The Federal Reserve raised interest rates by 25 basis points to a target range of 3.75–4.00%. The vote was 12–0. No dissents. First hike since July 2023.
None of this was the story.
The Dot Plot
Sixteen of eighteen FOMC participants expect at least one more hike this year. Four of those sixteen see two more. Only two members think 3.75–4.00% is sufficient. The median projection: one more 25bp increase before December.
This is the pivot. The question is no longer will they hike — it’s how many more.
| Sep 8 | Sep 11 CPI | Sep 16 Close | |
|---|---|---|---|
| Fed funds | 3.50–3.75% | 3.50–3.75% | 3.75–4.00% |
| 10Y yield | 4.79% | 4.92% | 5.01% |
| 2Y yield | 4.47% | 4.60% | 4.72% |
| DXY | 99.2 | 98.6 | 100.28 |
| Brent | $97.39 | $104.61 | $105.83 |
| Gold | $4,405 | $4,296 | ~$4,330 |
| HY OAS | 266 bps | 284 bps | ~265 bps |
| Hike prob. | 60% | 69% | Delivered |
Read the bottom row of that table. Two weeks ago, this hike was a 60% probability. Today it’s a fact. Read the HY OAS row. Credit spreads went from 266 to 284 on CPI stress — then tightened back to 265 as the hike arrived. Credit is telling you the economy absorbs 4%. The 10-year at 5.01% is telling you something else: that it doesn’t know where this stops.
Three Forces
Warsh was asked why long-term yields are at levels not seen since 2007. His answer had three parts, and each one matters more than the rate decision itself:
“The first is economic strength … Second reason: competition for capital … Third … geopolitics, with hot spots around the world driving long-term yields.”— Fed Chair Kevin Warsh, September 16, 2026
Economic strength. Retail sales this morning came in at +1.2% vs. 0.9% expected. The consumer is spending. The labor market added 162K jobs in August. GDP is tracking above 2%. The economy doesn’t need lower rates.
Competition for capital. This is the line that should get underlined. AI hyperscaler capex — Microsoft, Meta, Google, Amazon — is pulling hundreds of billions in corporate debt issuance through the long end. These companies are competing with the Treasury for capital. The 30-year at 5.36% isn’t just inflation and geopolitics. It’s a secular demand for capital that didn’t exist three years ago.
Geopolitics. Brent at $105.83. Saudi pipeline still offline. Global oil shut-ins at 6.7 million barrels per day. IEA coordinated a 400-million-barrel release and it barely dented the price. The energy channel into inflation is not closing.
The implication: even if one force fades, the other two hold yields up. Oil drops to $85 on a deal? AI capex and economic strength keep 10Y above 4.5%. This is a structural yield regime, not a temporary spike.
What Warsh Didn’t Say
He didn’t say where rates are going. He didn’t provide a dot. He didn’t give forward guidance. He said the Fed would be “quieter.” He said price stability is the goal.
This is the Warsh doctrine: act when the data demands it, explain as little as possible, commit to nothing. For a market that spent four years parsing Jerome Powell’s adjective choices, this opacity is its own form of tightening. You can’t front-run a reaction function you can’t see.
Tomorrow
The Bank of Japan meets September 17. Markets price a 61% probability of a 25bp hike to 1.25%. The ECB already hiked to 2.25%. German 10-year yields are at their highest since 2009. Three major central banks tightening in the same month — the last time that happened was late 2022, and it broke things.
The dollar surged to 100.28, its highest since July 31. Gold reversed. The Dow dropped 631 points. Nasdaq was flat — tech multiples survived because AI capex is both the disease (driving yields) and the cure (driving earnings).
The Regime
The hiking cycle has begun. The dot plot says it isn’t done. Three structural forces — strong economy, AI capital demand, energy supply disruption — keep yields elevated regardless of any single resolution. Sector rotation continues: value over growth, energy over duration, financials over biotech. The floor under rates just moved higher.
Five days ago I wrote that equities and bonds had split, and that when they split going into FOMC, bonds tend to be the one that was paying attention. Today’s 12–0 vote was the bond market’s answer arriving by registered mail.