The Fed’s higher long-run rate suggests expensive money may be the new normal—not merely a temporary phase of the current economic cycle.
KEY TAKEAWAYS
The Fed unanimously raised its target rate by 25 basis points to 3.75%--4.00%, despite political pressure to cut. That unanimity delivered a more forceful signal than the widely anticipated hike itself.
Short-term yields rose sharply while the 10-year yield retreated, flattening the yield curve. The move suggests markets are pricing both further tightening and greater long-term Fed credibility.
The Fed eliminated its previously projected 2027 rate cut and raised its longer-run rate estimate. That points toward a structurally higher cost of money, not simply a temporary tightening cycle.
The dot plot implies one additional hike, but futures markets are pricing a path closer to three. That disagreement has major implications for bonds, equities, mortgages, and housing.
AI infrastructure spending may be less sensitive to interest rates than traditional capital investment. If so, monetary policy may struggle to cool demand without a sustained equity-market decline.
MY HOT TAKES
The hike itself was almost irrelevant because markets already expected it. The real message came from the unanimous vote, the erased 2027 cut, and the higher estimate of the economy’s neutral interest rate.
The falling 10-year yield was not a rejection of the Fed’s hawkishness. It was a tentative vote of confidence that a Fed willing to inflict pain now may prevent worse inflation later.
The futures market believes the Fed is understating how much tightening will ultimately be required. A gap of nearly two additional hikes is too large to dismiss as ordinary forecasting noise.
Investors may be underestimating how persistent expensive money will become. The Fed’s higher longer-run dot means pressure on valuations and borrowing costs could survive well beyond this particular cycle.
AI spending may be weakening the traditional relationship between higher rates and lower investment. If competitive survival matters more than financing costs, only a meaningful stock-market drawdown may force executives to reconsider their plans.
You can quote me: “Being ‘left behind’ in tech is the equivalent of death.”
Fed…up! Fed Funds up, that is. The thermostat in my apartment and I have been in a quiet disagreement since Sunday, when the first properly cool morning of the fall rolled through New York and I bumped the heat on weeks earlier than I'd promised myself I would (my wife wasn’t paying attention–she likes it cooooold 😉❄️). It's a small capitulation–a few degress, one week–but it's the kind of decision you make constantly without noticing you're making it: a little more input, in exchange for a little less discomfort, until one day you look up and the number on the dial has crept somewhere you didn't plan for it to go. Kevin Warsh made his own version of that adjustment yesterday, except his dial doesn't control a furnace–it controls the cost of money for the entire economy, and unlike my sneaky move, he told everyone in advance exactly how much he was planning to turn it.
The headline part was almost a non-event, which is exactly how the Fed likes it when it can manage that. The FOMC voted unanimously to raise the federal funds rate by 25 basis points, to a target range of 3.75%-4.00%, the first hike in over three years. The unanimity is the detail worth sitting with, not the number itself. This came against a backdrop of very public pressure from the White House to cut, NOT hike–Trump said afterward he still had confidence in Warsh while taking a swing at what he called a "hostile" Fed board, and a unanimous vote in that environment isn't a reluctant compromise. It's a committee that agreed, without dissent, that the thermostat needed to move. Warsh's own language gave the game away. He said the hike "removed a dose of accommodation"--Fed-speak for policy is still, even after this move, somewhat stimulative, and further tightening remains on the table.
His press conference did nothing to soften that. Thirty minutes, one consistent message–a booming economy, inflation still sitting stubbornly above target, and a case for more restrictive policy rather than less. He again declined to submit his own dot to the Fed's rate projections, a pattern now rather than a one-off, consistent with his stated preference for leaning less on forward guidance and more on letting markets do their own thinking. In an interview, a former FOMC governor put words to what the market was already feeling, calling the unanimous vote "a very firm signal." When the man who used to sit in that room tells you the room just sent a message..well, you listen.
Here's where the "market already knew this was coming" thesis actually earns its keep, because the rate decision was priced. What wasn't priced was how yields would behave once Warsh started talking. The 2-year note–the maturity that reacts fastest to what the Fed is likely to do next–took the hawkish surprise straight on the chin, climbing to 4.740% on FOMC day, its highest level since 2024, before easing back slightly to 4.71% as I drank espresso number 2 this morning. The 10-year told a more interesting story. It had already been on an eight-day rising streak walking into the meeting, touching 5.025% on the 16th, and then it did something the 2-year didn't do–it snapped that streak this morning, falling about four basis points to 4.98%. That's the pivot point of this whole episode. The bond market wasn't just pricing a hike anymore by Wednesday morning. It was starting to price Warsh's credibility–the idea that a Fed willing to hike unanimously into political pressure is, over the longer run, a disinflationary force, even if it hurts in the short run.
The yield curve did exactly what a curve is supposed to do when the front end reprices harder than the back end. It flattened. The 2s/10s spread (the gap between 2-year and 10-year yields, and the market's shorthand for how much tightening versus slowing is priced across time) compressed from roughly +32.8 basis points on September 14 to +28.5bp yesterday and to about +27.3bp by this morning. I read a Bloomberg note this morning which flagged the tail risk plainly–keep hiking at this pace and you're flirting with an inverted curve, which historically hasn't been a friendly signal for what comes next. The dollar strengthened right alongside it, with the Dollar Index (which tracks the U.S. Dollar’s value relative to a basket of trading partner currencies) climbed from 99.39 on the 14th to 100.25 yesterday, breaking back above the psychologically loaded 100 level, before easing slightly this morning as risk appetite partially recovered. That dollar strength is already leaning on Asian currencies with the recently newsworthy Yen in particular, ahead of Friday's Bank of Japan meeting, with strategists flagging USD/JPY weakening toward its 200-day moving average near 158.
Stocks did what stocks do when the tone exceeds what was priced for the guidance, not the decision. The S&P 500 fell to 7,551–its lowest level since July–the Dow dropped 1.2%, and the Nasdaq 100 slipped 0.6%. This morning, futures clawed some of that back, helped along by Brent crude dropping more than 2% below $104 a barrel, which gave the market a reason to think inflation pressure might ease from at least one direction. Read that recovery for what it likely is: the market's first tentative vote that a Fed willing to be the bad guy now is preferable to one that loses control of inflation later.
Now to the part that actually matters more than the vote–the Summary of Economic Projections, or what the Fed-heads call the SEP, the Fed's quarterly report card on where it thinks rates, growth, and inflation are headed. The rate decision was expected. The SEP was the surprise, and the single sentence that should stop you cold is this one: the rate cut the Fed had penciled in for 2027 back in June is gone. Erased. The 2027 median dot now sits at 4.125%, identical to the new 2026 median–up from 3.625% in June, a 50-basis-point swing in a single quarter. The 2026 median itself rose from 3.75% to 4.125%, implying one more quarter-point hike before the year is out. Sixteen of eighteen Fed officials now see at least one more hike this year, four of them penciling in two. And buried a little deeper in the table is the detail I think matters most for anyone thinking in years rather than quarters–the longer-run dot, the Fed's estimate of where rates settle once the cycle is fully done, moved up to 3.25% from 3.0625%. That's not a cyclical call. That's the Fed quietly telling you it believes the economy's new normal, the "neutral" cost of money, is structurally higher than it thought in June. And it's doing this while projecting real GDP growth of 2.3% in 2026, 2.4% in 2027, and 2.2% in 2028. In other words, no growth slowdown priced in anywhere, despite all this tightening.
Which brings us to the gap that actually runs the show right now–the space between what the Dot Plot says and what futures markets believe. The Fed's own map shows one more hike this year, then a hold through all of 2027. The futures market is pricing something considerably more aggressive: roughly 57% odds of a hike by October, 75% by December, with the implied path running through March and April of 2027 to a cumulative move that works out to close to three additional hikes from here–not one. That's not a rounding error. That's the market telling the Fed, politely, that it doesn't fully believe its own homework.
Run that gap through every corner of the market and the implications compound rather than cancel out. Bonds sit at the center of it. The front end is most exposed if the roughly three-hike futures path proves right, the long end pulled in two directions at once, since more hikes are bearish for duration while a credible inflation-fighting Fed is, ironically, disinflationary over time, which is precisely why the 10-year rallied this morning even as the 2-year stayed elevated. The curve is likely to keep flattening, with real inversion risk if the Fed ultimately delivers more than the dot plot admits, and that higher longer-run dot is a headwind for bond valuations that doesn't go away when this particular hiking cycle ends. Equities face a version of the same math: if the market's more aggressive path is closer to correct, the discount rate applied to future earnings rises, and multiple compression hits long-duration growth names hardest–rate-sensitive sectors like utilities, REITs, and consumer discretionary feeling it most directly, while tech and growth's relative resilience on FOMC day suggests the market hasn't fully priced that risk in yet. Mortgages are the most direct pain point for ordinary families, with the 10-year near 5%, the 30-year mortgage rate is likely already sitting in the 7%-plus range, and with no 2027 relief anywhere on the Fed's own map, that's not a temporary discomfort, it's the new weather. Homebuilders are already feeling the dual squeeze in real time–Lennar down 2.04% over the past five trading days, the iShares Home Construction ETF down 1.32%, PulteGroup down 0.95%, D.R. Horton down 0.78%--buyers priced further out of affordability while builders' own financing costs climb (builders use shorter-term credit more directly tied to now-upwardly mobile Fed Funds), with no near-term relief from the Fed's projections to point to.
Here's the part I don't think is getting enough attention, and it's the reason this hiking cycle might behave differently than the ones you remember. The Fed's whole theory of the case is that higher rates make capital more expensive, which cools spending, which eventually cools inflation. That's the transmission mechanism–it's worked, more or less, for my nearly four decades in the biz. Buried in the SEP is a GDP path that doesn't budge despite the tightening, and I don't think that's an accident–a real slice of that resilience is AI infrastructure spending, and Warsh's own framing of a "booming" economy with inflation stuck above target is, whether he says it out loud or not, gesturing at demand-pull inflation pressure with a source. Here's the problem with expecting the traditional mechanism to work on it: hyperscaler capital expenditure right now isn't being underwritten on a cost-of-capital basis. It's being funded out of a survival calculus–whoever blinks first in the compute buildout risks losing the platform war for a decade, and no CFO wants to be the one who explains that decision to a board five years from now. That makes the spending close to rate-insensitive. Higher borrowing costs barely register against numbers that large when the alternative is falling behind permanently. Being “left behind” in tech is the equivalent of death. The one thing that probably would curb it–and, with it, pull some of the real non-energy demand pressure out of the inflation picture–is a serious, sustained equity decline, the kind that changes a board's actual risk tolerance for multi-year CAPEX commitments. A one-day drop to July lows followed by a next-morning bounce is not that. Not even close. And nothing in the past forty-eight hours suggests it's coming.
So here's where I've landed with my thermostat, one degree warmer than I meant to be. The Fed just turned its own dial, told you exactly how much, and the market spent two days deciding it doesn't fully believe the number on display. The rate hike was never the real story–the erased 2027 cut, the higher neutral rate, and a nearly three-hike gap between what the Fed says and what the market believes are. And the piece nobody's pricing correctly yet is the one lever that could actually close that gap–a real stock market drawdown–sitting conspicuously absent from a tape that bounced the very next morning. When I got out of bed this morning, my apartment was freezing–arctic in my assessment. I caught a glimpse of my thermostat enroute to my espresso machine–it was set at 68 degrees (F). My wife obviously changed it when I wasn’t paying attention. 🥶
YESTERDAY’S MARKETS
Yesterday, the S&P 500 fell to 7,551, its lowest close since July, while the Dow dropped 1.2% to 51,461, and the Nasdaq declined just 0.01%. The 2-year Treasury yield rose to 4.74%, its highest level since 2024, and the 10-year touched 5.0%. The dollar index (DXY) climbed to 100.25, moving back above the 100 level. The FOMC voted unanimously to raise the federal funds rate 25 basis points to a target range of 3.75%--4.00%...in case you missed it.
NEXT UP
Initial Jobless Claims (September 12th) is expected to come in at 207k, slightly above last week’s 206k claims.
Housing Starts (August) may have climbed by 6.7% after falling by 12.8% in July.
Building Permits (August) probably slipped by -1.5% after increasing by 4.3% in the prior period.
Pending Home Sales (August) may have fallen by -3.9% over the past year slightly worse than the -2.5% annual decline in July.