When Fed Governor Christopher Waller pushed back on the case for a September rate hike, the market's implied odds of that hike did not merely soften. They fell roughly 12 percentage points within minutes, from around 67% to about 54.6%, according to CME FedWatch. That single, fast-moving number underpins nearly every "traders now see a 54% chance of a hike" line in Fed coverage, including our own recent pieces on Waller's remarks. It is worth understanding exactly what that number is and how one Fed governor's comments can move it so quickly, because it is not a poll or a forecast. It is a price.

Where the number actually comes from

FedWatch is built entirely from 30-day Fed Funds futures contracts traded on the CME. These contracts are priced at 100 minus the rate the market expects the average federal funds rate to be over a given month. So if a contract is trading at 95.67, the market is implying an average rate of 4.33% for that period.

By comparing that implied rate to the Fed's current target range and effective rate, CME's model works out how much of a hike, hold, or cut is priced in, and distributes that across probability bands for each scenario at each upcoming FOMC meeting. This is the same principle that underpins all derivatives-based probability measures: markets aggregate dispersed views into prices, and those prices can be inverted into implied likelihoods. When Waller spoke, traders did not file a survey response. They bought and sold Fed Funds futures contracts based on his comments, and those trades repriced the contracts in real time — which is what FedWatch reflects.

The Waller move, step by step

Coming into Thursday, the odds of a September hike had already climbed to somewhere between 60% and 67%, largely on the back of Fed Chair Kevin Warsh's hawkish Jackson Hole speech in late August. That was the baseline FedWatch was pricing when Waller took the podium.

Waller then argued that the three-month annualised inflation trend — down to 3.05% from 4.76% in February — was a more reliable guide than headline annual figures, and described the pace of that improvement as "encouraging." Within minutes of those comments crossing the wires, Fed Funds futures were bought up enough to push the implied September hike probability down to about 54.6%. That is the entire mechanism in one example: a change in the language a Fed official uses about inflation directly altered what traders were willing to pay for a futures contract, and FedWatch translated that price change into a probability shift that could be pointed to and quoted.

It did not stop there. In the days since, as more positioning has come through, the implied odds have drifted further, toward something closer to a coin flip near 50%. None of that reflects a change in what the Fed has actually decided. The FOMC has not met yet. It reflects a continuous stream of traders repricing futures contracts as they absorb Waller's comments alongside everything else in the data flow.

What this tells you, and what it doesn't

Because it is derived from live futures pricing, FedWatch updates continuously, and the Waller episode is a clean illustration of how mechanically reactive it is. A single official at a single event moved the market's implied odds by 12 points in minutes. That is precisely what makes the tool useful for coverage like this: it converts a vague description such as "Waller sounded relatively dovish" into an actual, trackable number, showing readers not just that sentiment shifted but by how much.

Where it gets overread is when that number is treated as a forecast of what the Fed will actually do, rather than as a snapshot of current positioning. Waller's own remarks made this distinction explicit. He did not rule out supporting a hike; he said he would consider one if the upcoming August CPI print comes in hot. The FedWatch number moved because traders reassessed the odds, not because the underlying decision has been made. A reading of 50% does not mean the Fed's decision is a coin flip. It means futures traders currently see both outcomes as similarly probable — and that view could itself move sharply again on the next data point, such as the August CPI release that Waller himself flagged as decisive for his own thinking.

The practical takeaway

When you see a FedWatch-derived probability cited in coverage, including ours, the Waller episode is a useful reference point for what that number actually represents: the market's live, tradable view of Fed policy, capable of shifting materially in minutes on a single official's comments. It is a genuinely valuable real-time barometer, but it is built to move, and it will likely keep moving right up until the FOMC's 15–16 September decision itself — with each intervening data release and Fed speaker offering a fresh chance for traders to reprice.

Reading the FedWatch screen itself

1. The framing question at the top. Before any of the data, CME's own page states plainly what the tool answers: how likely interest rate traders currently think it is that the Fed will change its target rate at each upcoming meeting, based on pricing in 30-day Fed Funds futures. This line is worth internalising, since it is the whole point of the tool in one sentence, and it is why every number underneath is described as a probability "according to traders" — not a Fed decision or a forecast from CME itself. The page also carries a media attribution note asking that any rate probabilities used in reporting be credited to "CME FedWatch" specifically, which is why our own pieces attribute the figures that way.

2. Meeting date tabs. Across the top of the tool sits a row of upcoming FOMC meeting dates (16 Sep 26, 28 Oct 26, and so on). Each tab is a separate probability calculation for that specific meeting, since the Fed Funds futures contract expiring around that date is what the odds are drawn from. Selecting a different tab does not just relabel the same numbers; it pulls an entirely different contract's pricing.

3. Meeting information table. This row shows the mechanics behind the number: the contract used (for example ZQU6), when it expires, its current mid-price, and prior volume and open interest. The mid-price is the input converted into an implied rate, as explained above (100 minus the price). Prior volume and open interest are worth a glance because they indicate how much trading activity sits behind the number. This is not a thin market.

4. Probabilities summary. The three headline figures (ease, no change, hike) are the top-line output of the whole model for that meeting. In the Waller example, this is where the 54.6% and 50.2% readings actually live. This is the number that gets quoted in coverage, and it is a straight read of what sits directly above the bar chart.

5. The bar chart. This is the same probabilities summary shown visually, broken out by specific target rate range rather than just ease/hold/hike. Each bar corresponds to a possible target rate band (for example 350–375 versus 375–400), and its height is the probability, as priced by the futures, that the Fed lands there after that meeting. When a chart shows two bars close in height, near 50/50, that is the "coin flip" read discussed above.

6. The historical comparison table. This is arguably the most useful section for a story, and the one most likely to be skipped. It shows the same probabilities as they stood now, one day ago, one week ago, and one month ago. This is exactly how you would document a move like Waller's: reading across a row shows the probability shifting from 67.2% a month ago, to 64.6% a week ago, to 36.8% one day ago, to 49.8% now, for example. That is the moving picture behind the static number quoted in any single article, and it is the section to check whenever you want to show a shift rather than just a snapshot.