Ridiculous.

One and the same thing—two different accounts. One side says the probability of a rate hike on September 16 is 90%, while the other side says 70% of economists think they should hold steady.

Everything being said is about the Federal Reserve. The disagreement is so huge that it doesn’t even feel like they’re discussing the same central bank.

First, look at the market’s side of the ledger. CME FedWatch: on September 16, a 25-basis-point rate hike, with a probability of about 90% (as observed). The spark was the August CPI: year-on-year 3.4%, core month-on-month 0.3%, and a higher-than-expected 0.2% (as observed). Principal’s Seema Shah put it very plainly: "August core CPI came in clean and tidy at 0.3%, and with energy prices surging and tensions around Iran, it basically locks in the rate hike next week" (as observed).

Look at what the economists are saying over there. A Reuters survey conducted from September 4 to September 9 found that 70% of economists (65/93) think the Fed will hold steady in September, while 56% think it will hold steady for the entire year (as observed). Note the change: in the August survey, 90% said September would be unchanged. In one month, 20 percentage points flipped, but most people are still on the “hold steady” side.

They’re not looking at the same exam. The market is looking at prices: oil prices broke 100, PPI is 5.4%, and the yield on the 10-year U.S. Treasury is trending upward—every number is calling for another rate hike. Economists look at models: in July, the U.S. economy also lost 23,000 jobs (as observed). The labor market may be more fragile than it appears, and raising rates is essentially another hit to employment that was already soft.

Caught in the middle is the Fed Chair, Kevin Warsh. On August 28, he said the labor market was “stable” and policy should focus on prices (as observed). But he has a style that gives everyone headaches: no forward guidance, and he doesn’t tell you what he’s thinking. In June, he didn’t even submit his own interest-rate forecasts (as observed).

Without guidance, disagreements can only get louder. At the July meeting, three officials had already voted against, arguing for a direct rate hike (as observed). This is a split that’s public—it's not something the media made up.

In the Fed’s context, dissent is a very strong signal—officials generally don’t openly sing against the grain. (analysis) When three people flip at the same time, it suggests the “inflation hawks” are no longer a minority within the committee. This is also one of the reasons the market is confident in pricing in a 90% outcome: it’s not groundless—someone showed their hand early.

Now look at the year-end pricing. More than half of traders bet that October will keep rates unchanged at 3.75%-4%, while fewer than half bet on a December hike to 4%-4.25% (as observed, CME). In other words, the market doesn’t believe in a “one-and-done” hike—it’s betting on the start of a hiking cycle. Over on the economists’ side, 56% think rates stay unchanged for the whole year. The disagreement between the two sides isn’t just about this shudder in September—it’s a fundamental disagreement about what regime the second half of 2026 will be in (analysis).

There’s also an off-the-charts factor: “Chuanzi” has been calling for rate cuts nonstop (as observed). When the president wants cheap money, the central bank wants expensive money, and the White House and the Fed are at odds—2026 has never stopped that tension. What’s interesting is that the more the president calls for rate cuts, the more Warsh may feel compelled to hike—not for any other reason, but to prove independence (analysis).

So this isn’t a technical question of “whether to hike.” It’s a contest over pricing power. The market votes with real money—90%. The economists vote with models—70% saying hold steady. Eventually, someone has to admit they’re wrong, and the side that’s wrong will be wrong expensively. If the market’s 90% pricing misses, the dollar and U.S. Treasury yields will reverse the same day.

One detail worth watching: the dot plot. Each dot represents an official’s view of the future interest-rate path. In the June meeting, Warsh didn’t even submit his own dots (as observed). Whether he submits them this time, and where the dots land, is the most direct evidence for judging just how hawkish “Warsh” really is. He can give zero guidance with words—the dots on the dot plot are impossible to hide (analysis).

Even historical experience has to be discounted. The premise of a “shoe dropping” kind of rebound is that the shoe matches expectations. Under a 90% priced-in scenario, a 25-bp hike is “within expectations.” But if Warsh also releases signals suggesting another hike later in the year, then it’s “a hawk beyond expectations.” The market will vote with declines (analysis). What people fear is never the hike itself—it’s a hike that comes in beyond expectations.

The strongest counterargument has to be put on the table.

The counterargument says: both sides might be right. The Fed can hike once by 25 basis points, while the dot plot turns more dovish and implies that the hike is the last one. Then the market gets its “hike delivered,” and the economists get “that’s it for the year”—both sides take 50-50. No one has to admit they were wrong. Warsh’s zero-guidance style, ironically, creates room for exactly this kind of “bet both ways.”

This counterargument holds up. Historically, the combination of “hawkish rate hike + dovish-style guidance” isn’t rare. And a “shoe dropping” outcome can calm the market on its own.

Under what circumstances would I be wrong?

If there’s a rate hike on September 16, and the dot plot still shows one more hike during the year, along with an upward shift in the 2027 rate path—then the economists’ 70% view that rates will stay steady for the year would all be wrong. The split would be in academia, not in this piece of mine.

Or if the Fed holds steady, and the 90% market pricing turns out to be a collective misjudgment—then the dollar index and the yield on two-year Treasuries would drop sharply. In that case, the premise that “the market is always right” would need to be rewritten.

Or again, if Warsh suddenly gives clear forward guidance at the press conference—then the judgment that “zero guidance amplifies disagreement” would be wrong, showing that he’s just waiting for the CPI data.

Watch three things: the decision statement in the early hours of September 17 Beijing time, the dot plot, and the wording in the press conference. 90% vs 70%—there’s always a number that’s a joke.

(Source: CME FedWatch; Reuters September 9/September 11; USA Today September 14; Schwab Network)