Credibility Shock Persists Into Payrolls Week: Long-End Holds 19-Year Highs as "Credibility Premium" Builds; JPMorgan Turns Hawkish, BofA Says September Hike May Be Required

0. Weekly Arc

The post-FOMC narrative closed the week still trapped in the credibility shock: a 9-3 hold with three hawkish dissents, a Warsh press conference read as dovish execution despite hawkish rhetoric, long-end yields at 19-year highs, and a “credibility premium” building in the curve. September pricing held near two-thirds while JPMorgan turned hawkish and BofA framed a September hike as necessary. The arc ends with markets questioning whether the Fed will match its words with action — ahead of the pivotal July payrolls report.

1. Policy Narrative & Expectations

The net change over the past ~24h is a consolidation of the credibility-shock narrative rather than a new policy shift. September pricing holds near two-thirds: futures at ~67% ^[1], LSEG money markets at 68% after having fully priced a hike pre-meeting ^[2], and CME FedWatch showing an 88% probability of some hike within 2026 ^[3]. The institutional hawkish repricing continues — JPMorgan pulled its expected hike forward to December ^[4][5], while BofA now argues a September hike “may no longer be a policy choice but a necessary action to restore market trust and repair policy credibility” ^[6]. The dominant logic is unchanged: markets doubt the Fed will convert hawkish rhetoric into action, so the long end is absorbing an inflation-risk premium that policy is not yet delivering ^[7][8].

1.1 FOMC Officials’ Remarks

• [ONGOING] Hawkish — Lorie Logan (Dallas Fed), Beth Hammack (Cleveland Fed), Neel Kashkari (Minneapolis Fed): all three dissented in favor of a 25bp hike at the July 29 meeting; no new remarks beyond their published statements ^[9][1].
• [ONGOING] Chair Warsh — hawkish rhetoric, dovish execution: At his second “Fed Day” press conference (7/30), Warsh insisted there is no “soft inflation target” above 2% ^[10][11], said the Fed “cannot tolerate inflation running above target for more than five years” ^[12], and left a hike “entirely possible” if inflation stays elevated across the forecast horizon ^[10] — yet he declined to answer whether the Fed would hike in September and specified no data conditions for action ^[10]. Per USA Today, he understands the public’s “impatience” but has no “magic wand,” and: “Market participants are learning to play the ball, not the referee” ^[13]. Per the FT, he “all but refused to explain the Fed’s decision to do nothing, despite persistently above-target inflation” ^[7]. He also reiterated that markets have already delivered substantial tightening on their own and that the Fed will not be bound by market prices ^[12][1], and remained equivocal on whether PCE stays the preferred inflation gauge ^[4][5]. Marginal shift vs prior stance: no change in substance from the 8/1 briefing — the new development is analytical, with HSBC’s Ryan Wong saying the market is adding a “credibility premium” ^[7] and observers calling the press conference “one of the most confusing” in recent memory (Seema Shah) ^[13].

1.2 Policy Signals & Institutional Communication

• [ONGOING] FOMC outcome: 9-3 hold at 3.50%-3.75%, fifth consecutive hold; three same-direction dissents for a 25bp hike — the first such triple dissent since 2016 ^{[14][15][9][1]}. The statement reiterated the ample-reserves regime and flagged energy/supply shocks from Middle East escalation while noting strong productivity and capital investment ^[9][1].
• [NEW] Four major questions under study: the statement listed the policy impact of five years of high inflation; the effects of multiple external shocks and AI investment on output and employment; how shocks affect price changes; and the role of the policy rate and balance sheet in achieving price stability ^[12].
• [NEW] JPMorgan hawkish repricing: chief US economist Michael Feroli now expects a 25bp hike in December rather than H2-2027, sees the funds rate staying in the 3.75%-4.00% range afterward, and calls the risk of an additional September hike “real” ^[4][5].
• [ESCALATED] BofA: with resilient employment and sticky inflation, “the case for last year’s easing cycle is weakening while the case for a September hike is strengthening”; if coming weeks’ inflation data offer no dovish support, a September hike becomes a credibility-restoring necessity ^[6].
• [NEW] CICC July survey: 45% of fixed-income investors expect no move this year, 38% expect at least one hike, 17% expect a cut — more polarized than last survey; a majority now see the 10-year ending 2026 in the 4.5%-5% range ^[3]. CICC itself judges further hikes unlikely, citing weak Q2 growth and cooling core inflation ^[3].
• [NEW] AI working group: the Fed’s Productivity and Employment Working Group has three advisors who are all positive on AI — supporting CITIC’s view that the US equity AI narrative persists ^[14][15].
• [NEW] Political crosswind: President Trump tweeted after the FOMC meeting that “Warsh wants rate cuts,” undercutting the Fed’s hawkish communication ^[12].
• [NEW] Communication regime: Warsh said he has not decided whether his Jackson Hole speech will be a macro report or a September-December action plan; press conferences will continue through year-end ^[12].

2. Key Data & Market Read

• [ONGOING] June CPI — below expectations: the soft print (headline at 3.5% with a monthly decline, unemployment at 4.2%) would normally have been a clear positive, but Warsh’s PCE-gauge equivocation and JPMorgan’s rapid hawkish repricing turned the reaction into unease ^[4][5][12]. Narrative impact: a dovish data point overwhelmed by the credibility debate.
• [NEW] PCE — easing: US PCE inflation eased from May, slightly relieving Fed rate-hike expectations ^[11].
• [NEW] Q2 inflation breadth (analytic): Huachuang notes Q2 CPI ran 3.9% y/y and core 2.7%, with about 62-63% of CPI and core items rising more than 2% — a supply-shock character, not 1997-style overheating ^[1].
• [NEW] Inflation expectations (analytic): Cleveland Fed’s 10-year inflation expectation estimate was ~2.38% in H1 2026 vs ~3.29% in H1 1997, and the policy-rate/natural-rate gap was -25bp vs -47bp in 1997 — Huachuang argues the economy falls short of 1997 preemptive-hike conditions on every metric except the inflation headline ^[1].
• [NEW] July payrolls preview (projection): BofA forecasts a slightly below-consensus headline but resilient private hiring, the unemployment rate ticking up on a labor-force participation rebound, and wage growth showing no clear inflation pressure — a soft-landing-supportive profile that nonetheless strengthens the September-hike case ^[6].
• [NEW] August 7 payrolls — the pivotal test: the July nonfarm payrolls report is the most important data point of the week; JOLTS/ADP/claims and ISM PMIs (a gauge of business conditions under the high-oil-price shock) precede it, and weakness would further lower hike pricing ^[2][10].
• [NEW] EIA crude inventories: US crude inventories fell 7.2 million barrels last week to the lowest level in nearly eight years, deepening supply-tightness concerns ^[2].

3. Financial-Conditions Signals

• [ONGOING] Long-end at 19-year highs: after the meeting the curve bear-steepened — the 2-year yield fell while the 30-year jumped to 5.2%, the highest since 2007; the 10-year rose 6.9bp to 4.673% and the 30-year 10.9bp to 5.197% ^[7][1]. The FT calls the magnitude an “unusual reaction” following a hold, with the only similar instance after the November 2010 FOMC (QE2) ^[7].
• [ONGOING] Breakevens and the driver debate: the 5-year breakeven jumped 8bp on July 29 as the curve steepened ^[14]; post-press-conference moves — higher real long-end yields, wider breakevens, a steeper curve — suggest investors doubt the Fed’s long-run inflation anchoring ^[6]. BofA counters that the long-end rise may reflect repricing of larger fiscal deficits, stronger growth, higher risk premia, or higher inflation expectations rather than tighter financial conditions ^[6]; CICC reads it as markets pricing higher long-term inflation and policy risk ^[8].
• [NEW] “Credibility premium”: HSBC’s Ryan Wong says the market is adding a credibility premium to the curve ^[7].
• [NEW] Dollar: rising inflation expectations are weighing on the dollar, providing the window for Japan’s Ministry of Finance to intervene on July 30 ^[14][15].
• [NEW] Credit & housing: the 30-year fixed mortgage rate stood at about 6.66% after the July meeting, up from 6.47% in mid-June ^[13].
• [NEW] Liquidity & positioning: overseas investment banks estimate Korea’s leveraged-ETF unwind is basically complete and hedge-fund deleveraging is nearing its end ^[11]; Huachuang argues the market’s spontaneous tightening via higher Treasury yields supports the Fed staying on hold ^[1].
• [NEW] Treasury supply: the quarterly refunding plan and borrowing estimates arrive August 3; HSBC says the all-maturity rise in yields is the key factor constraining the Treasury from expanding medium/long issuance ^[2].
• [NEW] De-dollarization: CICC argues de-dollarization remains the trend — rising US government debt erodes the safe-asset consensus behind Treasuries — despite the recent stabilization in the dollar’s reserve share ^[8].

4. Global Central-Bank Linkages

• [NEW] PBoC: at its 2026 H2 work conference (8/2), the PBoC reaffirmed a moderately loose policy using reverse repos, the medium-term lending facility and government bond trading to keep liquidity ample; it will add an overnight reverse-repo instrument and narrow the temporary overnight repo rate band, and keep comprehensive social financing costs low; end-June aggregate social financing grew 7.4% y/y and M2 8.0% ^[16]. The Politburo said operations will pay more attention to “force, pace and timing” ^[8].
• [ESCALATED] ECB: euro-area July inflation rose again, breaking the ECB’s 2% target on energy, with services and core inflation ticking up; with Q2 growth beating expectations, markets are betting on a September hike ^[2]. SEB says the conflict’s duration is the decisive variable ^[2].
• [ONGOING] BOE: held at 3.75% with three of nine MPC members voting for a 25bp hike; the Bank says high energy prices have not yet triggered second-round inflation effects; market pricing for total 2026 hikes was trimmed from 37bp to 30bp ^[2].
• [ONGOING] BOJ/Japan: the MoF intervened on July 30 on the dollar’s weakness; CITIC warns that if the yen appreciates quickly and triggers carry-trade unwinding, high-valuation tech and semiconductor stocks could face periodic downward pressure ^[14].

5. Asset Implications

This section is inference — no [N]. Anchored to the facts above.

Quadrant	Current probability tilt	Key asset implication	Anchoring narrative
Growth↑ + Inflation↑	Rising	The reflation leg is alive but capped at the long end: oil holds above $90 with Hormuz flows down ~61% and crude inventories at ~8-year lows; TIPS/breakevens are the cleaner expression of inflation risk than nominals; the “credibility premium” keeps long nominal bonds under pressure	§1.2; §3
Growth↑ + Inflation↓	Falling, contested	The Goldilocks window is data-supported (soft June CPI, cooling PCE, BofA’s soft-landing payrolls profile) but credibility-blocked; front-end duration is favored while the survey consensus sees equities rangebound at high levels	§2; §1.2
Growth↓ + Inflation↑	Rising	The stagflation tail lives where the credibility shock sits: bond vigilantes selling long-end to force a hike, CICC flagging renewed stagflation risk in developed economies, and risk assets facing greater downward-adjustment pressure if long-end rates rise further	§1.2; §6
Growth↓ + Inflation↓	Falling	A disinflation-led bond rally is blocked at the long end by the credibility premium and fiscal-supply constraints; CICC sees a low probability of further hikes on weak Q2 growth and cooling core inflation, arguing for the belly rather than long duration	§2; §8

Stock-bond correlation call: The regime remains inflation-driven positive correlation — and the bear-steepening response to a hold, with the 2-year falling while the 30-year jumps to 19-year highs, is one of the most equity-unfriendly formats of a bond bear market. The FT notes the only similar post-hold reaction followed November 2010’s QE2, underscoring how unusual the current credibility repricing is. So long as markets doubt Warsh’s reaction function — September pricing near two-thirds yet a large “no hike” minority, and a chair who refuses to pre-commit — stocks and long bonds will tend to fall together. A flip to negative (growth-driven) correlation requires either a credibility-restoring signal (a September hike or a decisive Jackson Hole commitment) or payrolls soft enough to collapse the inflation-risk premium. Directional call: positive correlation persists through the August 7 payrolls and the September FOMC.

Risk-budget implication: Overweight front-end and belly duration — with September priced near two-thirds but Warsh refusing to pre-commit, the elevated hike premium remains vulnerable to another hold surprise; CICC’s survey majority expecting no move supports this. Overweight curve steepeners, since the credibility premium and fiscal-supply story keep pressure on the long end while the front end prices out near-term hikes. Underweight long-end nominal duration; express long-duration inflation risk via TIPS/breakevens given the driver-narrative split between inflation repricing and fiscal term premium. In credit, the mortgage channel (6.66% 30-year) shows the tightening is transmitting through rates, not spreads — favor IG quality over HY beta, and keep AI-exposed high-yield underweight. In equities, the late-session rotation out of heavy tech/Internet weights into HALO and defensive sectors (with $6.5bn of semiconductor ETF outflows) argues for quality over high-duration growth, particularly while yen carry-unwind risk is live. Gold is the two-sided hedge: capped near-term by real yields and hike expectations, but supported by record Q2 central-bank buying (289 tonnes, +62% y/y) — a dip-buying rather than momentum asset. On the dollar, inflation-expectation-driven weakness is the near-term driver, but intervention (Japan MoF already active) argues for options rather than spot expression.

6. Contrarian & Tail Risks

• Consensus fragility — “talk hawkish, act dovish” is being priced as a credibility gap: bond vigilantes, dissatisfied with the Fed’s delay in hiking, are selling long-end Treasuries to force a hike ^[14]; CICC warns that outsourcing tightening to markets in an above-target-inflation environment could easily undermine confidence in Fed credibility ^[8]; BofA warns that if the press conference fails to clarify the reaction function, uncertainty shifts to markets and can ultimately lead to unanchored inflation expectations ^[6]. Falsifiable: if payrolls print strong and the Fed still holds, the “behind the curve” narrative entrenches; if payrolls are soft and the Fed hikes anyway, both duration and equities get hit.
• Consensus fragility — September pricing is internally inconsistent: sourced probabilities range from ~57% (CITIC, down from 62%) to ~67% (futures) to 68% (LSEG), with CME showing 88% for any 2026 hike — the dispersion itself is a sign of a market unsure whether the Fed will convert words into action ^{[14][2][3][1]}.
• Consensus fragility — the 1997 Greenspan analogy is weak: Huachuang argues only the inflation headline exceeds 1997; inflation expectations are better anchored (~2.38% vs ~3.29%) and the policy-rate/natural-rate gap is not looser than 1997 — the precautionary-hike tail has risen after the divided FOMC, but it remains a tail ^[1].
• Second-order — fiscal/term-premium channel: the August 3 quarterly refunding arrives with long-end at 19-year highs; HSBC notes policymakers will carefully avoid any move that raises term premia, but BofA’s read that the long end is repricing fiscal deficits/risk premia implies supply is part of the story, not just inflation ^[6][2].
• Second-order — energy two-step: OPEC+ meets August 2 with most participants expecting another 188,000 bpd increase, but Capital Economics argues Middle East conflict limits members’ export/production capacity, not willingness — with Saudi Arabia facing both Houthi shipping blockades and Hormuz disruption, output decisions have limited practical impact ^[2].
• Second-order — yen carry and AI leverage: if the yen appreciates quickly, carry-trade unwinding pressures high-valuation tech and semis; CICC flags AI valuation-bubble worries and negative wealth effects on final demand as the AI chain’s downside channel ^[14][3].
• Source quality control: September-hike probability figures conflict across sources (57% vs 67-68%) and are presented as measured on different instruments (swaps/futures/CME) — treat as a range ^{[14][2][3][1]}. The Trump tweet is single-source reporting ^[12]. The reporter-cited “100% September probability” in ^[12] conflicts with CME’s 88% within-2026 figure and should be discounted ^[3][12]. The “credibility premium” framing is one strategist’s attribution via FT (Ryan Wong, HSBC) ^[7].

Appendix: Additional Sources

• ^[15] CITIC Securities — market split “short-end dovish, long-end hawkish”; AI narrative persists
• ^[9] Huaron Securities — 9-3 vote widening internal divisions; asset pricing shifting from neutral-rate anchor to geopolitical risk premium
• ^[11] Chengtong Securities — global liquidity marginally improving; market expected to build a bottom
• ^[17] Bloomberg — Warsh’s silent treatment has S&P traders bracing for wild swings
• ^[8] CICC Research — Fed hawkish forces strengthened; long-end rise reflects higher long-term inflation and policy risk

This report is a macro-mechanism analysis, not investment advice.