The Strong Dollar Is Forcing the AI Boom to Prove Itself

The dollar move is a rates story

The dollar index rose to 101.69 on June 24, its strongest level since May 2025, as Reuters reported investors moved into dollars during a tech and semiconductor selloff while also repricing the Federal Reserve toward possible hikes. The same report said CME FedWatch pricing put the probability of a July hike at 36%, up from 9% a week earlier, with September odds above 70%.

That matters more than the headline dollar level. DXY at 101.69 is not historically extreme. The practical issue is direction and speed: a stronger dollar plus higher expected short rates tightens financing conditions, raises discount rates, and makes distant earnings less valuable today.

The Fed gave markets reason to reconsider the easy-money assumption. On June 17, the FOMC held the federal funds target range at 3.50%-3.75% by a 12-0 vote and said inflation remained elevated. Its June projections put the median end-2026 funds rate at 3.8%, up from 3.4% in March, with 2026 PCE inflation projected at 3.6%.

AI stocks now face a funding-cost test

The selloff was concentrated where the market had been most willing to pay for future growth. Reuters reported June 23 that the Nasdaq Composite fell 2.2%, the semiconductor index fell 7.9%, and the Nasdaq was down more than 5% from its June 2 peak after a roughly 30% rally from early April.

Calling that automatically “the AI bubble popping” is too simple. Some AI demand is real. Nvidia’s own first-quarter fiscal 2027 results show the cash-generating side of the infrastructure boom. But the market is not made only of Nvidia. It includes chip suppliers, power and data-center plays, cloud platforms, software companies with AI narratives, and businesses relying on investors to fund heavy spending before returns are clear.

The stronger dollar raises the proof standard. For AI-linked companies, the test is no longer whether they can describe a large addressable market. It is whether spending turns into revenue, margins, backlog, cash flow, or lower unit costs. For hyperscalers, large cash flows make the buildout feasible, but higher rates, component inflation, power constraints, and rising debt issuance raise the hurdle for acceptable returns. Reuters reported earlier in June that Morgan Stanley expected global AI debt issuance to top $500 billion in 2026. That makes the rate backdrop central, not secondary.

The global risk is not just a stock-market story

A lazy framing treats dollar strength only as a safe-haven vote. It is partly that. Reuters quoted Ray Attrill of National Australia Bank calling the dollar the preferred safe haven, and Lee Hardman of MUFG tying euro weakness to Fed-versus-ECB policy expectations. But safe-haven buying does not make the move harmless. The contradiction is that investors are buying dollars for safety because the policy path looks tighter, and that same tightening raises borrowing costs.

The global channel is large. BIS data show U.S. dollar credit to non-bank borrowers outside the United States stood at about $14 trillion at end-Q3 2025, with 55% in debt securities. A rising dollar increases local-currency debt burdens for many borrowers and can pressure reserves, spreads, and domestic rates. The IMF’s April 2026 Global Financial Stability Report also flags emerging markets with high debt, low reserves, or weaker institutions as more vulnerable to capital-flow stress.

The uncertainty is real. Fed pricing can reverse if inflation, jobs, oil, or geopolitical data soften before the July or September meetings. The tech selloff may include profit-taking after a sharp rally, not only a fundamental downgrade. AI capex returns will take years to verify. But the market signal is clear enough: the strong dollar is forcing investors to separate durable AI cash flows from valuation momentum and financing stories.