Who’s Financing the AI Boom? (With Bryan Whalen)

PodcastOn InvestingBryan WhalenOct 16

The discussion weighs whether returns from AI investment can justify the buildout, as companies and debt markets take on substantial financing needs and credit risks. Speakers also examine persistent inflation and the implications for Fed policy, with Whalen and Colin Martin offering different rate-path expectations. Whalen highlights risks in data-center financing and corporate credit, while also outlining potential opportunities in Treasuries and selected securitized bonds.

AI Infrastructure Private Credit Fixed Income Interest Rates Inflation Risk Management

Preview · 23 more points in the app

Read the Full Summary

Key Takeaways

AI’s buildout is increasingly dependent on debt: investment-grade corporate net issuance was about $500 billion in 2025 and is expected to double in 2026, with nearly all of the increase attributed to AI and data centers. Analysts are expected to scrutinize whether AI adopters earn returns that justify their spending, while debt markets are already relying on the buildout to generate the earnings and cash flow needed to support it.

Whalen warns that AI-related credit risk extends beyond the largest technology companies: he says adoption could be less broad or models much cheaper than expected, while credit markets are reassessing whether software and insurance brokerage businesses can remain viable and repay debt. He says non-AI companies in the broader AI ecosystem could also suffer if equity valuations fall or the cost of capital rises, and that investors are generally not being compensated for those risks.

Inflation data strengthened the case for tighter policy: core CPI rose 0.33% month over month against a 0.24% consensus, about two-thirds of selected categories were running above 3%, and the implied probability of an October hike rose from 22% to 44%. The speakers cited pressures beyond oil—including tariffs, diesel, component prices, and AI buildout costs—and said a CPI-based PCE estimate could rebound into the high twos annualized, pending PPI data.

Whalen describes a policy bind in which rates high enough to slow the AI ecosystem could be hundreds of basis points above current levels, while the neutral rate for other parts of the economy may be below current rates. His call for one more Fed hike differs from Colin Martin’s expectation of two additional hikes into 2027 unless inflation reaccelerates, underscoring that the rate outlook remains conditional and contested.

TECHNOLOGY

AI Investment Returns — Analysts are expected to scrutinize whether companies adopting AI are earning returns that justify their spending, including the gap between the cost of capital and return on invested capital. Some companies are reining in spending as costs rise without clear productivity or profit-margin gains. Debt markets are also assuming AI investment will generate the earnings and cash flow needed to justify the buildout.

“Listening to the adopters of AI and trying to quantify, and I think analysts will be starting to ask more pointed questions of the leaders of the companies that they…”—

ECONOMY

Broad Inflation — Core CPI rose 0.33% month over month, above the 0.24% consensus, while headline CPI was in line with expectations. About two-thirds of selected categories were running above 3%. The speakers said inflation extended beyond oil prices driven by the war, citing tariffs, diesel, component prices, and AI buildout costs as contributors they were beginning to see in the data.

“So the month-over-month increase in core CPI was, to be precise, because these days rounding matters, point 33%, which was higher than the point two four gain that I think…”—

Inflation and Hikes — The CPI report supported the case for a rate hike later in the year, though the speakers were unsure it would change the Fed's decision later that month. Core CPI had risen at a rounded 0.3% rate for a second straight month, and its three-month annualized rate had reached 3.4%. John Williams has said core readings of 0.2% or less would help him feel confident inflation was moving in the right direction.

“It certainly supports the case for one later this year in our view…”—

PCE Estimate — Based only on CPI components that map to PCE, the speakers said the annualized PCE rate could rebound from around 2% in August to the high twos, with the annual rate returning to a little above 3%. They cautioned that the estimate was provisional pending PPI data.

“And once we have both the CPI and the PPI, that's when the economists start to map to the core personal consumption expenditure deflator, which is the Fed's preferred measure…”—

MARKETS

October Rate Odds — After the CPI release, the implied probability of an October rate hike doubled from 22% to 44%, according to the CME FedWatch tool.

“Well, this is your world, but I noticed that a quick doubling in probabilities for a rate hike at the October meeting, I looked at that on the CME Fed…”—

Small-Cap Quality — Among small caps, high interest coverage, stable profit margins, positive earnings revisions, and reasonable valuations had been working. Profitable Russell 2000 stocks had doubled the performance of unprofitable stocks, reversing the prior year's two-to-one outperformance by unprofitable stocks. The speaker linked the shift to a preference for quality in a rising-rate environment.

“But from a factor standpoint, what has been working, particularly down the size spectrum when you go into small caps, are factors like high interest coverage, profit margin stability, forward…”—

CCC Bond Spreads — Spreads on the Bloomberg High Yield CCC Corporate Bond Index were up 4.5 percentage points year to date, a much larger move than for BB- and single-B-rated bonds. CCC bonds had negative year-to-date total returns, cushioned by their high income payments. The speaker cautioned against focusing on CCC bonds and said bond income can help offset potential price declines if yields rise.

“The index we track, the Bloomberg high yield triple c corporate bond index, the average spread is up four and a half percentage points year to date…”—

FORECASTS

Rates and Duration — Colin Martin expects two additional rate hikes to be the most likely path into 2027 unless inflation reaccelerates. Under that outlook, he sees limited upside for short- and intermediate-term Treasury yields and prefers duration slightly below the Bloomberg Aggregate Bond Index's roughly six-year duration.

“We think an additional two kinda seems the most likely path right now as we look into 2027…”—

Divided Economy — Whalen describes a sharply divided economy: AI companies are relatively insensitive to interest rates, while other parts of the economy were already struggling and had worsened. He says this makes the Fed's decision difficult because higher rates may not restrain AI investment but can hurt the rest of the economy.

“And so we have this kind of, you know, it's two economies, a hyper bifurcated economy, one that's marching to its own beat and really is interest rate insensitive, and…”—

CPI Inflation Surprise

Colin and LaZanne review hotter-than-expected core CPI and broad price pressures, then consider how the report may affect Federal Reserve decisions and Treasury yields.

Fed Rate Path

The hosts weigh persistent inflation against the possibility of additional hikes, emphasizing that their rate outlook depends on whether inflation stabilizes or reaccelerates.

Inflation And Market Risks

They discuss the upcoming PCE inflation estimate, real earnings, and the impact of higher yields, while noting that quality and profitability are helping smaller companies outperform.

Small-Cap Credit Quality

LaZanne highlights the relative strength of profitable small-cap stocks, while Colin notes that the weakest high-yield credits face wider spreads and that bond income can cushion losses.

AI Earnings Questions

As earnings season begins, LaZanne says investors should probe AI spending, supply constraints, adoption costs, and whether companies can turn capital investment into measurable returns.

Bryan Whalen Interview

Colin introduces TCW fixed-income investor Bryan Whalen and asks how the unprecedented hyperscaler data-center buildout is being financed through corporate debt markets.

AI Buildout Financing

Whalen describes rapidly rising bond issuance and hyperscalers’ shift from abundant free cash flow to borrowing, while warning that exuberant assumptions may leave parts of the AI ecosystem exposed.

Data-Center Debt Risks

Whalen distinguishes resilient hyperscalers from AI-dependent companies and explains why construction-stage data-center deals may carry risks not reflected in their pricing.

Corporate Credit And The Fed

Whalen says AI disruption and the cost of capital are straining non-AI businesses, creating a bifurcated economy, and expects fewer Fed hikes than markets currently price.

Treasury Yield Drivers

Whalen attributes higher yields to energy-supply shocks, economic resilience, and heavy government and corporate borrowing, arguing that global debt supply is pressuring bond markets.

Treasury And Bond Opportunities

Whalen favors Treasuries for income and potential gains if yields fall, remains defensive on corporate credit, and sees selective value in agency mortgage-backed and other securitized bonds.

Economic Data Ahead

The hosts preview housing, employment, Treasury auction, purchasing-manager, and consumer-sentiment data, then close with their social media and research resources.