A plain-English guide to reading a contradictory Fed message.

Federal Reserve communications typically strive for internal consistency. The statement, the Summary of Economic Projections, and the press conference are all crafted to send a single coherent message about the committee's reaction function. When they don't, the market has to decide which part to weight more heavily. This piece is the framework for reading a contradictory Fed message: why contradictions happen, why the Q&A typically wins the interpretation battle, and how to weight the different signals when they conflict.

The paired analysis today reads Wednesday's FOMC as exactly this case: hawkish prepared remarks, dovish Q&A, market interpretation decisively dovish. This piece explains why the market's read was reasonable given the pattern and what to watch as the interpretation matures over the following sessions.

Where contradictions come from

Fed messages become contradictory for four specific reasons:

1. Committee dispersion that the prepared statement smoothed over. The FOMC statement is a consensus document; language is negotiated to accommodate a range of committee views. When the Chair takes Q&A, personal views can leak in ways that were not in the negotiated statement. If those personal views are more dovish or hawkish than the median, the message becomes contradictory.
2. Chair inexperience with the format. New Chairs (Warsh at his second meeting, Powell at his first, Yellen at her first) sometimes produce Q&A responses that inadvertently soften the prepared message. Chairs improve over their first year, and the market discounts early-tenure messages differently.
3. Genuine policy uncertainty. When the committee itself is uncertain about the next move, the message reflects that uncertainty. Prepared remarks tend to hedge conservatively (defensive to inflation risk), while Q&A tends to acknowledge the uncertainty more openly.
4. Deliberate optionality. Occasionally, the Chair wants to preserve maximum flexibility for the next decision. This produces messages that are deliberately open to interpretation in either direction; markets read them as contradictory but they are actually intentional.

Wednesday's Warsh press conference likely combined factors 1, 2, and 3: a hawkish committee median (three dissents plus half the dots) that the Chair softened in Q&A because he is early in his tenure and because genuine policy uncertainty is high given the mixed inflation and labor-market data.

Why the Q&A wins

When the prepared statement and the Q&A conflict, markets typically weight the Q&A more heavily. Three reasons:

• The Q&A is unscripted. The Chair's spontaneous responses reveal thinking that the prepared statement does not. Unscripted content is typically seen as more informative because it cannot be edited by staff.
• Specific answers to specific questions. The Q&A addresses hypotheticals ("would you hike if inflation stays above 3 percent?") that the prepared statement does not directly answer. The specific answers are what markets actually need to price the future path.
• Hedging patterns are readable. Chairs use specific hedging language ("we're not thinking about that," "we haven't decided that," "we are prepared to") that carries technical meaning. The Q&A is where this language actually surfaces; the prepared statement rarely uses such specific formulations.

What to watch when reading a Q&A

Five specific dimensions of Q&A that carry the highest signal:

1. Verb tense. "We're not thinking about" is dovish about the specific direction. "We haven't decided" is neutral. "We are prepared to" is hawkish. The tense choice carries technical meaning.
2. Bar-setting language. "It would take substantial" is hawkish (high bar to change). "We would consider" is neutral. "We are watching for" is dovish (low bar to change).
3. Time-frame specificity. "In the near term" implies September meeting weighting. "Over the medium term" implies December+ weighting. "Down the road" is essentially undated.
4. Committee-dispersion acknowledgments. "Some members thought" reveals internal splits. "We were unanimous" (or "broadly agreed") is a decisive signal about consensus. Explicit acknowledgment of dissent is the strongest signal that internal debate is active.
5. Referent selection. When asked about inflation, Chairs can reference specific inflation measures (core PCE vs headline CPI). Which one they choose reveals which measure the committee is actually watching most closely.

Warsh's specific patterns Wednesday

Analyzing Warsh's press conference against the five dimensions:

• Verb tense on rate hikes: "We discussed a range of scenarios" (neutral). Not "we are prepared to hike" (which would be hawkish).
• Bar-setting on further tightening: "Materially higher inflation than we currently project" (high bar; dovish).
• Time-frame specificity: Warsh used "in the coming months" repeatedly (broadly neutral, could accommodate September or later).
• Committee-dispersion acknowledgments: Warsh acknowledged the three dissents but framed them as "reasonable minds can disagree" rather than as a signal of imminent policy shift. Neutral-to-dovish.
• Referent selection on inflation: Warsh consistently referenced "core PCE" as the primary Fed measure, and specifically noted "core PCE has moved lower over the past three months." This is a dovish framing because core PCE has indeed softened while headline CPI has been more sticky.

Under the framework, four of five dimensions read dovish; one (bar-setting) reads strongly dovish. The market's overall dovish interpretation is well-supported by the specific Q&A patterns.

The credibility discount

When Fed communications are contradictory, markets impose a credibility discount. The specific form is the term-premium repricing (compensation for policy-path uncertainty over longer horizons). Wednesday's 30-year yield hitting the highest level since 2007 is the direct market signal of the credibility discount being applied to Warsh's first substantive FOMC as Chair.

Credibility discounts typically decay over 3-6 months as the Chair produces more consistent communications and markets calibrate to the new voice. Powell's first year (2018) is the closest recent parallel: contradictory early communications, a temporary term-premium expansion, then gradual normalization as the market learned his patterns.

Reading the follow-through

The market's Wednesday reading was decisively dovish. The Thursday-Friday follow-through depends on two specific inputs:

• Post-meeting Fed speeches (Thursday-Friday). If regional Bank presidents deliver hawkish speeches that push back on the dovish press-conference read, the Wednesday move partially reverses. If they align with the dovish read, the Wednesday move extends.
• Data prints (Thursday Q2 GDP, Friday PCE). A hot GDP or a hot PCE reprices the hawkish tail; a soft reading confirms the dovish read.

Related references

• FOMC-day tape: the base framework for reading the meeting release end-to-end.
• Fed committee splits: the internal-dispersion framework relevant to reading the 9-3 vote.
• Fed Chair testimony: the specific dynamics of scripted vs unscripted Fed communications.
• Term premium: the specific channel through which the credibility discount showed up in the 30-year yield.

Contradictory Fed messages are a specific type of event that the framework has to read carefully. The Q&A typically wins the interpretation battle, but the specific patterns in the Chair's hedging language are what actually drive the market's read. Warsh's first substantive FOMC produced a strong-form example of this pattern; future meetings will show how the market's calibration evolves.