Latest / The National Debt Podcast with Fexingo: Treasury, Borrowing, and Long-Term Fiscal Outlook / How Sovereign AI Spending Reshapes US Debt Strategy
Transcript
- Lucas: If you want to understand where the United States’ borrowing costs are heading next year, you probably shouldn’t be looking at Washington right now. You should be looking at Seoul. South Korea has just unveiled a record-breaking budget dedicated entirely to building what they call a 'world’s highest performance' sovereign artificial intelligence model. It’s a massive bet on compute capacity, essentially arguing that controlling the hardware and algorithms of tomorrow is more critical than balancing today’s ledger. Luna: It’s interesting because it flips the script on traditional fiscal conservatism—where do we put our money? In savings or in the engine that drives future growth? Lucas: Exactly. And this matters for us because the mechanics of funding that kind of industrial policy are bleeding into global capital markets. When major economies decide to borrow heavily to buy GPUs and build data centers, it changes the demand curve for Treasury bonds. We’re seeing the ten-year yield sitting around four point nine four percent today, but the real question is whether that number stays stable when every other G20 nation starts competing for the same finite pool of high-grade credit. Luna: So you’re saying the US isn’t just competing with China anymore; we’re competing with countries like South Korea who are willing to run bigger deficits just to secure their tech sovereignty. Lucas: Precisely. Let’s look at the numbers on our side of the Atlantic. As of early October, the federal debt total was sitting at roughly thirty-nine point five trillion dollars. That sounds abstract until you realize the deficit for the third quarter alone widened to nearly two trillion dollars annualized. The government is already borrowing aggressively to fund existing operations, and now there’s a new category of spending—strategic tech—that demands even more liquidity. Luna: And with the Fed Funds rate holding steady near three point eight eight percent, the cost of servicing that debt isn’t cheap anymore. Lucas: Right. Interest on reserve balances are at three point nine zero percent. So the government is paying almost four percent just to keep cash parked at the Federal Reserve, let alone what it pays investors in the secondary market. If South Korea’s example triggers a broader shift where other nations prioritize AI infrastructure over debt reduction, we could see sustained upward pressure on yields. That squeezes mortgage rates, corporate borrowing, and eventually, consumer credit. Luna: It feels like we’re entering an era where fiscal discipline takes a backseat to technological supremacy. Lucas: That’s a fair characterization, but I’d argue it’s more nuanced. It’s not about abandoning discipline; it’s about redefining what counts as an asset. A bridge is a physical asset. A sovereign AI model is an intellectual and computational asset that can generate exponential economic returns. The risk is that if the returns don’t materialize fast enough, you’re stuck with the debt without the payoff. Luna: But doesn’t the sheer scale of that investment distort the bond market? If everyone is buying similar tech-heavy assets, aren’t we creating a bubble in government-backed innovation? Lucas: I think the distortion is already happening in the short end of the curve. Look at the three-month Treasury yield; it’s at four point one six percent. That’s tight to the Fed Funds rate, which means the market is pricing in very little change in monetary policy. But the long end, the thirty-year at five point three one percent, tells a different story. Investors are demanding a premium for lending money out over decades, likely because they’re worried about inflation coming from exactly this kind of supply-side spending shock. Luna: So the bond market is whispering that we’re overspending on the future, even if GDP looks strong at three point four percent growth. Lucas: Stronger than expected growth usually cools fears of recession, but it also cools the incentive for the Federal Reserve to cut rates aggressively. If the economy is humming along at three point four percent while the government adds trillions to the debt pile for AI projects, the Fed might stay higher for longer. That keeps borrowing expensive for everyone else. Luna: It’s a strange paradox: the stronger the economy gets, the harder it is to afford the debt that helped get us here. Lucas: Well said. Before we dive deeper into how this plays out for your portfolio, I want to check in with you. If walking through the economy with us has made something click, the way these stay ad-free is listener support—buy me a coffee dot com slash fexingo — that's F. E. X. I. N. G. O. Luna: Yeah, and honestly, keeping these conversations grounded in real data rather than hype is what makes the difference. Your support literally lets us dig into the spreadsheets instead of reading headlines. Lucas: Thanks. Now, back to the core tension. South Korea’s move is aggressive, but is it sustainable? Their GDP per capita is high, but their population is aging rapidly. How do they justify this level of capital expenditure when their domestic labor force is shrinking? Luna: That’s the key. They’re betting that automation and AI will offset the demographic decline. If they get it right, the productivity gains will pay off the debt. If they get it wrong, they’re just adding leverage to a stagnant system. Lucas: And here’s the kicker for US investors: the supply chain for that AI infrastructure is heavily concentrated. Most of the advanced semiconductors required for those sovereign models come from a handful of suppliers, many of whom rely on US technology or US capital markets for financing. So when Seoul spends big, Tokyo spends big, and Brussels follows suit, the money flows through New York. Luna: So the US becomes the banker for the world’s AI race, effectively financing our competitors’ infrastructure. Lucas: In a way, yes. We sell the bonds, they buy the bonds, then they use the proceeds to build factories that compete with American firms. It’s a circular flow of capital that benefits Wall Street in the short term through issuance fees and trading volume, but it may weaken American industrial dominance in the long term if we don’t match their investment pace. Luna: Doesn’t the current deficit make it politically impossible for Washington to launch a matching program? Two trillion dollars a quarter is already a tough sell. Lucas: It is incredibly tough. The debt to gdp ratio is at one hundred twenty-one point two percent. Most economists would say that’s a red line. But geopolitics often overrides economics. If the US refuses to fund its own AI infrastructure, it risks ceding the standard-setting power to Europe or Asia. The political calculation is shifting from 'can we afford this?' to 'can we afford not to?' Luna: That’s a dangerous precedent. Once you start treating national security budgets as immune to fiscal constraints, where does it stop? Lucas: History suggests it stops when the interest payments become unmanageable. Right now, interest expense is a significant portion of federal outlays, but it’s still dwarfed by defense and entitlement spending. However, if the spread widens—if that ten-year yield jumps above five percent consistently—interest alone could become the fastest-growing line item in the budget, crowding out everything else. Luna: Which brings us back to the VIX and market volatility. Are investors starting to price in that risk? Lucas: Subtly. The VIX itself is calm, but the term structure of Treasury yields is flattening. The spread between the two-year and the ten-year is only thirty-two basis points. That’s historically low. It means the market thinks short-term rates will stay high, but long-term growth prospects are uncertain. It’s a classic sign of stagflationary anxiety, or at least slow-growth anxiety masked as stability. Luna: So we’re not in a panic yet, but the warning lights are flickering on the dashboard. Lucas: Exactly. And here’s what I want you to take away from this episode. When you hear about foreign governments spending billions on AI, don’t just see it as a tech story. See it as a debt story. Every dollar spent on a sovereign model is a dollar borrowed from the global pool of capital. That competition drives up the cost of money for all of us. Luna: It’s a reminder that fiscal policy isn’t just about taxes and spending bills; it’s about where we place our bets on the future. Lucas: Right. And as that future gets built, the interest rates you pay on your mortgage, your car loan, and your business line of credit are directly influenced by how much other countries are willing to borrow to win the AI race. Luna: So the next time you see a headline about a new chip factory in Asia, remember it’s also a headline about your monthly payment. Lucas: Spot on. The connection is closer than most people realize. We’ll keep tracking these yields and the global policy shifts as they happen. Thanks for listening.