money and won't have to repay until 2126. Meanwhile, I can't even commit to a gym membership. And this isn't the first time a hot new technology went on a borrowing spree. Well, in the late 1990s, it was the internet. Phone companies borrowed billions to bury fiber optic cable across America. And by 2002, only about 2% to 5% of that cable was actually being used. And in 2001, telecom companies accounted for about a quarter of all the corporate bond dollars that went into default in America. Well, that means for every $100 in company debt that businesses failed to pay, as promised, 25 bucks came from phone and internet companies. Now, I know, today is not the same, Amazon and Google make far more cash than those phone companies ever did, but the pattern is worth watching. Build first, borrow big, and hope the customers show up in time. And who was holding those telecom bonds when they went bad? Pension funds, insurance companies, bond funds owned by everyday investors like you and me. You know the same big careful lenders we're about to meet? Remember that because it matters later. And there's a reason investors are paying attention. In 2024, AI-related companies accounted for roughly 1% of investment-grade corporate bonds sold in the US, and this year, roughly 18%. And the shift is even more striking when you look at long-term debt. Goldman Sachs estimates that AI-related companies accounted for about 40% of investment-grade corporate bonds sold in the U.S. this year, with maturities of 15 years or longer. Well, in English. This year, for every $10 that highly-rated American companies borrowed through bonds due in 15 years or more, about $4 was borrowed by AI-related companies. Think about what that means. AI didn't just become a bigger customer for computer chips. It became a much bigger customer for long-term money. And long-term money is exactly what the mortgage market depends on as well. So what does any of this have to do with a family trying to buy a house? Well, let's meet Bob. Bob is an investor with money to lend. And he can buy government bonds, corporate bonds, or mortgage-backed securities, which means bundles of home loans that investors help finance. Say Bob has $100 to lend for 30 years. Uncle Sam wants $60. Well, Jake and Emily, they want $40 for their first home. Everybody's happy because everyone got their money. Then, Google walks in. Well, I would like $30 too, please. Now here's $130 in requests, and Bob still only has $100. So he lends to whoever pays more. Google can pay more, of course, and Uncle Sam won't borrow less. So Jake and Emily have to pay a higher rate to stay in line. Nobody raised the rates on purpose, the line just got longer. And of course, more investors can step in with more money, but sometimes it takes higher interest rates to bring them in. And in real life, Bob is pension funds, insurance companies, and other big investors who lend money for decades. When there's suddenly a lot more long-term debt for them to absorb, borrowers have to offer higher yields, which means higher interest rates, to attract enough lenders. But here's the twist. If you have a 401(k), you might be Bob. Many retirement funds own bond index funds and when Google or Amazon sells new bonds, those funds automatically buy some. So you may already be lending money to the AI boom without even realizing it. Which means you could feel this twice. Once on your mortgage rate like Jake and Emily, and once in your retirement fund or account like Bob, if some of those AI bets don't really pay off. Today's borrowers are far stronger than the fiber companies were, of course, but it's the same kind of lender. Different decade, same Bob. And here's why all this matters for Jake and Emily. The 10-year treasury yield, the interest rate the US government pays, is one of the key benchmarks investors watch when pricing long-term loans. Mortgage rates often move in the same direction. Why? Because lenders and investors compare the returns they can earn from mortgages with the returns available on other long-term investments, like government IOUs. Think about it from Bob's side. If the US government, about the safest borrower on earth, will pay him 5% a year, why would he lend to Jake and Emily for 4%? Well, he wouldn't. And right now that 5% isn't hypothetical because this month, the 10 year, hit its highest level since 2002. So the government's rate works like a floor. Mortgage rates get stacked on top of it. And when the floor rises, everything on top of it rises too. Now, here's the part that deserves a closer look. Bank of America Economics have estimated that the surge in corporate bond sales together with more mortgage-backed securities pushed the 10-year treasury yield up by roughly 0.3% points this year. Now, picture that 0.3 points landing on Jake and Emily's mortgage rate. They borrow $400,000 on a 30-year fixed-rate mortgage. Their monthly payment rises by roughly $80. Well, that's close to $1,000 a year. That's money they could have spent on diapers, groceries, or saving for their kids. But instead, it's money they have to find because the cost of borrowing went up. And to be clear, that Bank of America estimate covers all the new company borrowings plus more mortgage bonds, not AI alone. So we can't pin all 80 bucks on AI. And AI isn't the only reason rates are high. It may not even be the biggest. Government borrowing, inflation, and the Fed all matter too, right? What we can say though is that the AI boom is generating an extraordinary amount of new borrowing. Economists see reasons that this flood of debt could push long-term interest rates higher, and mortgage borrowers are exposed to the same market forces. So here's what comes next. Goldman Sachs expects big tech to issue roughly $400 billion in bonds in 2027. The question is whether investors will keep buying that debt at today's prices or start demanding higher returns. That's one of the signals worth watching as AI spending accelerates and the next time you hear about a giant data center, a record bond sale, or billions of dollars being poured into AI, remember this. The bill doesn't always arrive where the spending happens. Sometimes it shows up somewhere else entirely like a family's mortgage. So here's my question for you. If paying a little more on your mortgage meant America wins the AI race, would you take that deal? Well, tell me in the comments below. Thanks for watching, and I'll see you in the next one.