Beyond the December hike: why did the probability of a January hike jump from 18% to 54% in two days?

Executive summary

A rate hike at the Federal Reserve meeting on 8 and 9 December is no longer disputed in markets. The market-implied probability of a hike of any size reached 96.55% on Polymarket on the evening of 11 November. This probability is derived from contract prices on a prediction market, where participants trade contracts on the outcomes of future events, and EL7 records these prices daily. It stood at 80.05% before the inflation release on 10 November. But this figure has limited significance. A single December hike is exactly what 12 of 18 members projected in September’s dot plot, a chart showing each participant’s assessment of the appropriate future interest rate. Markets have therefore caught up with the Fed, rather than moved ahead of it.

The more significant shift in expectations concerns what follows December. Between 9 and 11 November, the probability of a second hike at the January meeting rose from 18.2% to 54.15%, and the probability of a March hike from 36.45% to 59.1%. Adding up what these markets price in for each meeting produces a path approaching two quarter-percentage-point hikes by March, or around 54 basis points. A basis point equals one hundredth of a percentage point. The median, the middle value when projections are ranked, of members’ September projections was one hike in December followed by unchanged rates throughout 2027. The gap is therefore one hike, which the Fed has yet to signal.

Nor will its chair signal it before the meeting. Kevin Warsh said at the 28 October press conference that he does not believe in forward guidance, a central bank’s public commitment to a future rate path, except in crises. Markets therefore infer the path from other members, most clearly Lorie Logan, president of the Dallas Fed, who said on 10 November that the committee needs “at least two additional hikes” before pausing. This broadly matches market pricing.

But our reading adds a factor that prices do not clearly reflect. Four voting members change in January under the annual rotation, with three of this year’s most hawkish members leaving their voting seats: Beth Hammack, Neel Kashkari and Logan herself. Of the five dissenting votes calling for rate hikes in 2026, four came from regional Fed presidents whose voting terms end in January. This does not mean a January hike is unlikely: the Federal Reserve Board in Washington, comprising the Fed’s governors, holds seven of twelve votes. It does mean, however, that January’s voting committee will be less inclined to raise rates than the 2026 committee.

In bonds, the two-year yield rose 17 basis points to 4.73% on 10 November, but remained below its 28 September peak of 4.92%. The front end of the yield curve, which reflects yields on bonds approaching maturity, has therefore yet to price in everything it did six weeks earlier. Before the inflation release, leveraged money, investment funds that use borrowing to increase their positions, held a net short position, meaning their sell contracts exceeded their buy contracts. This position was close to its largest recorded size in futures linked to the Secured Overnight Financing Rate (Secured Overnight Financing Rate (SOFR)), the overnight borrowing rate secured by Treasury securities. Futures are agreements entered into today to settle obligations at a later date; these funds were betting on higher rates before the data release.

Three dates will determine the size of the gap before 9 December: Personal Consumption Expenditures (PCE) data on 25 November, the spending data used to derive the Federal Reserve’s preferred inflation measure; the jobs report on 4 December; then the new dot plot itself. A shift by just two members in the December dot plot would close the gap with the market.