WARSH SAID "WE HAVE
WORK TO DO"
In July we published The Warsh Fed Watch and said the new chair had killed forward guidance and that a hike was live. Friday he stood up at Jackson Hole, in his first keynote as chair, and removed the ambiguity. The bond market repriced inside of an hour. Equities barely moved. That gap is the whole story.
What He Actually Said
Three passages did the damage. The first is an admission no sitting Fed chair has made this plainly:
"The responsibility for 65 months of sustained, elevated inflation sits squarely with the central bank."Kevin Warsh · Jackson Hole keynote · Aug 28, 2026
The second is the conditional that traders spent the afternoon pricing:
"We must be confident that underlying inflation is moving to our objective, clearly and at sufficient speed. Otherwise, we have work to do."Kevin Warsh · Jackson Hole keynote · Aug 28, 2026
And the third buried a policy tool that has been in continuous use since 2008:
"Forward guidance as a regular practice was adopted by my colleagues and me during the Global Financial Crisis. It was essential at the time. I believe that the practice has overstayed its welcome."Kevin Warsh · Jackson Hole keynote · Aug 28, 2026
He framed his commitment as being to a discipline, not a decision, and rejected mechanical rate paths outright on the grounds that "our knowledge just doesn't extend that far." He also spent a meaningful block of the speech on AI as a genuine macro variable, citing annualized token sales north of $100 billion, up more than 500% year over year, and the open question of what that does to productivity, employment and capital allocation.
The Scoreboard
| Instrument | Aug 28 level | Move | Read |
|---|---|---|---|
| Sept 25bp hike odds | 59.7% | from 35.4% the prior Thursday | Coin flip became a lean |
| 2-year Treasury | 4.32% | +9 bp | One-month high. The front end did the work |
| 10-year Treasury | 4.71% | +4 bp | Barely participated |
| 30-year Treasury | 5.19% | flat | Did not move at all |
| DXY | 99.57 | +0.4% | Tightening bid |
| Gold (spot) | $4,526 | -1.7% | Real yields up, gold down. Textbook |
| Bitcoin | ~$78,100 | lower | Same trade, higher beta |
| Dow Jones Industrial Average | 53,598.14 | +0.05% | Green |
| S&P 500 (cash) | 7,722.06 | -0.12% | Flat to soft |
| Nasdaq Composite | 26,450.59 | -0.34% | Weakest of the three, again |
Equity levels are late-session cash-index quotes on Aug 28, not official settlements, and not front-month futures prints. YM, ES and NQ track the Dow, S&P 500 and Nasdaq-100 respectively, so do not compare these to the futures levels in our July note. Rate levels per Benzinga's post-speech curve note.
Why A Bear Flattener Is Different From A Selloff
If the market had decided Warsh was going to break something, the long end would have rallied and the curve would have inverted harder. If it had decided inflation was structurally unanchored, the thirty-year would have sold off with everything else. Neither happened. The entire move lived in the two-year.
Translated: traders believe the Fed hikes once or twice, believe it works, and believe the terminal destination is unchanged. That is a narrow, mechanical repricing. It is also exactly the kind of move that punishes long-duration equity multiples while leaving cash-flowing value alone, which is why the Dow closed green on a day the Nasdaq did not.
The Data He Is Actually Looking At
Here is the part most people got wrong last week. If you have been reading CPI headlines you think inflation is nearly beaten. If you read the measure the Fed actually targets, it is not close.
| Measure | 12-month | Released | Note |
|---|---|---|---|
| CPI, headline | 3.4% | Aug 12 (July data) | +0.1% m/m. Cooled from 3.5% |
| CPI, core | 2.5% | Aug 12 | +0.2% m/m. Near target |
| CPI, energy index | +14.7% | Aug 12 | Doing nearly all the headline work |
| PCE, headline | 3.7% | Aug 26 (July data) | Held flat. Expected 3.6% |
| PCE, core | 3.3% | Aug 26 | Held flat. Expected 3.2%. This is the one that matters |
Core CPI at 2.5% and core PCE at 3.3% is an unusually wide spread between two measures that normally track each other closely. They weight housing and healthcare differently and cover different baskets, and we are not going to pretend to resolve the gap here. What matters for your P&L is simpler: the Fed targets PCE. Warsh cited 3.7% on the twelve-month PCE measure from the podium. If you anchor your Fed view on core CPI, you will spend the next two months confused about why the front end keeps selling off.
The energy line is the other half of it. Headline is elevated because the Iran situation keeps cycling between de-escalation and re-escalation around the Strait of Hormuz, and crude has been sitting near $80 with real tail risk above that. A central bank that is already worried about credibility does not get to call an energy shock transitory twice.
The Sept 15-16 Setup
- Target range is 3.50% to 3.75%. A hike takes it to 3.75% to 4.00%.
- July was a 9-3 hold. Three members already dissented in favor of a quarter point. Warsh only needs to move two more votes.
- September carries a Summary of Economic Projections. That means a dot plot, on a committee whose chair just publicly argued that projecting a rate path is bad practice. Watch how that tension resolves. It is the most interesting thing on the calendar.
- Two inflation prints land first: August CPI in mid-September and PCE at the end of the month, after the meeting. The CPI print is the one that can move the September decision.
- There is no forward guidance coming. He said so from the podium. Stop waiting for a leak and start trading the data.
Futures-Trader Implications (Research, Not Advice)
- The front end is the driver now, not the long end. ZT is where the information is. If you only watch the ten-year you will keep missing the move that is actually repricing equities.
- The NQ / YM divergence we wrote about in July is still the regime. Same mechanism, new catalyst. Higher front-end rates compress the multiple on long-duration growth first. Money leaving NQ has been landing in YM's book all summer, and Friday was another clean instance of it.
- Data days are now the volatility events. With guidance dead, the entire information load has shifted onto CPI, PCE and payrolls. Size accordingly and expect the reaction to be larger than it used to be for the same surprise.
- Gold and crypto are trading as one position. Both broke on the same headline for the same reason. If you are long both as a "diversified hedge," you are not diversified.
- The asymmetric risk into September is a hot August CPI. Odds are already near 60%. A hot print takes them toward certainty and hits both legs of the rotation at once, because at that point the long end has to move too.
Key Takeaways
- Warsh used his first Jackson Hole to blame the Fed for 65 months of inflation and to end forward guidance as standing practice.
- September hike odds went from 35.4% to 59.7% on the speech.
- The curve bear-flattened. Two-year +9bp, thirty-year unchanged. A narrow repricing, not a regime break.
- Core PCE at 3.3% is the number he is trading off, not core CPI at 2.5%.
- Sept 15-16 carries an SEP. A dot plot from a chair who dislikes dot plots is the event to watch.
Sources
- Federal Reserve · Keynote remarks by Chairman Warsh at the 2026 Jackson Hole Economic Policy Symposium
- Benzinga · Gold, Bitcoin sink as Warsh lifts September hike odds; curve flattening detail
- Benzinga · Fed September hike odds spike after Warsh's Jackson Hole speech
- CNBC · September Fed decision now a coin flip as hike odds increase
- Yahoo Finance · Market live blog, Friday Aug 28, 2026
- Federal Reserve · 2026 FOMC calendar (Sept 15-16, with projections)
- BLS · Consumer Price Index, July 2026 (released Aug 12)
- CBS News · July PCE held at 3.7%, core 3.3%, hotter than expected
- CNBC · Fed's preferred gauge shows core prices rose 3.3% annually in July
- Chase · September rate hike now expected amid energy shocks; July's 9-3 hold
- Washington Post · Fed chair Warsh says the bank "has more work to do"