PodcastEurodollar TalkPeter Boockvar

The Stock Market Has NEVER Been This Dependent on One Theme

The guest argues that the US economy, corporate earnings, and the stock market have become unusually dependent on AI-related capex, while the scale of the investment does not yet establish its eventual payoff. The discussion also points to constraints and risks, from debt-market competition and rising borrowing costs to possible AI liability and fraud. Economic conditions are uneven, and stress in lower-rated credit could deepen if refinancing becomes more expensive.

Artificial IntelligenceEquitiesCorporate Earnings

Preview · 15 more points in the app

Read the Full Summary

Key Takeaways

The guest describes an AI investment cycle that could reach about 3.5% of GDP from 2025 to 2032, with Wall Street expecting hyperscaler capex of $800 billion in 2026 and $1.2 trillion in 2027. He says the US economy, S&P 500, and earnings growth have become unusually reliant on that spending, even though it is too early to quantify the payoff and some AI-built capacity may fail to attract users.

Debt-market constraints and competition for capital could interrupt AI capex, the speakers warn, while the guest says Meta’s 7% financing for an off-balance-sheet data center and CoreWeave’s 9% borrowing raise the financing benchmark for lower-rated borrowers. He argues those costs point to pressure for a rerating of triple-C debt.

The guest says higher rates are prompting closer scrutiny of interest coverage and refinancing needs; CCC spreads have risen above 950 basis points, and some private-equity-backed companies refinancing 2020–21 debt could see their cost of capital rise from about 6% to 12%. He is watching whether high-yield stress spreads to single-B and double-B debt.

The guest characterizes the economy as two-speed: AI data-center construction and upper-income spending are strong, while housing is in recession and lower- and middle-income consumers face pressure. He says roughly $700 billion of the government’s $1 trillion in interest expense goes to US Treasury holders and can support upper-income spending, but that demand could weaken if the stock market falls or white-collar job losses push upper-income consumers down the income ladder.

GEO-POLITICS

Multipolar Shift — The guest argues that the world is shifting from a US-centric order toward a multipolar one, with countries pursuing more independent trade and capital flows and China seeking a global role, particularly in technology. He says tariffs and the approach of Trump 2.0 have angered countries, leaving investors unable to rely on a familiar historical playbook.

“And I'll add one more thing. You know, for decades, particularly post the fall of the Berlin Wall, we have this very US- centric world…”—

TECHNOLOGY

AI Capex Scale — The guest cites a Columbia professor's estimate that the AI capex cycle could reach about 3.5% of GDP from 2025 to 2032, exceeding prior investment cycles, particularly railroads. He cautions that factories can be built for products that fail to attract users, even as successful technologies may benefit users.

“Now I think what is interesting about this moment just based on forecasts and we'll see whether the forecasts are realized or not but this particular capex spend and there…”—

Hyperscaler Spending — Wall Street estimates cited by the guest put hyperscaler spending at about $800 billion in 2026 and $1.2 trillion in 2027, a projected 50% increase in capex. He describes those estimates as the current expectations bar for the spending cycle.

“So when I quantify that that the where these companies are spending money with these other companies that okay take let's take the hyperscaler spend Wall Street estimates for 2026…”—

ECONOMY

Capex Dependence — The guest argues that the US economy, stock market, earnings growth, and historically high profit margins have become unusually dependent on the capex buildout. He says the spending generates almost half of US economic activity, nearly half of the S&P 500, and about half of the earnings growth rate. Any slowdown in capex growth could disrupt that reliance.

“I don't recall a time period that has been so dependent on one theme and I like to say the US economy, the US stock market, US earnings growth story…”—

Fed Safety Net — The guest cautions investors against assuming the Fed can provide a reliable safety net in the next market downturn. He argues that inflation has reduced the Fed's influence over interest rates.

“They were the gods of interest rates. Well, that is no longer the case. Inflation has completely changed the game…”—

Technology and Rates — The guest says productivity gains and deflation recur with technological change and are not unique to generative AI. He argues that this history alone does not mean interest rates or the Fed's response need to follow a different course.

“You know, we hear from people that, you know, this is going to be hugely productivity enhancing and it's going to be greatly deflationary…”—

MARKETS

AI Payoff — The guest says it is too early to quantify the benefits of current capex relative to the spending, even if AI investment represents a step change from earlier cycles. He argues that familiar market cycles, valuations, and interest-rate changes still matter to investors.

“We're beginning to get a little bit of the evidence of where the benefits sort of acrue in terms of how people are using it, but I think it's still…”—

Chinese Memory Risk — The speaker warns that if China develops a homegrown DRAM and HBM producer, Chinese spending could shift away from Samsung and SK Hynix, and possibly Micron. He says that could put Micron's 85% gross margin at risk; the scenario is conditional.

“And what this means is that if China is able to create an a homegrown DRAM HPM type company, well, that means China's spend is less with Samsung, less with…”—

Nvidia Margins — The speaker says Nvidia reported 75% gross margins in multiple prior quarters and indicated that higher component costs could bring margins down to 71%–72%. He notes that Nvidia stock has chopped around since.

“If you look at Nvidia their last quarter, they went into that quarter with 75% gross margin in their in their multiple prior quarters and they said because of higher…”—

Investment Environment Shifts

The guest says markets are always challenging, but AI concentration, changed interest-rate dynamics, and a more multipolar world make today’s investment landscape unusually unfamiliar.

AI Capex Concentration

The guest argues that AI-related capital spending now drives a remarkable share of US economic activity, earnings growth, and profit margins, creating unusual dependence on one theme.

AI Bubble And Human Nature

The speakers compare the AI investment surge with past technology booms, warning that investors may overbuild and overestimate near-term returns even when the technology ultimately benefits society.

AI Payoff Uncertainty

The guest says it remains too early to measure AI’s benefits against the enormous buildout, leaving investors exposed to both substantial upside and the risk of disappointing adoption.

AI Capex Warning Signals

The guest tracks hyperscaler spending expectations, chipmaker margins, and Chinese competition, arguing that any slowdown in AI investment could threaten stocks priced for continued acceleration.

China Competition And Margins

The guest highlights China’s effort to build domestic memory-chip capacity and warns that competition could pressure incumbent suppliers’ exceptionally high margins.

AI Liability And Regulation

The speakers consider whether harmful AI-enabled activity could expose model developers to lawsuits and liability, potentially prompting companies to restrain spending or deployment.

Debt Markets And AI Spending

The guest sees rising borrowing costs and competition for capital as risks to continued AI investment, while warning that duration losses make investors less willing to fund large debt needs.

Credit Stress And Refinancing

The guest flags widening CCC spreads and private-equity refinancing needs, saying higher rates could expose weak interest coverage and deepen stress if credit conditions deteriorate.

K-Shaped Economy And Consumers

The guest describes a fast AI and upper-income economy alongside weak housing and pressure on lower-income consumers, and explains how rates and household spending complicate the Fed’s choices.

Labor Risks To The AI Boom

The guest says the AI-led economy is not insulated from weakness: job losses among upper-income white-collar workers could reduce spending and weaken the market’s fast lane.

Related Episodes

Important Disclaimers

Not Investment Advice: For education only — not financial, tax or legal advice.

AI-Generated: Made by AI from source transcripts and may contain errors.

No Guarantees: Past performance does not predict future results; investing carries risk.

Our Commentary · Request Removal