Week recap · Macro · FOMC · rates · June 19, 2026

We said the surprise skewed up. Warsh skewed it down.

Author Brandon Leon Posted 2026-06-19 (Juneteenth; week closed Thursday) Coverage Week recap · grading the FOMC call

TL;DR: Three days ago we made a falsifiable call into Warsh's first FOMC: the hawkish hold was priced, so the surprise would skew dovish-of-fear and the rally would hold. We were wrong. Warsh out-hawked the tape — the dots flipped to a 2026 hike, the 2-year jumped 16bp (the exact gauge we named as the kill-switch), and the S&P fell to 7,420. The call is dead by our own rule. The one consolation: we named the kill-switch, war-gamed this exact path, and sized it a lean — and equities shook the whole thing off by Thursday. Below: the honest grade, the lesson, and the refreshed book into Micron's June 24 print.

Where we landed

The week was a round trip into a reversal. A U.S.–Iran peace deal collapsed oil and a soft core CPI carried the S&P to ~7,550 by Monday June 15. Then Kevin Warsh took the chair. On Wednesday June 17 the Fed held at 3.50–3.75% — 12–0, as expected — but the projections did the talking: the 2026 median dot rose to 3.8% from 3.4%, flipping from an implied cut to an implied hike; the statement was gutted to 130 words with the easing bias struck; and Warsh declined to submit his own dot, instead announcing task forces to overhaul the Fed.¹

The market took it as a regime change. The 2-year Treasury yield ripped 16 basis points to 4.216% — its biggest jump on a Fed day since March 2008 — while the 10-year barely moved (4.45%, −1bp): a violent bear-flattening. The S&P closed Wednesday down 1.21% at 7,420.10, the Nasdaq off 1.34%, the Dow down 507.²

So the call died. On June 16 we wrote: "a 2-year yield jump of 15+ basis points on the statement = genuine hawkish shock = call dead." It jumped 16. The "Warsh underwhelms the hawks / pain-trade-is-up" thesis was wrong; the bear scenario we assigned 25% is what printed. We own it.

The honest twist: equities didn't stay scared. Thursday June 18 the S&P rose 0.99% back to roughly 7,490 — above our 7,450 line — with the Nasdaq up nearly 2% on a chip-led comeback and the Russell 2000 up 2.12%, even as the front end stayed repriced.³ That doesn't save the call — the regime read was wrong and the gauge we named fired. But it tells you where the damage landed: in rates, not in the equity bid.

1. The week in one arc

Five trading days, two regimes. The week opened with the dovish setup intact and closed with a hawkish Fed re-pricing the front end of the curve.

Wed June 10 — CPI. May headline hit 4.2% Y/Y, a three-year high, but core was soft at 0.2% M/M (below the 0.3% expected) and 2.9% Y/Y. An energy print, not a broadening one.⁴
Thu June 11 — PPI. Hotter still on the headline (+1.1% M/M, 6.5% Y/Y, highest since November 2022) but four-fifths of the goods move was energy.⁴
Mon June 15 — Iran peace, oil collapse. A preliminary U.S.–Iran framework — 60-day ceasefire, an agreement to reopen the Strait of Hormuz — sent WTI from ~$95 into the low $80s, a multi-month low. The energy-inflation premium drained out; the 10Y fell to ~4.4% and the S&P ran to 7,554.⁵
Tue June 16 — our call. Into the FOMC we read the hawkish hold as fully priced and called the surprise to skew dovish-of-fear, with the rally holding 7,450. Falsifiable on a sub-7,400 close or a 2-year jump of 15+bp.
Wed June 17 — the Warsh shock. Hold, but dots to a hike, easing bias struck, 2Y +16bp, S&P −1.21% to 7,420. The call's kill-switch fired.¹
Thu June 18 — the shrug. S&P +0.99% (~7,490), Nasdaq ~+2% chip-led, Russell +2.12%; the front end stayed elevated. Markets closed Friday for Juneteenth.³

Chart 1 — S&P 500: the peace-rally, then the Warsh reversal

Up to 7,554 on the Iran deal. Down to 7,420 on the hawkish Fed. Back to ~7,490 by Thursday.

S&P 500 path, June 11–18. The Iran peace deal carried it to a June 15 close of 7,554; Warsh's hawkish debut knocked it to 7,420 on June 17; a chip-led rebound recovered roughly 1% Thursday to ~7,490 (derived from the verified +0.99% off 7,420.10). The June 18 level dipped below and then climbed back above our 7,450 line — but the call was already dead on the 2-year. Sources: TheStreet daily summaries (June 15, 17, 18); CNBC.

Simply: Stocks rallied early in the week on an Iran peace deal that crushed oil prices. Then the new Fed chair delivered a much tougher message than expected, and stocks dropped Wednesday. By Thursday they'd mostly recovered — but the bond market stayed rattled. Markets were closed Friday for Juneteenth.

2. The decision: a hold that read like a hike

The Federal Open Market Committee left the target range at 3.50–3.75% by a unanimous 12–0 vote — exactly the priced outcome. Everything that moved markets was in the projections and the communication.¹

The dot plot was the headline. The 2026 year-end median rose to 3.8% from 3.4% in March — not the “drift toward zero cuts” the consensus expected, but a flip to an implied quarter-point hike above today's midpoint. Nine of eighteen officials now see rates higher by year-end, and the Committee raised its 2026 PCE inflation forecast to 3.6% from 2.7%. The statement itself was cut to 130 words from 341, with the forward-guidance / easing-bias language removed entirely. By the close, fed funds futures had moved to price a 2026 hike.

And then there was the Chair. In his first meeting, Warsh declined to submit a dot to the very grid the market was bracing for, and used the press conference to announce task forces to overhaul major Fed operations — an early signal that he intends to reshape the institution, not just set a rate. The net message the market heard: inflation is the problem, higher-for-longer is the base case, and a cut is off the table.

Chart 2 — The 2026 rate path: March vs. June

The median dot flipped from a cut (3.4%) to a hike (3.8%). That's the regime change in one number.

FOMC Summary of Economic Projections, 2026 year-end median federal funds rate: 3.4% in March (an implied cut from the current ~3.625% midpoint) to 3.8% in June (an implied hike). Nine of eighteen officials project a higher rate by year-end; the 2026 PCE inflation forecast was raised to 3.6%. Sources: Federal Reserve SEP (June 17, 2026); CNBC; investingLive.

Simply: The Fed didn't change rates, but its own forecasts flipped from expecting a cut this year to expecting a hike. The statement was stripped down, the “we might cut” language deleted, and the new chair signaled he wants to run a tougher, leaner Fed. Markets read it as: no cuts coming — if anything, a hike.

3. The 2-year tell — the kill-shot

On June 16 we said the hawkish surprise, if it came, would register in one place: the front end. We made it the explicit falsifier — a 2-year yield jump of 15+ basis points on the statement marks a genuine hawkish shock and kills the call. The 2-year jumped 16 basis points to 4.216% in the hours around the decision, its largest move on a Fed day since March 2008.² The 10-year, by contrast, fell a basis point to 4.45%. That divergence — front end up hard, long end flat — is a textbook bear-flattening: the market pricing tighter policy now without raising its long-run growth or inflation expectations. It is exactly the signature of a credible hawkish shock.

We could not have asked for a cleaner refutation. The gauge we nominated did the refuting. There is no reading of the tape on which this call survives.

Simply: We told you the one number to watch that would prove us wrong: the 2-year Treasury yield. It spiked 16 basis points — the biggest Fed-day move in 18 years. That's the market saying “tighter policy is coming.” Our call was wrong, on exactly the gauge we picked.

4. The scorecard, graded honestly

We carried a call, a scenario set, a rate range, and three trip-wires into the meeting. Here is how each closed.

June 16 item	What we said	Outcome	Verdict
Forward call	Warsh underwhelms; surprise dovish-of-fear; S&P holds ≥7,450	Warsh out-hawked; dots to a hike; S&P to 7,420 on the day	KILLED
Falsifier — 2Y +15bp	that = hawkish shock = call dead	2Y +16bp to 4.216% (biggest Fed-day move since 2008)	FIRED
Falsifier — S&P <7,400	close below 7,400 (6/17–18)	7,420 Wed, ~7,490 Thu — held the floor	NOT BREACHED
Scenario weight	bear (hawkish shock) 25%	the 25% bear is what printed	WRONG TILT
Dots read	drift hawkish toward zero cuts	median flipped to a hike (3.8%)	UNDERSHOT THE HAWK
Easing bias struck	expected struck	struck (130-word statement)	RIGHT
Front-end duration stance	no longer underweight at the front	front end repriced +16bp against us	WRONG
CapEx-raise call (dies 6/25)	open	still open; MU prints 6/24	OPEN

No way to dress this up: the central call was wrong, the scenario tilt was wrong, and the duration stance was wrong. The only items we got right — the hold and the struck easing bias — were the consensus, not our edge. The S&P falsifier (sub-7,400) technically did not breach, and equities recovered above our 7,450 line by Thursday; but grading the call on the equity floor while ignoring the rates gauge we ourselves nominated would be exactly the kind of self-serving scoring this letter exists to avoid. The 2-year fired. The call is dead.

Simply: We got the big call wrong. The Fed was more hawkish than we bet, and the gauge we said would prove us wrong did. Stocks held up better than the bond market, but that's not a save — our read of the Fed was the call, and it missed.

5. The lesson — we named the risk, then faded it

This miss is worth more than a generic “the Fed surprised us,” because we wrote the right answer into the piece and then bet against it. From June 16, in the section war-gaming the other side: “a brand-new Chair has every incentive not to look soft on inflation in his first appearance, even when the dots and the decision land exactly as priced. If the 2-year jumps on the statement, that is the channel that did it — and the call is wrong.”

That is what happened, almost to the word. The error was not analytical blindness; it was probability. We identified the single most incentive-aligned risk on the board — a hawkish debut to establish credibility — and assigned it 25% when the incentive structure argued for more. The lesson we are writing down: when a risk is that cleanly aligned with the actor's incentives, it is not a tail. Do not fade it. A new chair's first meeting is, structurally, a hawkish event until proven otherwise.

The discipline that kept this from being expensive: we made the call falsifiable, named the exact kill-switch, and sized it as a lean rather than conviction. That is the difference between a wrong call and a damaged book. The position came off when the 2-year moved; it did not have to be argued out of.

Simply: We actually wrote down the exact way we could be wrong — a tough new chair proving himself — and then bet it wouldn't happen. The takeaway: a brand-new Fed chair's first meeting is a tough-message event by default. Don't bet against it. Because the call was small and falsifiable, the mistake was cheap.

6. The refreshed book: a hawkish-regime Fed into Micron

The framework changes because the regime did. As of June 17 the base case is no longer “dovish drift” — it is a Warsh Fed that has put a 2026 hike on the table and stripped the cut from its own forecasts. We position to that.

Rates — the front end leads

The damage was concentrated, and that is the actionable read. The 2-year repriced ~16bp; the 10-year did not move. We want to be underweight front-end duration — the stance we wrongly abandoned a week ago — while the long end is closer to fair: a flatter curve is the path of least resistance under a credible-hawkish chair. Refreshed ranges through July: 2-year 4.05–4.45%, 10-year 4.35–4.70%.

Equities — the bid that absorbed the shock

The most important tell of the week may be Thursday: a 16bp front-end shock, a flip to a hiking Fed, and the S&P recovered nearly all of it within a day on a chip-led bid. That is a market whose marginal driver is the AI-infrastructure earnings cycle, not the policy rate — the structural call we have carried since May. It does not make equities cheap; it makes them resilient to rate shocks in a way the bear case underrates. We hold the AI-infrastructure breadth basket; we do not chase it here.

Three trip-wires, refreshed

Trip-wire #1 · The hike gets pulled forward

A July or September hike enters pricing, or a Fed official names a specific meeting.

The dots put a hike on the 2026 table; the risk is the market pricing it sooner than December. Watch the 2-year above 4.45% and any Warsh-aligned official naming a date. Response: add to the front-end-duration short; trim equity beta into strength.

Trip-wire #2 · Hyperscaler CapEx confirmation

META, MSFT, AMZN, or GOOGL raises 2026 CapEx — or guides cautiously — before the carried-over call expires June 25.

Unchanged and now urgent: the equity resilience thesis rests on the AI-capex cycle holding. A raise validates the breadth basket; a cautious tone is the tell to fade it. Response: see the forward-call status below.

Trip-wire #3 · Iran peace unwinds / oil round-trips

The 60-day framework cracks or WTI sustains back above $90.

The disinflation help the market is leaning on came from the oil collapse. A reversal hands the hawkish Fed its excuse to act sooner, and re-arms the inflation forecast it just raised. Response: re-add energy hedge; expect the front end to lead higher again.

The forward call — status

Carried-over call · hyperscaler CapEx raise (dies June 25)

No raise yet. Micron's FQ3 print — Wednesday June 24 after the close — is the next read on the AI-capex cycle, the day before the call expires.

Micron sits above a $1 trillion market cap and up roughly 244% year-to-date — priced for the supercycle to keep compounding.⁶ Its gross-margin guide and HBM commentary are the cleanest tell we get on whether hyperscaler demand is still accelerating. A strong print plus a hyperscaler raise validates the equity-resilience read; silence on both, and the breadth basket faces the same multiple pressure under a higher discount rate. Falsification of the CapEx call: June 25 with no raise = call dies; we do not double down.

Simply: The Fed is now the tough one, so we lean against short-term bonds (where the pain landed) and leave longer bonds alone. Stocks proved resilient because the AI spending cycle — not the Fed — is driving them, so we hold our AI-infrastructure names without chasing. The next big test is Micron's earnings June 24.

Get the next note by email → All writing

Sources & footnotes

June 17, 2026 FOMC (Kevin Warsh's first meeting as Chair): federal funds target range held at 3.50–3.75% by a 12–0 vote. Summary of Economic Projections: 2026 year-end median federal funds rate raised to 3.8% from 3.4% in March (an implied hike); 9 of 18 officials project a higher rate by year-end; 2026 PCE inflation forecast raised to 3.6% from 2.7%. Statement shortened to ~130 words from 341, with easing-bias / forward-guidance language removed. Warsh declined to submit a dot and announced task forces to overhaul Fed operations; futures moved to price a 2026 hike. Sources: Federal Reserve statement and SEP (June 17, 2026); CNBC (“Fed holds rates steady,” June 17, 2026); investingLive (“Warsh rewrites the Fed playbook,” June 17, 2026). ↩
Market reaction, June 17, 2026: 2-year Treasury yield rose more than 16 basis points to 4.216% — the largest move on a Fed meeting day since March 2008; 10-year yield fell ~1 basis point to 4.453%. S&P 500 closed down 1.21% at 7,420.10; Nasdaq Composite down 1.34% to 26,021.66; Dow Jones Industrial Average down 507 points (−0.98%). Sources: CNBC (Treasury yields, June 17–18, 2026); TheStreet “Stock Market Today” (June 17, 2026). ↩
June 18, 2026: S&P 500 rose 0.99% (to roughly 7,490, derived from the close off the prior day's 7,420.10); Nasdaq Composite climbed nearly 2% on a chip-led recovery; Russell 2000 up 2.12%. Treasury yields little changed (2-year ~4.18%, 10-year 4.453%) with the front end remaining elevated. U.S. markets closed Friday June 19 for the Juneteenth federal holiday. Sources: TheStreet “Stock Market Today” (June 18, 2026); CNBC (“S&P 500 closes higher, Nasdaq climbs nearly 2%,” June 17–18, 2026). Exact June 18 closing index level to be re-verified against the print before reuse. ↩
May 2026 CPI (BLS, June 10): headline +0.5% M/M, +4.2% Y/Y (three-year high, in line with consensus); core +0.2% M/M (below 0.3% expected), +2.9% Y/Y. May 2026 PPI (BLS, June 11): final demand +1.1% M/M (vs. 0.7% expected), +6.5% Y/Y (highest since November 2022); roughly 80% of the goods advance from energy. Sources: BLS Consumer Price Index and Producer Price Index releases; CNBC; Morningstar. ↩
U.S.–Iran preliminary peace framework announced June 15, 2026: 60-day ceasefire and an agreement to reopen the Strait of Hormuz over the coming weeks. WTI crude fell from ~$94.58 on June 8 to roughly $81 by June 15, a multi-month low. S&P 500 closed June 15 at 7,554.29; 10-year Treasury yield ~4.43%. Sources: NPR; PBS NewsHour; Axios; TheStreet (June 15, 2026). ↩
Micron (MU): FQ3 2026 results scheduled after the close on June 24, 2026; stock above a $1 trillion market cap and up roughly 244% year-to-date as of mid-June. Approximate; re-verify against the live quote and the 8-K on the print. Sources: StockTitan (earnings date); StockAnalysis; CNBC. ↩

Nothing on this page is investment advice. Forward-looking statements are scenarios, not promises. See disclaimer.