📚 THE SETUP · WAR DAY 85 · THE WEEK THAT REDREW THREE MAPS · WARSH DAY 1 · DOW RECORD · SPACEX $1.75T IPO · BTC BELOW $76K · LONG WEEKEND IRAN WATCH
50,579
Dow Record Close
Week ended at all-time high
Day 1
Warsh as Fed Chair
First FOMC June 16–17 · 25 days
44.8
UMich Final May
Record low · All-time since 1966
$1.75T
SpaceX IPO Target
June 12 · $4.69B Q1 revenue · SPCX
📚 The Week in Framework — Three Maps Redrawn. One Long Weekend to Navigate.
War Day 85 — Saturday Analysis
The Week That Redrew Three Maps Simultaneously. Here Is What Each One Says About Next Week.
Editorial Desk
The week of May 18–22, 2026 produced three structural changes that will define the markets for the remainder of Q2. Each arrived through a different channel. Together they create a framework for Tuesday’s open and the five sessions that follow.
The AI Supercycle Map. Nvidia confirmed $81.6 billion in Q1 revenue, 85% year-over-year growth, and a $91.0 billion Q2 guide that exceeded even the buy-side whisper range. Workday beat and raised margins. Zoom beat and authorised a $1 billion buyback. Ross Stores beat on revenue and earnings. The AI thesis was not just confirmed at the semiconductor layer — it was confirmed across the enterprise software and consumer spending layers simultaneously. Then SpaceX filed its S-1 and revealed $4.694 billion in Q1 revenue, with Starlink contributing $3.257 billion. OpenAI and Anthropic announced IPO intentions. The AI infrastructure cycle is now broad enough to support multiple simultaneous IPOs at trillion-dollar valuations. The map is larger than it was a week ago.
The Monetary Policy Map. Kevin Warsh was sworn in as Federal Reserve chair on Friday. His swearing-in at the White House — the first Fed chairman to take the oath at 1600 Pennsylvania Avenue since Alan Greenspan in 1987 — was not ceremonial detail. It was the visual statement of the institutional tension he will navigate: executing independent monetary policy with a president watching from across the room. JPMorgan has forecast rates unchanged through mid-2027. Wolfe Research has established that 40 basis points of the yield rise is structural and only 10–15 would reverse on an Iran deal. The monetary policy map now has a named face on it and a date attached: June 16–17 is the first Federal Open Market Committee (FOMC) meeting under his chairmanship. The map tells one story: higher for longer, with no near-term cuts regardless of who is asking.
The Consumer-Equity Divergence Map. The University of Michigan’s final May consumer sentiment reading came in at 44.8 — the lowest in the survey’s 60-year history. On the same day, the Dow Jones Industrial Average set an all-time closing record. Both readings are simultaneously accurate. They describe two separate economies running in parallel on the same country. Equity markets are pricing the forward AI earnings curve and diplomatic optimism. Consumers are pricing gasoline at four-year highs, mortgage rates at 6.51%, and groceries running at 3.8% inflation. Walmart, Deere, and Ross Stores all beat Q1 earnings but missed or pulled full-year guidance. The war is in corporate annual plans now — not as an abstract risk factor but as a confirmed cost input with a dollar amount attached. The gap between these two maps will close only when oil falls sustainably. That requires Iran. That requires navigating a long weekend where the HEU directive is in effect and the military remains on standby.
📍 Where Things Stand
S&P 5007,473.47 · 27pts from ATH
Dow Jones50,579.70 · RECORD
WTI Crude$96.36 · Week’s low
Bitcoin~$75,830 · Below $76K daily
WarshDay 1 · FOMC Jun 16–17
IranHEU directive in effect
Samsung strikeDay 2 of 18
UMich May final44.8 · Record low
SpaceX IPOJune 12 · SPCX · $1.75T
Iran Scenarios for Tuesday
Deal confirmedOil $80–$85 · S&P gap up · BTC above $80K
EscalationOil $115+ · S&P gap down · BTC below $74K
No new signalTuesday opens near Friday’s close · AI earnings week begins
🏭 Warsh & Monetary Policy — Three FOMC Paths · The Inheritance · June 16–17
June 16–17 · First FOMC · 25 Days
Warsh Inherits an Economy That Cannot Be Fixed by Rate Policy Alone. That Is His First Problem.
Analysis Desk
Kevin Warsh arrives at the Federal Reserve with a mandate to restore price stability at 2% inflation. He inherits an economy running 3.8% CPI, a 30-year mortgage rate at 6.51% (up from 6.36% the prior week), a University of Michigan consumer sentiment index at an all-time low of 44.8, and an oil market that will remain structurally above $96 per barrel even in the diplomatic positive scenario because of a 6–8 million barrel-per-day inventory deficit that persists through Hormuz reopening. Wolfe Research has established that 40 basis points of the yield rise in the 10-year Treasury is structural — attributable to stronger corporate earnings and a reversal of AI-related fears, not the war. Only 10–15 basis points would reverse on an Iran deal. Rate cuts do not solve the inflation problem because the inflation is not primarily demand-driven. It is supply-driven through oil. Cutting rates in a supply-driven inflation environment makes the inflation worse. Warsh knows this. The question is whether he can communicate it clearly enough to satisfy markets, insulate the institution from political pressure, and begin the structural reforms he has signalled.
Three June 16–17 FOMC Paths
Path 1 — Hold + Neutral (Most Likely, JPMorgan Consensus)
Rates unchanged. Statement language balanced. No signal toward cuts or hikes. Market reads as Warsh establishing credibility before acting. Equity markets likely to hold or advance. 10-yr yield stable. This is what JPMorgan’s forecast of rates unchanged through mid-2027 implies as the base case.
Path 2 — Hold + Hawkish Lean
Rates unchanged. Statement signals next move is more likely a hike than a cut. This is the path Wolfe’s structural yield analysis implies: 40bp elevated, only 10–15bp reverses on a deal, the rest is earnings-driven and persistent. Markets: long-end rates spike, growth stocks correct, energy and financials outperform. Consistent with the ~80% rate-hike-this-year pricing from Reuters.
Path 3 — Hold + Dovish Signal (Trump’s Request)
Rates unchanged. Statement language implies cuts possible. This is what Trump has demanded and what Warsh’s White House swearing-in visually suggested. Markets: immediate credibility concern, long-end bond market sells off as inflation premium rises, dollar weakens, gold rallies. This path damages institutional independence on Day 1 and is the least likely given Warsh’s stated reform-oriented posture.
“I will lead a reform-oriented Federal Reserve, learning from past successes and mistakes, both escaping static frameworks and models and upholding clear standards of integrity and performance.” — Kevin Warsh, upon being sworn in
🚀 AI IPO Cycle — SpaceX · OpenAI · Anthropic · Is This 1999?
Forward-Looking · June 12 · May 27 AI Earnings
The Cycle Is Broadening From Hardware to Connectivity to Models. Here Is What Separates It from 1999.
Analysis Desk
SpaceX’s S-1 confirmed what the AI infrastructure supercycle has been building toward: a company with $18.674 billion in 2025 revenue, $4.694 billion in Q1 2026 revenue, and a $3.257 billion Starlink connectivity segment that is the primary revenue driver — not a future promise. OpenAI and Anthropic have announced IPO intentions for later in 2026. All three companies are targeting public markets in the same year that Nvidia confirmed $81.6 billion in quarterly revenue. The dot-com comparison is the question that will define how analysts frame this cycle. The 1999 parallel is real but imprecise.
In 1999, the primary internet IPOs were pre-revenue. Pets.com raised $82.5 million in February 2000 at a valuation of approximately $290 million with minimal revenue and no path to profitability. The distinction with 2026: SpaceX has $18.7 billion in annual revenue and a $1.127 billion adjusted EBITDA in Q1 alone. Nvidia has $81.6 billion in quarterly revenue growing 85% year-over-year. The infrastructure revenue is confirmed and accelerating. The frothy element is the multiple: SpaceX at $1.75 trillion represents approximately 94 times annualised Q1 2026 revenue. That is a valuation multiple consistent with a belief that Starlink, SpaceX’s space systems, and future AI connectivity businesses will grow revenue at 50%+ annually for a decade. The 1999 parallel applies to the valuation methodology, not the underlying revenue base. Companies with $18 billion in annual revenue can still be priced wrong — but the risk of zero is fundamentally different from a pre-revenue speculative IPO.
The week of May 27 is the first AI cycle test beyond the hardware layer. Marvell Technology (MRVL) is the semiconductor networking read — can AI data center interconnect demand sustain the NVDA-driven buildout in the networking silicon layer? Salesforce (CRM) is the enterprise AI software read — is the Monday.com +26% from War Day 80 and Workday’s beat-and-raise a pattern, or are those isolated beats? Synopsys (SNPS) is the chip design automation read. Snowflake (SNOW) is the AI data infrastructure read. If all four beat and raise, the AI cycle is confirmed across every layer from silicon to software. If any miss meaningfully, the cycle is narrowing to the hyperscaler core.
SpaceX (SPCX)$4.69B Q1 revenue · $1.75T target · June 12
Valuation multiple~94x annualised Q1 revenue · Pricing decade of growth
OpenAI / AnthropicIPO intentions 2026 · Pre-profit · Different risk profile
May 27 cycle testMRVL + CRM + SNPS + SNOW · All layers simultaneously
📊 Consumer vs. Equity — UMich 44.8 vs. Dow Record · The Two Economies
The Dow Hit a Record the Day Consumer Confidence Hit an All-Time Low. Both Readings Are Right.
Analysis Desk
The University of Michigan’s final May reading of 44.8 — the lowest in the survey’s nearly 60-year history — and the Dow’s all-time record close of 50,579.70 occurred on the same Friday. Understanding why both are simultaneously correct requires understanding who each measure is tracking. The S&P 500 is a forward-looking pricing mechanism for the earnings of the largest publicly traded US companies. It has been pricing: Nvidia’s confirmed $81.6 billion quarter, Workday’s margin expansion, the AI infrastructure supercycle, Warsh’s monetary certainty, and a diplomatic positive that oil will eventually fall. The University of Michigan consumer sentiment survey tracks how American households feel about their current and expected financial situation. It has been measuring: gasoline at four-year highs, mortgage payments up $475 per month since the war began, groceries running at 3.8% inflation, and three consecutive retailers (Walmart, Deere, Ross Stores) cutting or missing their full-year guidance while explicitly naming the Iran conflict as a cost driver.
These two measures describe the same economy from two different vantage points. The equity market vantage point is elevated and forward-looking: it prices the AI cycle’s future earnings and a deal-scenario oil crash that hasn’t happened yet. The consumer vantage point is ground-level and present-tense: it prices what gas costs today, what the mortgage payment is this month, and what groceries cost this week. The guidance recession — confirmed across multiple sectors this week — is the early signal that the ground-level view is starting to reach corporate annual plans. If WMT, Deere, and Ross are cutting guidance because of war fuel costs, the equity market’s forward earnings estimates are vulnerable to downward revision if the Iran conflict persists through Q3.
What closes the gap: A confirmed Iran deal → Hormuz reopens → oil falls toward $80 → gasoline falls below $3.50 within 60–90 days → CPI drops from 3.8% toward 2.5% within two quarterly readings → rate-hike probability collapses → mortgage rates fall from 6.51% toward 5.5% → consumer sentiment recovers → retailers restore FY guidance → equity forward earnings estimates rise. This is a six-to-nine-month transmission sequence. It requires the deal to close this month or in June. “Final stages” from Wednesday and the bilateral signal from Friday suggest it might. The HEU directive from Thursday’s morning suggests it might not.
UMich May final44.8 · Record low since 1966 baseline
Dow close50,579.70 · All-time record close
Guidance recessionWMT + Deere + Ross · Beat Q1 · Cut or pulled FY
Gap-closing catalystConfirmed Iran deal → oil crash → CPI relief → 6–9 months
⚔️ Samsung Memory Strike — Day 2 of 18 · Supply + Demand Collision
Day 2 of 18. The Demand Is Accelerating. The Supply Is Stopping. The Pricing Signal Arrives Week of May 27.
Samsung’s 18-day strike entered its second day Friday with no settlement news. The structural collision is straightforward: Nvidia’s $91.0 billion Q2 guide confirms that AI compute demand is accelerating, which drives HBM (High Bandwidth Memory) and NAND flash demand at rates that factories, per Seagate’s CEO, “take too long” to expand. A simultaneous 18-day production halt at the world’s largest memory manufacturer tightens the supply side while demand is increasing. Spot memory pricing typically takes five to seven days to reflect a supply disruption of this magnitude. That means the pricing signal — if Samsung does not settle over the long weekend — should appear in spot NAND and DRAM rates during the week of May 27. Companies holding existing memory inventory — Seagate (STX), Western Digital (WDC), and Micron (MU) — are the direct beneficiaries of any supply tightening. If Samsung does settle over the Memorial Day weekend, the pricing signal may not materialise, and the near-term supply constraint thesis reverses.
Strike statusDay 2 of 18 · No settlement entering weekend
Demand contextNVDA $91B Q2 guide · AI memory demand accelerating
Pricing signal windowDays 5–7 = Week of May 27 · Watch spot NAND / DRAM
Settlement riskLong weekend · Samsung has settled under hyperscaler pressure before
BeneficiariesSTX · WDC · MU — inventory holders
₿ Bitcoin May Close Watch — $75,830 Daily · May 30 Monthly Close · 8 Trading Days
The Daily Close Is Below $76K. The Trigger Is the Monthly Close. Eight Trading Days to Navigate.
Bitcoin closed Friday at approximately $75,830 — the lowest daily close of the week and the first below $76,000 since the week began with the war at $77,040 (War Day 80). Tom Lee’s $76,000 trigger is explicitly a May monthly close threshold: the last trading day of May 2026 is Friday May 30, and that is the close that determines whether the bull market signal is confirmed. A daily close below $76,000 on May 22 does not trigger it. It does, however, establish the recovery requirement: Bitcoin must close above $76,000 on Friday May 30 to confirm the signal. The current deficit is approximately $170 from Friday’s close. That is within one session’s range in normal trading conditions — but Bitcoin is not in normal conditions entering a geopolitically elevated long weekend.
The binary for the next 8 trading days is defined by Iran. A confirmed deal sends BTC above $80,000 by the momentum mechanics of the diplomatic gauge that has tracked every Iran development since the war began. A military escalation over the Memorial Day weekend sends BTC further below $76,000 and makes the May 30 recovery increasingly difficult. A sustained stalemate — the most likely outcome given the HEU directive in effect — leaves BTC oscillating in the $74,000–$78,000 range and makes the May 30 close a genuine nail-biter. The CLARITY Act (Creating Legal Accountability for Rulemaking In Technology) structural positive remains intact: Senate floor scheduling pending, July 4 White House target, 69% Polymarket probability. The structural positive has not moved. The cyclical negative — Warsh’s no-cuts posture — has strengthened.
Friday daily close~$75,830 · Below $76,000 · First time this week
Monthly close trigger$76,000 · May 30, 2026 (last trading day)
Recovery needed~$170 above Friday’s close · One session’s range
Deal scenarioBTC above $80,000 · Trigger easily confirmed
Escalation scenarioBTC further below $76,000 · Trigger missed
Stalemate scenario$74K–$78K range · May 30 close is the decision point
📅 Week Ahead — Five Sessions That Test Everything the Week Just Confirmed
Memorial Day Opens Into Iran. May 27 Tests the AI Cycle. May 30 Closes the Month.
Monday May 25 — Memorial Day (Closed)
Three days of Iran news, Samsung strike developments, and any weekend diplomatic or military events will be unpriced until Tuesday. Geopolitically the highest-risk long weekend of the conflict: HEU directive in effect, military on standby, bilateral signal from Friday creating an ambiguous but potentially productive negotiation environment.
CLOSED
Tuesday May 26 — Consumer Confidence + AutoZone + Zscaler
The Conference Board’s May consumer confidence index — read alongside Friday’s UMich 44.8 record low for a complete consumer picture. AutoZone (AZO) earnings as an auto parts retail bellwether for war-era discretionary spending. Zscaler (ZS) earnings as the cybersecurity software read in the AI infrastructure context. Tuesday’s open will gap for whatever happened to Iran over the weekend.
TUE MAY 26
Wednesday May 27 — AI Earnings Round 2 + New Home Sales
The week’s dominant catalyst: Marvell Technology (MRVL), Salesforce (CRM), Synopsys (SNPS), and Snowflake (SNOW) all reporting. If all four beat and raise, the AI cycle is confirmed across every layer from semiconductor networking to enterprise software to AI data infrastructure. Also: April new home sales — the first confirmed transaction data since 30-year mortgages crossed 6.51%.
KEY DATE
Friday May 30 — Last Trading Day of May (BTC Monthly Close)
The May monthly close for all assets. For Bitcoin specifically: $76,000 or above on this close confirms Tom Lee’s bull market signal. Below $76,000 on this close does not. The close also determines May’s performance for all major indices entering the June seasonality period.
MAY CLOSE
June 12 — SpaceX IPO (SPCX on Nasdaq)
The largest IPO in history if achieved at the $1.75 trillion target. Road show reportedly June 5. Starlink’s $3.26 billion Q1 revenue is the primary asset being sold. The IPO will set a valuation benchmark for the AI connectivity infrastructure layer that every subsequent AI IPO — OpenAI, Anthropic — will reference.
JUNE 12
June 16–17 — Warsh’s First FOMC
The first Federal Open Market Committee meeting under Kevin Warsh’s chairmanship. Three paths as analysed above: hold + neutral (most likely), hold + hawkish lean (Wolfe structural scenario), or hold + dovish signal (Trump’s demand, least likely). The policy decision and Warsh’s inaugural press conference will set the market’s read on his independence for the rest of 2026 and potentially the rest of his term.
JUN 16–17
📖 Key Terms — Issue 68
This Edition
Federal Open Market Committee (FOMC)
The policy-making body of the Federal Reserve System responsible for setting the target range for the federal funds rate — the interest rate at which banks lend overnight reserves to each other, which serves as the benchmark for borrowing costs throughout the US economy. The FOMC consists of twelve voting members: the seven members of the Board of Governors (all permanent voters), the president of the Federal Reserve Bank of New York (permanent voter), and four of the remaining eleven regional Federal Reserve Bank presidents on a rotating annual basis. The FOMC meets eight times per year in scheduled sessions, each lasting two days. At the conclusion of each meeting, the committee votes on a policy statement that sets the target rate range and communicates the committee’s assessment of economic conditions. The chair of the Federal Reserve chairs the FOMC meeting and delivers the post-meeting press conference. Kevin Warsh’s first FOMC meeting as chair is scheduled for June 16–17, 2026 — 25 days from Friday’s swearing-in. The significance: this meeting will be Warsh’s first public communication of his monetary policy framework under his own chairmanship. The statement language he chooses — particularly any signals about whether the next move is more likely a hike or a cut — will set the market’s interpretation of his tenure for the remainder of 2026. With rates ~80% likely to be hiked this year per Reuters, and JPMorgan forecasting no change until mid-2027, the committee faces a genuine internal disagreement about the trajectory. Warsh’s management of that disagreement in his first press conference will be the market event of June.
AI Infrastructure Cycle
The current phase of technology investment characterised by the construction of the physical, software, and connectivity infrastructure required to train, run, and deploy large artificial intelligence models at commercial scale. The AI infrastructure cycle in 2025–2026 encompasses four layers, each confirmed by specific earnings this week. The semiconductor layer: Nvidia’s $81.6 billion Q1 2026 revenue, up 85% year-over-year, confirming that GPU demand for AI training and inference has not plateaued. The connectivity layer: SpaceX Starlink’s $3.257 billion Q1 2026 revenue, representing the satellite internet infrastructure that delivers AI capabilities to the edge. The enterprise software layer: Workday’s beat-and-raise on margins, Salesforce’s Q1 performance, confirming that enterprise spending on AI-enabled software continues. The data infrastructure layer: Snowflake and Marvell Technology reporting May 27, which will either confirm or challenge the cycle’s extension into data management and networking silicon. The distinction from the 1999 dot-com cycle is primarily one of revenue: the leading AI infrastructure companies have confirmed, growing, high-margin revenue that the pre-revenue dot-com era lacked. The risk is one of multiple: SpaceX at $1.75 trillion values the company at approximately 94 times annualised Q1 2026 revenue, which prices a decade of 50%+ annual revenue growth into the IPO. Even companies with real revenue can be priced wrong. The AI infrastructure cycle has real foundations. The question entering May 27’s earnings is whether those foundations extend beyond the semiconductor layer into every layer above it.