PREDICTION CHECK · Macro & Monetary Policy
We Got the Hike Right, but Underestimated the Fed’s Further-Tightening Bias
Macro & Monetary Policy | Prediction Check | FOMC 16 September 2026
The 25bp decision and target range were correct, and the communication framework broadly held. The miss was underestimating the committee’s further-tightening bias, rather than getting this meeting’s action wrong.
Target range raised to 3.75%–4.00%
Participants projecting at least one further hike; not an event probability
Median projected rate at end-2026 and end-2027
Decision: Hit | Communication: Hit | Path signal: Partial miss | Subsequent action: Pending
01 · RESEARCH BRIEF
The one-minute brief
This review separates the decision, chair communication, participant projections and market pricing. ACIS assigned a 92% subjective probability to a quarter-point hike and expected inflation-focused, data-dependent communication without pre-commitment. Those calls held. The dot plot, however, signaled stronger support for further tightening than our base-case interpretation. We record a partial miss on the path signal. The review reflects the 17 September Asian session and preserves the original forecast.
Audio transcript
ACIS correctly anticipated a twenty-five basis point hike and the target range. Inflation-focused communication with data dependence also broadly matched expectations. However, we underestimated the committee’s further-tightening bias and record a partial miss. Sixteen of eighteen participants expected another hike, but that is not an eighty-nine percent event probability or a realized future decision. We will continue checking inflation, employment, long yields and subsequent meetings while preserving the original forecast.
Known facts and open questions
- Classification
- Prediction Check, independently archived without a weekly issue number.
- Evidence
- The decision and projections are official evidence; communication scoring and asset implications are ACIS interpretations.
- Timing
- Event: 16 September. Review: 17 September Asian session. Subsequent policy action remains unresolved.
Decision
Verify the current action and target range.
Chair
Verify inflation emphasis and pre-commitment.
Participants
Verify the median dot and distribution.
Pricing
Assess actual prices separately from forecast accuracy.
02 · THESIS → EVIDENCE → UPDATE
What changed in the thesis?
The current decision was correctly anticipated
- Prior thesis
- 92% subjective probability of a 25bp hike to 3.75%–4.00%.
- New evidence
- The FOMC unanimously raised rates by 25bp to the forecast range.
- Updated view
- Record a hit, while recognizing that a widely anticipated action provides little evidence of an investment edge.
Communication held; participant hawkishness was underestimated
- Prior thesis
- Data-dependent hawkishness without pre-commitment, with a pause and at least one further hike roughly evenly balanced.
- New evidence
- Warsh emphasized inflation, while 16 of 18 participants projected at least one further hike this year.
- Updated view
- The communication framework broadly held. The committee signal was more hawkish than our base-case interpretation: partial miss. Actual further hikes remain to be observed; 16/18 cannot replace the original approximately 50% event probability.
03 · EVIDENCE & ANALYSIS
Evidence and analysis
01｜What was locked before the event?
The original forecast assigned 92% to a 25bp hike, 7% to a hold and 1% to a 50bp increase. It anticipated renewed emphasis on the 2% inflation objective and persistence, without promising another hike. A pause and at least one subsequent hike each carried roughly 45%–55%. The report already described a more hawkish scenario in which the dots and communication pointed to further tightening. Its importance was underestimated; it must not be retrospectively relabeled as the base case.
02｜The outcome: matching action, a higher projected path
The FOMC voted 12–0 for a quarter-point increase. The 18 projection participants and 12 voting members are different populations. Relative to June, September median 2026 projections raised GDP growth from 2.2% to 2.3%, lowered unemployment from 4.3% to 4.1%, and raised PCE inflation from 3.6% to 3.7% and core PCE from 3.3% to 3.4%. The year-end policy-rate median rose from 3.8% to 4.1%; the end-2027 median was also 4.1%.
03｜The chair and participants convey different information
Warsh stressed persistent inflation and the lack of meaningful improvement in underlying summer trends, without mechanically fixing future actions. That fits the data-dependent hawkish framework. The dots instead show each participant’s preferred policy path under their own assumptions. Optionality in the chair’s language can coexist with widespread expectations of further hikes; it does not imply a near-neutral committee.
04｜A resilient economy can still pressure valuations
Higher growth, lower unemployment and higher inflation projections suggest that resilience may give policymakers more room to prioritize inflation. ACIS describes this as a resilient economy with more expensive money. That is an analytical interpretation, not an official conclusion. Earnings can remain supported while higher discount rates reduce the present value of distant cash flows, particularly for expensive, long-duration and externally financed assets.
05｜Error attribution: a hit is not a clean sweep
The current action was correct and the communication framework useful, but the committee’s tightening bias was underestimated. A 50% probability is not a forecast that an event cannot occur, and one event cannot establish a model’s calibration. The partial-miss label concerns the committee and path signal; subsequent policy action remains under review. The manuscript labels the market framework a hit, but lacks consistent event-window price observations. This publication therefore does not score asset-price predictions as verified hits.
06｜What the review changes
Attention shifts from whether September brings a hike to whether inflation, employment and consumption sustain the case for further tightening, and whether long yields independently tighten financial conditions. Future forecasts will distinguish the decision, chair communication, participant distribution and what markets already price. Only this public methodological summary is presented; internal scoring and model details remain private.
04 · INVESTMENT IMPLICATIONS
Industry and asset implications
Quality growth
Earnings and cash flow can support fundamentals without eliminating the valuation headwind from higher rates.
High-beta, long-duration growth
Rich valuations and higher discount rates can act together; dependence on external funding amplifies exposure.
Dollar, gold and digital assets
Rate differentials may support the dollar. Gold and digital assets still depend on real yields, risk appetite and flows; a hike alone does not determine price direction.
Long-duration bonds
An announced hike is not proof of a yield peak. Inflation, term premia and long-end yields remain essential tests.
These are conditional transmission channels, not verified event-window return attribution or trade instructions.
05 · VALIDATION & RISKS
What to verify next
Inflation and growth
Monitor core PCE, CPI, employment and consumption.
Failure signal: Material disinflation or weaker growth would reduce the case for sustained tightening.
Subsequent policy
Record actual later decisions and resolve the at-least-one-further-hike event by year-end.
Failure signal: Participant projections can change; they cannot be settled as realized actions today.
Market response
Compare short yields, the 10-year, the dollar and risk assets using consistent pre- and post-event timestamps.
Failure signal: Without a specified baseline and window, a plausible transmission narrative is not a verified price-forecast hit.
What would change our view?
Dots are not guarantees. Energy, growth or financial-stability shocks can change the path. A correct rate forecast can coexist with an opposite asset-price response. One hit does not establish persistent forecasting skill, and a partial miss should not be obscured by hindsight.
06 · FAQ
Key questions
Was the forecast correct?
The action and target range were hits, and communication broadly matched. Committee hawkishness was underestimated and is recorded as a partial miss.
Does 16 of 18 imply an 89% hike probability?
No. It is a distribution of participant projections, not a market-implied or statistical event probability.
Is another hike now certain?
No. The dots express conditional views of appropriate policy. Actual decisions remain dependent on subsequent data.
Why might a correct hike call fail to make money?
If the action is priced, returns depend on surprises, positioning and financial conditions rather than the action alone.
07 · TERMS & SOURCES
Terms, sources and related research
Key terms
- FOMC
- The Federal Open Market Committee, a principal US monetary-policy decision-making body.
- SEP and dot plot
- Economic projections and individual rate views, rather than a collective policy commitment.
- Basis point
- One-hundredth of a percentage point; 25bp equals 0.25 percentage points.
- Core PCE
- Personal consumption expenditures inflation excluding food and energy, used to assess underlying price pressure.
- Forward guidance
- Central-bank communication about future policy, not necessarily an unconditional promise.
- Higher for longer
- An expectation that policy rates stay elevated for an extended period.
[1] Federal Reserve | 16 September statement ↗
[2] Federal Reserve | September projections ↗
Based on the supplied finalized review and the original public forecast. Official evidence, communication interpretation and ACIS judgments are distinguished. The 16/18 distribution is not an event probability; subsequent hikes remain unresolved. Asset implications are conditional analysis, not a formal event study or return attribution. For research and education, not individualized investment advice.