Context
This note follows Is AI Productivity Real?, Is AI 1996 or 1999?, and our June macro event reviews. The question here is what actually changed after the June FOMC.
TL;DR
- The June FOMC did not announce a rate hike. The target range stayed at 3.50-3.75%.
- The more important shift is communication. Chair Warsh did not submit his own SEP projections, the statement became shorter, and forward guidance was deliberately reduced.
- The SEP was hawkish: the 2026 federal funds rate median rose to 3.8%, while 2026 PCE and core PCE projections rose to 3.6% and 3.3%.
- But the longer-run federal funds rate median stayed at 3.1%. This is not yet a confirmed 1999-style multi-hike cycle.
- The investment implication is not “no liquidity.” It is expensive money. Capital still flows, but it becomes more selective and prefers AI infrastructure, power, defense, energy, physical bottlenecks, and proven cash flow.
What Changed
The Federal Reserve’s June 17 materials confirm three facts.
| Item | June FOMC | Read-through |
|---|---|---|
| Policy rate | 3.50-3.75% hold | No actual hike yet |
| 2026 funds-rate median | 3.8% | Higher than March’s 3.4% |
| Longer-run funds-rate median | 3.1% | Structural neutral rate not yet repriced |
| 2026 PCE / core PCE | 3.6% / 3.3% | Inflation forecast moved up |
| Warsh SEP | Not submitted | Dot-plot interpretation is less mechanical |
| Statement | Shorter, less guidance | Market must read the data itself |
The key is not only the dot plot. It is the removal of the market’s comfort mechanism. For years, investors have asked what the Fed would do with each data point. Warsh is trying to push the market back toward pricing the data itself.
That weakens the Fed put. It does not eliminate it, but it makes valuation support less automatic.
Expensive Money Is Not No Money
This regime does not mean capital disappears. Governments still spend. Big technology companies still invest. Energy security, defense, reshoring, AI data centers and power grids still pull capital.
But money has a higher price. That changes the question.
Old regime: How large can the TAM become?
New regime: Who actually earns cash flow through the bottleneck?
In the old easy-money regime, long-duration growth, software multiples and unproven TAM stories could be priced generously. In the expensive-money regime, the market asks whether a company has pricing power, recurring demand, high ROIC and a direct link to big wallets such as AI hyperscalers, governments, defense budgets and utilities.
1997 Or 1999?
The June FOMC looks closer to a 1997-style insurance-hike risk than a fully confirmed 1999-style tightening cycle.
The reason is the long-run dot. If the Fed were formally accepting a much higher structural neutral rate, the longer-run federal funds rate median would likely move higher. It did not.
But the risk is clear. If the September SEP lifts the longer-run dot, if inflation stays sticky, or if AI CapEx keeps easing financial conditions through equity markets, the 1999 analogy becomes stronger.
Korea Market Translation
For Korea, the implication is selective exposure rather than broad beta.
- Samsung Electronics and SK hynix remain core because HBM and AI memory are direct bottlenecks.
- Second-line semiconductors need stricter proof: orders, process bottlenecks, margins, and customer CapEx links.
- Power equipment, grid, defense, shipbuilding, nuclear and infrastructure remain important because they connect to hard spending needs.
- Pure theme stocks with no cash flow become more vulnerable when the Fed put weakens.
Final View
The June FOMC was not mainly a rate-hike shock. It was a change in the market’s operating system. The Fed is reducing guidance. Investors must read the data directly. In that world, capital does not vanish, but it moves more narrowly toward real bottlenecks and cash flow.
Sources: Fed statement, Warsh press conference transcript, June 2026 SEP, Wallga Ajae FOMC review, Hankyung Global Market.