Hawkish Warsh regime and higher-for-longer
What changed
The headline signal — how confidence moved from the previous snapshot, and why.
Stepped 0.87 to 0.89 with the band held at 0.04. The defining event was not a speech but a failed intervention: Treasury tripled its long-dated buyback to $6B specifically to cap borrowing costs, and
the 30-year rose through 5.3% to 5.307% and the 10-year to 4.85% in the same session, with
TLT making new lows this morning. A long end that sells off into a tripled buyback is repricing the path itself, not reacting to guidance, and that is the durable-regime claim of this thesis being validated by the market rather than asserted by the chair. The data cooperated:
payrolls at +162,000 against +53,000 expected with
55,000 of upward revisions removed the labour-softness case for holding, and
September hike odds moved to 60.6%. The band is deliberately NOT tightened, for one reason:
core PPI at +0.2% against a +0.3% forecast is the first genuinely softer inflation reading this thesis has faced, and with CPI tomorrow and the decision six days out, a hold framed dovishly remains live and would be the first real contrary evidence.
The thesis
The claim and where confidence stands now.
Kevin Warsh used his first FOMC (June 16-17 2026) to establish an explicitly hawkish, price-stability-first regime, succeeding the dovish-tilt uncertainty that the retired Fed leadership transition policy uncertainty thesis tracked. The Committee
held the target range at 3.50-3.75% on a unanimous 12-0 vote but
dropped its easing bias, sharply shortened the statement, and published a dot plot that removed the prior 2026 cut and put a HIKE on the table - nine of eighteen participants see at least one hike this year, six see multiple, only one sees a cut.
Warsh withheld his own dot, said "inflation is a choice", and committed the Committee to delivering price stability, deliberately devaluing forward guidance and reorienting the Fed around inflation. The market priced it immediately -
the 2-year yield jumped ~16bp on the statement,
the 10Y rose to ~4.49% and the 30Y to ~4.93%, and
equities posted the worst new-chair Fed-day reaction since 1994. The thesis is that this is a DURABLE regime, not a one-meeting tone: higher-for-longer policy with a live hike bias and no 2026 cut, a steeper-than-priced rate path, and a Fed that tolerates above-target growth and energy inflation rather than easing into it. Update July 2: Warsh reinforced the regime - at
the Sintra forum July 1 he emphasized Fed independence, rebuffed calls for cuts, and pledged to deliver price stability, warning against tolerance of above-2% inflation - and markets now price
~66% odds of a December hike in the Fed's own dots, with
core PCE at 3.4% validating higher-for-longer. Update September 3: the Jackson Hole keynote resolved hawkish.
Warsh said inflation is too high, signalled rates may have to rise, named PCE as the gauge he will act on, and said the better summer prints do not show underlying improvement, and
called broad financial conditions not restrictive against a strengthened economy at 4.1% unemployment.
September hike odds went 35.4% to 57.5% in a session and on to roughly two-thirds by September 1, with
the 10-year at 4.81%, its highest since October 2023, and the 30-year at 5.26%. Update September 10: the long end repriced through a direct attempt to cap it.
Treasury tripled its long-dated buyback to $6B, three times the $2B communicated, to stem rising borrowing costs - and in the same session
the 30-year rose five basis points through 5.3% to 5.307% while the 10-year hit 4.85%, its highest since October 2023, with
TLT down to 81.00.
August payrolls at +162,000 against +53,000 expected and
55,000 of upward revisions removed the labour-softness case for holding, taking
September hike odds to 60.6%. The one contrary datapoint is
core PPI at +0.2% against +0.3% expected, which keeps a dovish hold live into the September 16 decision.
Drivers
The underlying macro forces this thesis expresses - the loading mean is how much each force drives the thesis, the stddev our confidence in the mapping.
Supporting evidence
Typed, citation-backed observations across time, grouped by strength. Hover a point for the claim.
What would invalidate this
The machine-evaluable conditions that would falsify the thesis.