Research and idea generation for personal use. Not investment advice. See full disclaimer at the bottom.

Top of mind

Broadcom (AVGO) delivered a clean Q2 beat — $22.2B revenue (+48% YoY) and AI semiconductor revenue of $10.8B (+143% YoY) against a $10.7B guide — but CEO Hock Tan held the FY2026 AI semiconductor forecast at $56B rather than raising it, and infrastructure software missed consensus by $140M, sending AVGO down nearly 14% in after-hours trading. CrowdStrike (CRWD) compounded the damage: record Q1 net new ARR of $256M (+32% YoY) and a 4-for-1 stock split announcement weren't enough to offset a billings miss, and the stock fell ~9% in after-hours. Both reactions land on a tape already stressed by Iran — drone strikes on Kuwait's international airport on June 3 and a House war powers rebuke of Trump pushed Brent crude to $99.50, and the path to $100 now looks shorter than the path back to $90.

Market snapshot

(Levels reflect June 3, 2026 confirmed closes — the most recent prior trading-day close before AVGO and CRWD reported after the bell. June 4 session opened under pressure from both; the AVGO intraday range touched a low of approximately $405 before stabilizing.)

Asset	Level	Change	Notes
S&P 500	7,553	-0.74%	Iran and pre-earnings caution; broke a 3-day record run
Nasdaq Composite	26,852	-0.89%	Chip names gave back the Marvell/AI surge from prior week
Dow Jones	50,776	-1.04%	Three declines in four sessions; Iran the macro override
10Y Treasury	4.49%	-1bp	Slight flight-to-safety bid; yield curve flattening
VIX	~16.1	+~3%	Rising into the dual earnings catalyst; pre-AH levels
Brent Crude	~$99.50	+2.5%	Iran drone escalation; Hormuz risk premium expanding again

Read-through: The June 3 tape was already pricing risk-off before AVGO and CRWD opened their books. The S&P pulled back from all-time highs not because the AI story broke, but because Iran is the macro governor that AI beta cannot override. After-hours confirmed what the session had telegraphed: the market was positioned for a beat-and-raise from Broadcom, not a beat-and-hold, and the gap down on June 4 represents an expectations recalibration rather than a fundamental break. The question the next two sessions answer is whether institutions treat the gap as a buying opportunity or the start of a broader AI-multiple compression.

Headlines & analysis

1. Broadcom beats AI metrics — market sells the held guide

Source: CNBC, Motley Fool Earnings Transcript, MarketBeat (June 3, 2026) So what: AVGO Q2 FY2026 delivered $22.2B revenue (+48% YoY, above the $22B guide), AI semiconductor revenue of $10.8B (+143% YoY, above the $10.7B guide), and EPS of $2.44 versus $2.32 expected. Q3 FY2026 guidance of $29.4B (+84% YoY) was strong in absolute terms. But CEO Hock Tan reaffirmed — not raised — the FY2026 AI semiconductor revenue forecast at $56B, and infrastructure software came in at $7.18B against the $7.32B expected. AVGO also confirmed long-term supply agreements with Google, Anthropic, OpenAI, and Meta for multi-gigawatt AI compute deployments, and booked $6B in AI orders from two additional customers. The new hyperscaler partnerships are structurally important; the market chose to focus on the $140M software miss and the unraised guide. AVGO fell nearly 14% after-hours.

2. CrowdStrike: record ARR, raised guidance, 4-for-1 split — and still sold off

Source: TechTimes, Motley Fool Earnings Transcript, SEC Form 8-K (June 3, 2026) So what: CRWD Q1 FY2027 produced record net new ARR of $256M (+32% YoY) — above the $250M consensus — and EPS of $1.10 beat by $0.22. Full-year FY2027 net new ARR growth guidance was raised by 520 basis points at the midpoint. The company announced a 4-for-1 stock split, which management teams typically reserve for periods of high confidence. Despite all of this, Q1 billings disappointed relative to expectations and CRWD fell ~9% after-hours. Billings are timing-dependent and structurally less informative than ARR for a subscription security business; the market disagreed and punished the miss. The setup: a stock with record fundamentals and raised guidance trading lower on a metric that can swing quarter to quarter. That is either a dip or a warning, depending on whether Q2 billings normalize.

3. Iran escalation enters a new phase — oil approaches the $100 line

Source: Bloomberg markets wrap, Britannica (2026 Iran war event), CNBC (May 28 – June 4, 2026) So what: The US launched targeted airstrikes on southern Iran on May 26; Iran responded with drone strikes that hit Kuwait's international airport on June 3. The House of Representatives passed a war powers resolution on June 3 rebuking the Trump administration and calling for an end to military action — a political signal with limited enforcement power but significant symbolic weight. Brent crude reached approximately $99.50, with commercial traffic through the Strait of Hormuz severely disrupted. At $100 Brent, US CPI stays structurally above 3.5% YoY through at least Q3 2026, and the Fed loses its primary justification for keeping rates on hold. Every session oil holds near $100 makes Warsh's language on June 17 harder to frame as dovish.

4. Prediction markets reprice June 17 to a hawkish hold — not a hike

Source: Polymarket, Kalshi, TradingKey analysis (June 3–4, 2026) So what: Despite the Iran oil shock and resilient labor data, prediction markets assign approximately 97% probability to rates held at the June 16–17 FOMC — Warsh's first meeting as the 17th Fed Chair (he took office May 22). The base case is a "hawkish hold": Warsh removes the easing bias, signals rate hike optionality, but does not move in his inaugural meeting. Longer-run pricing shows 35% probability of a 2026 hike and 63% by July 2027. For equity investors, this framing is important: a hawkish hold that formally ends the easing cycle delivers most of the multiple compression of an actual hike without the accompanying recession signal that typically follows hike cycles. Duration-sensitive tech names face the same re-rating either way.

5. BNP Paribas initiates CoreWeave at Outperform — $192 target implies 71% upside

Source: Barchart, Seeking Alpha, FXLeaders (June 2–3, 2026) So what: BNP Paribas analyst Stefan Slowinski initiated coverage of CoreWeave (CRWV) with an Outperform rating and a $192 price target on June 2, calling the company "one of the most strategically important companies within the AI infrastructure ecosystem." CoreWeave reported Q1 2026 revenue of $2.1B and a revenue backlog of $99.4B — roughly 12x its current annual run rate — and is up 74% year-to-date. The initiation timing matters: it arrived one day before AVGO's guide non-raise. If AVGO's held forecast signals that hyperscaler AI capex is on track but not accelerating unexpectedly, CRWV's backlog provides the forward demand visibility that chip suppliers don't have. Note: Magnetar entities sold $14.6M in CRWV shares around the same period — a yellow flag worth monitoring.

Ideas — long-term core

Quality businesses, durable competitive advantages, reasonable valuation. Hold horizon: years.

AVGO — Broadcom

Thesis: Tonight's selloff is an expectations reset, not a thesis break. AVGO confirmed AI revenue of $10.8B (+143% YoY) in Q2, announced supply agreements with four of the largest AI spenders globally (Google, Anthropic, OpenAI, Meta) with multi-gigawatt deployments starting FY2027, and booked $6B in orders from two new unnamed hyperscaler customers. The FY2027 AI semiconductor revenue target of $100B+ is intact, implying roughly 3x growth from the current annual run rate over two years. VMware infrastructure software — $7B/quarter in recurring revenue — is a durable compounding floor the market perpetually discounts while fixating on the AI story.
Valuation note: AVGO at $405–425 (post-AH range) trades at approximately 27–28x forward estimates — below the 30x premium heading into earnings and at a meaningful discount to Nvidia's 35x. For a business with contractual multi-year AI supply relationships plus $7B/quarter in recurring software, 27–28x is the most attractive entry point since February 2026.
Why now (or why patient): Patient accumulation over 3–5 sessions after the gap makes more sense than chasing the open. The June 4 open establishes price discovery; institutional behavior in the first two hours sets whether this is a dip or a re-rating.
Risks / bear case: The non-raised guide could signal that hyperscaler AI capex plateaus sooner than modeled — if Alphabet's $80B infrastructure commitment does not monetize, custom ASIC orders pause before FY2027 deployments begin. Software miss raises execution risk on VMware integration. Oil at $100 plus a Warsh hawkish pivot compresses 27–28x multiples simultaneously, turning the dip into a deeper re-rating. Google customer concentration remains the single-largest fundamental risk.

CRWV — CoreWeave

Thesis: CoreWeave is the largest independent GPU cloud platform with a $99.4B revenue backlog — 12x current annual revenue — providing forward demand visibility that most software companies cannot match. As hyperscalers build proprietary AI infrastructure, the overflow and burst capacity demand routes to specialists like CRWV that provision at speed. BNP Paribas' initiation adds analyst coverage depth, and the AI infrastructure buildout thesis (Broadcom's held guide, Marvell's Jensen endorsement) is about demand at the infrastructure layer, where CRWV lives.
Valuation note: Traditional P/E doesn't apply to a high-growth capital-intensive platform. Backlog-to-market-cap is the relevant anchor; at $99.4B in backlog versus current market cap, CRWV is priced for conversion, not discovery. Modeling depends heavily on backlog timing and margin trajectory as GPU supply commoditizes.
Why now (or why patient): The AVGO AH selloff creates sympathy pressure on AI infrastructure names broadly — CRWV may offer a better entry than the post-BNP initiation spike. Watch the June 4–5 tape for any washout that disconnects CRWV from the AVGO-specific narrative.
Risks / bear case: Magnetar insider selling ($14.6M) is a yellow flag. Customer concentration risk — if one or two hyperscalers reduce CRWV usage, backlog converts slower than modeled. Heavy dependence on NVIDIA A100/H100 supply means any Nvidia supply disruption flows directly through margins. GPU commodity pricing pressure in 2027 could compress the margin story that underpins current valuation.

Ideas — opportunistic

Catalyst-driven, time-bound, sized smaller. Hold horizon: days to months. Define exit before entry.

CRWD — Post-billings-miss recovery

Catalyst: CRWD fell ~9% after-hours on a Q1 billings miss despite record net new ARR ($256M), a $0.22 EPS beat, a 520bp raise to FY2027 ARR guidance, and a 4-for-1 stock split. Billings are timing-dependent; the structural ARR trajectory is intact and the raised guidance is the forward signal. The dislocation between record fundamentals and a negative stock reaction creates a short-term entry if Q2 billings normalize.
Time horizon: 2–4 weeks, through the next analyst commentary cycle or any intra-quarter ARR update.
What would invalidate: A second consecutive billings miss in Q2 FY2027 would indicate the ARR → billings conversion is structurally slowing, not timing-dependent. Any guidance reduction or commentary suggesting churn is re-emerging from the July 2024 Falcon incident recovery period invalidates the thesis entirely.
Risk note: This is a sympathy-sell in a market digesting AVGO + Iran simultaneously. The cleaner entry is the June 4–5 price discovery after the market absorbs the full picture, not pre-market futures. Size for a 2–4 week hold at 1–2% of portfolio.

XOM / Energy — inflation hedge against sustained $100 oil

Catalyst: Brent at $99.50 with active Hormuz disruption and no credible ceasefire signal. If oil holds above $100, it keeps US CPI structurally elevated into Q3 2026, extends the Warsh hawkish pivot timeline, and re-rates energy multiples upward while simultaneously compressing AI-adjacent multiples. Energy (XOM, broad sector) has significantly underperformed the AI trade YTD — a $100 oil regime reverses that relative performance.
Time horizon: Tail-risk hedge through June 17 FOMC. Not a growth trade; a portfolio hedge against the Iran scenario playing out fully.
What would invalidate: Any credible ceasefire, Hormuz reopening, or House war powers resolution with enforcement sends Brent back toward $80–85 and collapses the energy hedge thesis immediately.
Risk note: This is not a conviction trade — it is a positioned hedge against a tail risk currently underpriced relative to the geopolitical newsflow. Size at 2–3% of portfolio maximum.

Portfolio-level guidance

Allocation and risk observations. Not specific buy/sell calls — those depend on a full picture this report doesn't see.

The AVGO and CRWD after-hours moves are a portfolio event, not just earnings events. The two AI-adjacent leaders reporting the same night, with the same "beat the metric, miss the expectation" pattern, suggests the market was positioned for perfection across the AI trade. That positioning is now unwinding. Reducing overall AI semiconductor concentration before the June 4 open makes more sense than waiting for the first hour of price discovery to decide.
A hawkish hold on June 17 may be more dangerous for equities than a hike. A hike creates a clear signal: Warsh is data-reactive, implying cuts are available when data softens. A hawkish hold that formally removes the easing bias while holding rates sends a structurally bearish message: rates are going higher on a timeline that can't be priced through CME FedWatch. Duration-sensitive tech multiples compress the same amount in either scenario; the difference is that a hold removes the "it'll be over soon" safety valve. Reduce duration exposure before June 17.
NFP tomorrow (June 5, 8:30 AM ET) is the week's most consequential remaining catalyst. Capital Economics projects May NFP at approximately 65K new jobs — down sharply from April's 115K. A soft print gives Warsh data-driven cover to hold tone neutral on June 17 and is mildly risk-on for equities. A hot print (+100K) validates the inflation-via-labor story, raises the 2026 hike probability above 40–45%, and creates a simultaneous equity-and-bond headwind. The next 18 hours carry more information density than the prior four sessions combined.
Goldman's INTU downgrade to Sell is the canary in the software sector. If AI coding and productivity tools are eroding Intuit's tax and accounting revenue base, "AI eats software" is entering Phase 2: platform incumbents facing margin pressure from AI-native entrants, not just foundational model providers winning. Watch for follow-on software-sector downgrades in the June 4–10 window.
Cash positioning: Hold 10–15% through June 17. The AVGO/CRWD after-hours moves compress near-term risk appetite across AI names. Entry on quality names after price discovery (3–5 sessions) is structurally better than chasing the gap.

Watch list — tomorrow / this week

Economic data: May NFP at 8:30 AM ET on Friday, June 5 — the final major labor print before the FOMC blackout period begins. Capital Economics projects 65K (vs. April's 115K). The ADP read published June 3 morning is the leading indicator; any divergence between ADP and NFP creates conflicting signals for the June 17 language. Earnings: No major S&P 500 earnings remaining this week after AVGO and CRWD reported. Veeva Systems (VEEV) if results are pending; Five Below (FIVE) as a consumer health read-through on the lower-income spending base amid oil-driven inflation. Fed / central bank: June 16–17 FOMC — Warsh's inaugural meeting. Watch for any pre-blackout governor statements this week; with Warsh's known preference for limited public communication, any Warsh-attributed language before the blackout begins carries amplified signal weight. Iran / Hormuz: House war powers vote passed June 3 but Trump has not signaled military pullback. Iran drone strikes on Kuwait's airport on June 3 signal escalation, not de-escalation. Any credible ceasefire or Hormuz reopening signal is the single most positive macro catalyst available — it would send Brent below $85, collapse June 17 hawkish probability, and trigger a simultaneous equity and bond rally. AVGO price discovery: Watch institutional activity in the June 4 first 90 minutes. If long-only funds accumulate into the gap near $405–425, the selloff is temporary noise. If price drifts lower on volume, the market is re-rating the AI capex premium structurally. Analyst price target revisions (June 4–5) will set the near-term resistance levels.

Disclaimer

This report is prepared for personal research and informational purposes only. It does not constitute investment advice, an offer, or a solicitation to buy or sell any security. Information is drawn from public sources believed to be reliable but is not guaranteed accurate or complete. Markets change rapidly; data may be stale by the time of reading. Any "ideas" mentioned are research candidates, not recommendations, and do not consider any specific person's financial situation, objectives, or risk tolerance. Consult a licensed financial advisor before making investment decisions. Past performance does not predict future results.

Broadcom beats the quarter — the AI trade gets sold anyway