The Federal Reserve held rates steady, but the Treasury market delivered a very different message. Here's what investors should actually be watching.
KEY TAKEAWAYS
The Federal Reserve held rates steady for a fifth consecutive meeting, but the policy statement offered almost no new forward guidance. Markets were left to interpret the Chairman's broader comments rather than any formal change in policy.
Kevin Warsh repeatedly emphasized that tighter financial conditions created by the bond market have effectively done some of the Fed's work. This suggests greater reliance on market-driven tightening rather than additional policy rate increases.
The Chairman questioned the limitations of relying on a single inflation measure and indicated the Fed is reviewing its framework. That introduces additional uncertainty into how future policy decisions may be made.
Treasury markets responded very differently across the yield curve. Short-term yields declined while long-term yields climbed sharply, creating a significant bear steepening despite the Fed remaining on hold.
Higher long-term rates continue to pressure longer-duration assets, particularly technology and AI infrastructure investments that rely heavily on discounted future cash flows.
MY HOT TAKES
Sometimes the most important central bank meetings are the ones where almost nothing changes on paper. Markets often react more to shifting frameworks than to actual rate decisions.
Bond markets frequently identify policy uncertainty faster than equity markets. Watching yield curves can reveal concerns that headlines fail to capture.
Communication is now becoming a policy tool itself. When investors lose confidence in how policy decisions are made, they demand greater compensation for long-term risk.
The Fed appears increasingly willing to allow market interest rates to perform some of the tightening traditionally accomplished through policy moves. That shifts more influence toward Treasury markets.
Investors should spend less time trying to predict the next meeting and more time understanding the evolving reaction function behind future decisions.
You can quote me: "The Treasury market politely told the Federal Reserve it disagreed—and the long end is rarely polite for long."
Nothing burger. If you are from the New York / New Jersey area, you are probably unfamiliar with our famous Greek diners. If you are, hang tight. If you’re not–well, it’s important to understand that a menu at a Greek diner quite literally has what feels like hundreds of selections–it’s like a textbook. There are–in addition to the traditional Greek dishes–food from all corners of the planet. Most have huge salad bars–some even survived the pandemic. Every one of them has freshly baked pies and cakes. I love Greek diners, and you can quote me on that.
Yesterday, there were only 2 things on the menu for the FOMC meeting. Hike rates or don’t hike rates. Pretty basic. They used to offer a third option “cut rates,” but that has been removed from the menu, despite its vast popularity. It is supposedly President Trump’s favorite item. Now, the two remaining items on the menu can be cooked in many different ways. Yesterday, we were hoping to learn precisely how the “don’t hike rates” menu was prepared. We wanted details. What we ended up with was–well, nothing–absolutely nothing.
Yesterday afternoon I watched the server say almost nothing for forty minutes, and as it was going on, I watched the bond market lose its mind. Kevin Warsh delivered a NOTHING BURGER–no dots, no guidance, no hints. Wall Street's talking heads called it boring. Then the 2-year Treasury Note yield fell while the 30-year went to its highest level since 2007. That is not boredom. That is the Treasury market politely informing the Federal Reserve that it disagrees, and the long end (insider speak for long maturity Treasuries) does not do polite for very long. Let's talk about what actually got said yesterday, because none of it was subtle.
Start with the box score, because it's the least interesting part. The Committee held the Fed Funds rate at 3.5% to 3.75% on a nine to three vote, the fifth consecutive hold. Three of Warsh's colleagues wanted to hike a quarter point right now and said so on the record (that is a dissension with a capital D). The statement itself was reaffirmed rather than rewritten, which under this chairman is a policy choice rather than laziness–he has cut the word count of these things roughly in half since he arrived, on the theory that the Fed talks too much. And buried in there, unchanged, sits the single most consequential sentence the Federal Reserve has published this year: “inflation remains above the two percent goal, in part reflecting supply shocks that have driven price increases in certain sectors, including energy.”
Read that again. The Fed wrote down its own excuse. And nobody made a fuss about it, because it has been sitting there since June, and because in this business we have all been trained to skim over the boilerplate.
Now, you would think a Committee that just got a friendly inflation print–June CPI cooled to 3.5% from 4.2%, with core at 2.6%--would take the win and tell you that's why they held. Warsh was asked exactly that. How much did the June core print have to do with not going in July? His answer, in two words, was "not much." Then he said it again, in case anybody missed it. He added that the historic problem with data dependence is the data and the dependence (pretty clever, actually), which is the kind of line I write down in my notebook of Wall Street sayings for later use. Fine. So it wasn't the data. Then what was it?
Here is where the nothing burger starts leaking through the wrapper. Warsh returned, over and over, in nearly every answer, to one theme: the market has already done the tightening. Nominal and real yields are materially higher across the Treasury curve, he said, and some of the inter-meeting increases in market rates rank in the top decile of the last two decades. The markets have done quite a bit even while we haven't done much in 42 days. Financial conditions have tightened, and that tightening gave the FOMC comfort. “Market participants,” he said, “are learning to play the ball, not the referee.” When one reporter tried to call the decision a pause, Warsh said market prices would take the other side of that characterization, meaning, we didn't move, but rates moved, so something happened.
That is a chairman explaining that he has outsourced his tightening to the bond market. Which is a perfectly respectable thing for a central banker to believe. It is also, and this matters, the most dovish sentence available to him, dressed up in the language of market discipline.
Then he told us what the Committee actually argued about, and this is the part the wires mostly ignored. He enumerated four questions. Has the past really passed, after five-plus years of inflation above target. Do shocks that differ in their sources–pandemic supply chains (remember those?), military conflicts, energy disruptions, tariffs, and the AI capex surge–also differ in their effects on output and employment. Are the price increases coming out of the capex boom, memory and logic chips and the AI infrastructure build, evidence of a broad inflationary dynamic, or do we just stare at them because they happen to be standing under the bright streetlight. And finally, if interest rate policy is supposed to be the primary tool, how much accommodation are we actually getting from the balance sheet.
Those are four excellent questions. They are also, and I say this with some affection for the man's intellect, precisely the four questions you ask when you are constructing the intellectual permission to look through inflation you have concluded you cannot fix.Yeah, I just said that. 😉
And then it got interesting. Somebody asked Warsh which measure of inflation he is actually targeting. He gave the correct answer first–PCE, per the strategy statement the Fed publishes every January. Then he volunteered that who knows what that document says after next January, and that his task forces might have something to add. Then, unprompted, a sitting Federal Reserve chairman invoked Goodhart's law and the Lucas critique about his own inflation target: when you declare a measure to be consistent with your objective, you risk making it a lousy measure and a lousy objective. Then he said he is looking at a broader set of inflation data than PCE. And then, and I promise I am not embellishing, he said he was telling us this without fully revealing his cards.
Let that land. The man swore to deliver two percent and not a whisper more, and in the same breath told a room full of reporters that the 2% PCE yardstick is under review, that the review reports to him, that he is already using a different yardstick privately, and that he isn't going to show it to you. Meanwhile the Millers are standing in the checkout line watching the number that actually governs their life, and I can assure you their grocery bill does not care which index the Federal Reserve adopts next January. 😠
But the answer that should have moved markets, and the reason I think the long end did what it did, came from a Wall Street Journal journalist. He asked the only question that mattered. Rates bring inflation down by cooling demand, and that usually shows up in the labor market–so if that isn't the channel you're relying on, what is? Warsh's reply: fine-tuning aggregate demand so that it catches supply is not his mental model, and he doesn't think the Fed is any good at the fine-tuning business.
Now stack that against the reaction function he volunteered a few minutes earlier, unprompted, as if to prove he had one. Any central banker, he said, with labor markets roughly at equilibrium, who sees underlying inflation moving higher is more inclined to tighten. Reuters immediately did the arithmetic out loud: underlying inflation has been moving higher, with one exception, for 63 months, and you haven't tightened. Axios did the other arithmetic – the funds rate is running about 75 basis points below the two-year yield and roughly a hundred below most Taylor rule estimates, so why aren't rates higher today? The answer to all these questions–the ones we really sought–was “I am not saying.”
So here is what the Treasury market heard yesterday afternoon, and I think it heard correctly. It heard a chairman describe a reaction function and then decline to follow it. It heard him say the tool everyone expects him to use doesn't work through the channel that would make it effective. It heard him say the measuring stick is being rebuilt in private. And it heard him say, cheerfully, that he is content to let bond prices do the work.
The front end took that deal at face value. The 2-year fell four basis points to 4.25%, and September hike odds slid from roughly around 75% to about 60%. That is the market saying, fine, we hear you, no hike.
The long end took the other side. The 30-year jumped 10.5 basis points to 5.20%, touching 5.244% intraday, the highest since July of 2007. The 10-year climbed almost seven, to 4.671. The 2-year/30-year yield curve spread went from something like 82 basis points to something like 96 in a single session. A bear steepener into a DOVISH HOLD. And the composition is the tell. The move was skewed toward higher breakeven inflation rather than higher real yields, a rotation from earlier this month, when real yields led. The Dow gave up 1,153 points, with semis and tech leading, because the long bond is the discount rate on every twenty-year AI infrastructure bet financed with borrowed money.
The lazy read is that bonds are pricing more inflation. I don't think that's quite it. Notice that the market didn't cancel the hike–it moved it. December is fully priced. What the curve is pricing is not inflation and not tightening. It is pricing the absence of a mechanism, and charging term premium for the uncertainty.
In fairness, some of that long-end move wasn't Warsh at all. WTI settled up 6.6% at $84.46 the same afternoon after the President said we would be hitting Iran hard, following Iranian missile strikes on American forces. Geo-petro-politics doesn't wait for the Fed's calendar. And breakevens are still below their May highs, so this is a warning shot, not a rout. It's also entirely possible Warsh is right, and that a bond market left alone tightens more honestly than a committee ever could.
He used two metaphors on himself yesterday. He has no “magic wand,” and he is trying to “land the plane.” Both are fair. He also told us, while insisting Jackson Hole is still a blank piece of paper, that there will be plenty of action between September and December. Two CPI prints, a symposium in the thin mountain air, and a Committee that has just discovered it enjoys arguing. Say what you want about a chairman who won't tell you what he's thinking–the long end can't be spoon-fed anymore, so it has gone back to doing what it did before the Fed started narrating everything, which is pricing risk out loud and daring the central bank to prove it wrong. That's not a comfortable conversation. It is, at least, an honest one.
Quite honestly, I feel quite full after consuming that “not hike rates” cooked up as a nothing burger. In the days and weeks to come, it appears–at least based on the servers' statements–that the menu is going to get a lot more items–and a lot more complicated.
YESTERDAY’S MARKETS
Yesterday, the Dow Jones Industrial Average fell 1,153 points, or 2.19%, to close at 51,594–its worst decline since April 2025–while the S&P 500 dropped 1.52%, and the Nasdaq Composite lost 1.74%, finishing more than 10% below its all-time high. The 2-year Treasury yield fell four basis points to 4.24%, the 10-year rose roughly seven basis points to 4.67%, and the 30-year jumped 10.5 basis points to 5.2%, its highest level since July 2007. West Texas Intermediate crude advanced 6.4% to settle at $84.31 a barrel and Brent gained 6.6% to $89.61. The FOMC held the Federal Funds target range at 3.50%–3.75% on a 9-3 vote with three members dissenting in favor of a quarter-point hike.
NEXT UP
Personal Income (June) gained 0.3% higher than estimates but below last month's 0.7% print.
Personal Spending (June) came in 0.3% lower than estimates and last month’s 0.9% increase.
PCE Price Index (June) came in at 3.7% as expected and below May’s 4.1%.
Initial Jobless Claims (July 25th) came in at 197k, below last week’s 188k claims.
Annualized Quarterly GDP (Q2) came in at 1.5%, below estimates and last quarter’s final print of 2.1%.
Important earnings today: Blue Owl Capital, Enterprise Products, Air Products, Valero Energy, Regeneron, PBF Energy, Norwegian Cruise Line, Cinemark, Life Time Group, Hershey, Labcorp, Exelon Energy, KKR, Hyatt Holdings, Martin Marietta Systems, Southern Co, Cigna, Crocs, Huntington Ingalls, American Electric Power, Builders FirstSource, Altria Group, Yum! Brands, Sirius XM, International Paper, Mastercard, Stryker, Corteva, Monolithic Power Systems, Rivian, Apple, ROBLOX, Dexcom, First Solar, and Amazon.