Wall Street has spent much of the past year assuming the Federal Reserve is nearly done raising interest rates. But one research firm argues that the artificial-intelligence investment boom could keep the economy running hot enough to force the central bank to push its peak rate back to the levels seen in 2023-24.
In a note, TS Lombard economist Dario Perkins says the terminal fed funds rate—the highest point the Fed takes its benchmark rate in this cycle—could match or even exceed the 5.25%-5.50% range that prevailed through much of 2023 and 2024. That's a stark contrast to what markets are currently pricing.
What markets expect vs. what could happen
Right now, traders are treating the Fed as closer to done than not. Interest-rate futures and other market instruments imply the central bank's policy rate will eventually settle around 4.50%-4.75%. That's roughly a full percentage point below the current target range of 4.25%-4.50% (the Fed cut rates in late 2024 and early 2025).
Perkins thinks investors may be underestimating how long demand could stay strong. His reasoning centers on the massive capital expenditure—capex, in Wall Street jargon—that the biggest technology companies are pouring into AI and cloud infrastructure. This isn't small change: it's long-term spending on data centers, specialized chips, and networking gear that can run into the hundreds of billions of dollars.
He points to the so-called hyperscalers—Amazon, Alphabet, and Meta—as the main drivers. Meta, for instance, has guided to 2026 capital spending of $135 billion to $145 billion, a figure that would have seemed unthinkable just a few years ago. When companies invest at that scale, the money flows through the economy: it creates jobs in construction, manufacturing, and engineering, and it boosts demand for everything from electricity to cooling systems.
If that investment wave keeps lifting growth and keeps hiring resilient, inflation could prove stickier than the Fed hopes. That would make it harder for the central bank to declare victory over price pressures and would argue for keeping rates higher for longer.
Why a higher peak matters for your portfolio
The gap between the market's roughly 4.50%-4.75% terminal-rate view and Perkins' 5.25%-5.50% scenario is more than an academic debate. It has direct consequences for borrowing costs and asset prices.
When traders start pricing a higher peak, short-term Treasury yields and rates on instruments like SOFR (the Secured Overnight Financing Rate, a benchmark for dollar funding) tend to move first. That higher path then filters into the discount rate investors use to value future profits. A higher discount rate reduces the present value of cash flows expected years from now.
That's a particular headwind for what are called long-duration stocks—companies where a large share of value comes from profits far in the future. Mega-cap growth names, including Microsoft, Alphabet, and Amazon, often fit that description. Their valuations are built on expectations of strong earnings growth over many years, so they're more sensitive to changes in interest rates than, say, a utility or a consumer staples company.
The irony is that the same AI capex boom that supports upbeat revenue stories at these companies can also compress their valuations if it keeps the Fed's peak rate in play. As earnings season approaches, investors will be watching both the revenue numbers and what management says about future spending.
Perkins' bottom line: if the economy is still holding up, a "restrictive" policy stance—meaning rates high enough to cool demand—may require a higher peak than today's curve suggests. The 5.25%-5.50% zone could be revisited before something in the economy cools meaningfully.
What to watch next
For everyday investors, the key takeaway is that the path of interest rates is far from settled. The Fed has signaled it wants to see more progress on inflation before cutting further, but the AI investment boom could keep upward pressure on prices.
That means the "higher for longer" narrative could persist, affecting everything from mortgage rates to the returns on savings accounts. It also means the corporate bond market, where companies like the hyperscalers are borrowing heavily to fund their AI buildouts, will remain a key barometer of investor confidence.
If the Fed does end up pushing rates back to 5.25%-5.50%, it would be a significant shift from current expectations. But as Perkins suggests, the same forces that are driving the AI boom—massive investment, strong demand, resilient hiring—could be the very forces that keep the central bank's foot on the brake.
For now, the message for investors is to stay alert to how the data evolves. Inflation reports, jobs numbers, and any hints from Fed officials about their thinking will all matter. And as Microsoft's recent buyout offers show, even the biggest tech companies are adjusting their cost structures to fund their AI ambitions—a sign that the spending spree is far from over.
