To The Moon

Kevin Warsh’s debut as Federal Reserve chair sent an unmistakable signal: the easy-money era is over. The central bank held its benchmark rate steady at 3.5% to 3.75%, but the updated projections told a different story. Nine officials now see at least one rate hike this year, a sharp pivot from earlier expectations of cuts. Sticky inflation, hovering near 3.6% on core PCE measures, driven by resilient growth, energy pressures, and lingering geopolitical tensions, has forced even patient policymakers to reconsider.

I have followed monetary policy long enough to respect this shift. Warsh, known for his hawkish leanings, is confronting an economy that refuses to cool on schedule. Higher-for-longer rates or potentially higher rates raise the cost of capital across the board. That matters profoundly for the market narrative of 2026: the unrelenting AI boom.

The scale of investment underway is staggering. Hyperscalers Amazon, Microsoft, Alphabet, Meta, and Oracle are collectively channeling $600 billion to $750 billion or more into capital expenditures this year, the bulk targeting AI infrastructure: data centers, chips, networking, and power. Some estimates push the figure toward $800 billion, with three-quarters of that spend flowing directly into AI-related buildout. This is not incremental spending; it is a generational infrastructure sprint.

I remain genuinely impressed by the momentum. Tech giants continue to deliver strong earnings, semiconductors ride unprecedented demand, and the “Magnificent Seven” or so names still dominate index returns. The S&P 500 has climbed to fresh records largely on the back of this concentration. AI is not mere hype it is reshaping how companies compute, innovate, and compete. Productivity gains may eventually validate much of the optimism.

Yet I grow increasingly skeptical that this surge alone can indefinitely prop up elevated valuations in a world of elevated borrowing costs. Higher interest rates act like gravity on long-duration assets. They elevate discount rates, compress multiples, and make the massive debt-financed portions of these buildouts more expensive to service. Many of these projects were greenlit in a lower-rate environment. If returns on invested capital disappoint—if monetization trails the hype or efficiency gains prove slower than expected—the reckoning could be sharp.History offers cautionary parallels. Technology booms have powered markets before, only to falter when capital becomes scarcer. Today’s concentration is extreme: a handful of AI-linked stocks command an outsized share of market capitalization. That has masked weakness elsewhere and left the broader market vulnerable to rotation or disappointment. A pullback in crowded tech trades would not surprise me.Markets have demonstrated remarkable resilience, absorbing rate uncertainty while AI spending accelerates. But resilience should not be confused with immunity. Volatility is likely to rise. We may see profit-taking in overextended names, pressure on high-valuation growth stocks, and opportunities in sectors less dependent on cheap capital.In my view, investors should temper enthusiasm with realism. The AI transformation is real and likely transformative over the coming decade. Extraordinary capital spending reflects genuine conviction from the world’s most sophisticated operators. Yet extraordinary spending does not guarantee extraordinary returns when the cost of capital climbs. Diversification beyond the AI leaders, selective exposure to companies with clear paths to monetization and strong balance sheets, and maintaining liquidity for volatility feel prudent.The Warsh Fed has drawn a line. AI will have to deliver not just promise, but tangible, rate-resistant cash flows. The boom is powerful, but in this environment, it is not bulletproof.