ArchiveFM

We are heading into a very active Q4 for AI infrastructure, private credit and real assets. CoreWeave is being repriced, Treasury yields remain high, and the market is finally forcing sponsors to prove that projects can survive expensive money.

⛽ The inflation story matters because energy is doing real damage. September CPI was 3.7% year over year, core CPI was 2.6%, gasoline jumped 9.1% in one month, and diesel is roughly 69% higher than a year ago. Using current trucking cost data, that diesel shock alone can translate into roughly a 14% increase in truck operating cost, all else equal.

🚛 Diesel sits underneath the physical economy. Food, construction materials, machinery, consumer goods and data center components all move through logistics, so higher fuel costs eventually become higher operating costs and higher prices.

🌍 This is why the Strait of Hormuz matters to AI finance. Reduce geopolitical pressure, normalize energy flows, cool inflation, ease Treasury yields and the cost of infrastructure capital can fall.

🏭 America also has a structural capacity problem. Smaller refineries have been built and existing facilities expanded, but the last new U.S. refinery with significant downstream processing capacity came online in 1977. We want reshoring, AI, manufacturing and energy security without building enough of the physical infrastructure underneath them.

⚡ My Q4 view is straightforward: the AI infrastructure boom is not dead; indiscriminate capital is. The next two years should reward projects with secured power, credible customers, real offtake, disciplined leverage and capital structures that can actually survive underwriting.

Until tomorrow, be the best human you can be.

~ Teresa Grobecker

#AI #DataCenters #Infrastructure #CapitalMarkets #PrivateCredit #Energy #Inflation

Duration
15m
Broadcast on
18 Oct 2026
Audio Format
other
The price of AI is really the price of energy. We are heading into what I believe will be a remarkably active fourth quarter, particularly in AI infrastructure, private credit, and real asset finance. That may sound counterintuitive when the 10-year treasury is above 5%, when AI infrastructure equities are being reprised aggressively, and when companies such as CoreWeave are reminding investors that the market will eventually demand proof that growth can support the cost of capital. But I do not see the current environment as evidence that the AI infrastructure thesis is collapsing. I see it as evidence that the easy money phase is ending, and the underwriting phase is beginning, and that distinction matters enormously for anyone allocating serious capital. For the past several years, the AI narrative has been dominated by demand. Investors have focused on how much compute will be needed, how many GPUs will be deployed, how much power data centers will consume, and how quickly hyperscalers, neo-cloud providers, and sovereign AI initiatives will expand. Those remain important questions, but they are no longer sufficient. The more consequential question now is whether the infrastructure required to meet that demand can actually be financed at today's cost of capital, because a project can have extraordinary demand and still fail if its debt, construction schedule, power economics, or capital structure do not work. That is where the current inflation debate becomes directly relevant to AI infrastructure. September CPI increased 3.7% from a year earlier, which was enough to reignite concern that inflation remains too persistent. Yet core CPI, excluding food and energy, was materially lower at 2.6%. Gasoline increased 9.1% in September alone, and accounted for more than half of the monthly increase in headline CPI, while the broader energy index was more than 20% higher than a year earlier. That does not mean inflation is imaginary, and it would be intellectually lazy to pretend that consumers are not paying higher prices. The more important issue is what is causing the marginal acceleration in those prices. There is a major difference between inflation created by excessive domestic demand and inflation created by a supply side shock in energy and transportation. If the cause is different, then the remedy should also be different, and treating every inflation problem as a monetary policy problem risks attacking the symptom while ignoring the underlying constraint. This is where diesel becomes far more important than most policy conversations acknowledge. Diesel is one of the foundational inputs of the American physical economy, because nearly every tangible product eventually touches trucking somewhere in the supply chain. Food, construction materials, machinery, replacement parts, industrial equipment, consumer goods, and the electrical components required to build AI data centers all have to move. When diesel rises dramatically, that cost does not remain isolated inside the trucking industry. It migrates through the entire cost structure of the economy. The national average price of diesel is now roughly $6.18 per gallon, approximately 69% higher than it was a year ago. According to the American Transportation Research Institute, the average cost to operate a commercial truck in 2025 was $2.336 per mile, while the cost excluding fuel was $1.85 per mile. That implies that fuel represented roughly 20% of total operating cost. If we apply the year-over-year increase in diesel to that fuel component as a simple sensitivity, the fuel shock alone can increase total trucking operating costs by approximately 14%, assuming the other major inputs remain constant. That does not mean every business in America suddenly has a 14% higher cost of doing business, and making that claim would overstate the data. What it does mean is that one of the most important transportation inputs underneath the American economy has absorbed an extraordinary cost increase, and that cost cannot disappear. Freight operators either accept lower margins or pass part of that expense downstream,

Transcription truncated.

We are heading into a very active Q4 for AI infrastructure, private credit and real assets. CoreWeave is being repriced, Treasury yields remain high, and the market is finally forcing sponsors to prove that projects can survive expensive money.

⛽ The inflation story matters because energy is doing real damage. September CPI was 3.7% year over year, core CPI was 2.6%, gasoline jumped 9.1% in one month, and diesel is roughly 69% higher than a year ago. Using current trucking cost data, that diesel shock alone can translate into roughly a 14% increase in truck operating cost, all else equal.

🚛 Diesel sits underneath the physical economy. Food, construction materials, machinery, consumer goods and data center components all move through logistics, so higher fuel costs eventually become higher operating costs and higher prices.

🌍 This is why the Strait of Hormuz matters to AI finance. Reduce geopolitical pressure, normalize energy flows, cool inflation, ease Treasury yields and the cost of infrastructure capital can fall.

🏭 America also has a structural capacity problem. Smaller refineries have been built and existing facilities expanded, but the last new U.S. refinery with significant downstream processing capacity came online in 1977. We want reshoring, AI, manufacturing and energy security without building enough of the physical infrastructure underneath them.

⚡ My Q4 view is straightforward: the AI infrastructure boom is not dead; indiscriminate capital is. The next two years should reward projects with secured power, credible customers, real offtake, disciplined leverage and capital structures that can actually survive underwriting.

Until tomorrow, be the best human you can be.

~ Teresa Grobecker

#AI #DataCenters #Infrastructure #CapitalMarkets #PrivateCredit #Energy #Inflation