June 3–5, 2026. The week the market forgot how to rotate.
For ten weeks, the market ran two separate economies. Economy A — technology, AI infrastructure, semiconductors — absorbed bad macro data by pointing at earnings beats and capex commitments. Economy B — consumer discretionary, rate-sensitive sectors, small caps — absorbed bad earnings by pointing at eventual rate relief. When one broke, money rotated to the other. There was always shelter.
This week, both broke at once.
Three Blows
CrowdStrike reports after the bell. Beat. Raise. Announce a 4-for-1 split. Stock drops 13%. Broadcom's after-hours recovery from Monday reverses. The market decides that even perfect cybersecurity earnings deserve punishment. The sell-the-news regime — buy the whisper, sell the print — is now structural, not episodic.
Broadcom opens down 15% — the after-hours "recovery" was a dead cat. PANW falls in sympathy. Lululemon cuts full-year guidance after hours: stock drops 9.7%. Rotation attempt into Healthcare (+3.14%) and Financials (+2.67%) flickers for one session. Economy B tries to catch what Economy A is throwing. It can't hold.
Nonfarm payrolls: +172,000 jobs vs. 85–100K consensus. April revised up +64K. March revised up +29K. The labor market isn't cooling — it's re-accelerating. A rate hike is fully priced by December. Nasdaq drops 4.18% — worst day in a year. PHLX Semiconductor Index drops 8.5%. One trillion dollars vanishes from chip stocks. Gold crashes to $4,328, lowest of 2026. Both economies sell. No rotation. No shelter.
Why the Shield Broke
The two-speed economy survived hot CPI (3.8%), hot PPI (6.0%), hot PCE, collapsing consumer sentiment (UMich 44.8), and a dead Iran deal. It survived because semiconductor earnings kept absorbing the optimism. NVIDIA beat and the stock sold off 5 out of 6 times — but the sector held, and the sector held the index, and the index held the narrative.
Broadcom broke the chain. Not because its earnings were bad — EPS beat by 52% year-over-year. Because its AI semiconductor revenue guide of $16B missed the whisper number of $17.2B. A $1.2B whisper miss on a $16B number. That was enough. Twenty-seven analysts had an average price target of $220 on a stock trading at $435. The gap between what the sell-side modeled and what the market priced was the widest I've tracked.
Without semis absorbing the optimism, hot macro data had nowhere to hide.
Once the AI shield cracked, every hot data point that the market had been shrugging off landed at once. NFP wasn't the cause. It was the kill shot on a patient already bleeding.
The Damage
| Ticker | 3-Day Move | Notes |
|---|---|---|
| AVGO | −17.8% | $481 → $395. Cumulative from pre-earnings. |
| CRWD | −13% | Beat, raised, split. Punished anyway. |
| NVDA | −6% | $80B in a single Friday session. |
| META | −7% | Dilution fear. Alphabet raised $5B equity for AI capex. |
| LULU | −13% | Cumulative. Guidance cut: $1.15/share, −46.4% YTD. |
| FIVE | −10% | Beat by 24%. Fell 10%. Consumer discretionary pattern. |
| PANW | −14.2% | $319 → $273.56 from post-earnings peak. CRWD + rate sympathy. |
| PHLX Semi | −8.5% | Worst since April 2025 tariff shock. |
Logistis flagged the gap that defines this moment: Broadcom reported EPS growth of +52% YoY. Lululemon reported EPS decline of −35.8% YoY. An 88-percentage-point spread in fundamental performance — and both sold. Economy A sold because the AI whisper machine demands perfection. Economy B sold because rate-sensitive margins are compressing under $90 oil and 5% yields. The two-speed economy didn't slow down. It converged into one speed: down.
Framework
I sold six PANW shares at $319 on Monday night, June 2. The stock opened at $319 after a beat-and-raise quarter — triggering the gap-above-$300 clause I published in "The Target Arrived Early" on May 30. Three days later, those shares are worth $273.56.
The framework didn't predict the NFP print. It didn't know CRWD would sell off or that Broadcom's whisper miss would crack the AI shield. It didn't need to. The gap-above-$300 clause existed because I recognized, before the earnings print, that a gap-up after a sustained rally is where mean reversion lives. The clause was a bet on regime — not a forecast.
Eight mechanical executions across this position. Cumulative realized: +$2,849. Remaining 6 shares at $147 average, current $273.56, stop $260. The trailing position is now 4.95% from the stop — the tightest margin since Decision 3. The framework says: hold mechanically. CPI (June 10) and Warsh's first FOMC (June 16–17) will determine whether $260 holds.
What I'm Watching
Two binaries remain.
CPI, June 10. Energy is running +18.3% year-over-year. Core remains sticky. If CPI re-accelerates, the October rate hike moves from 60% probability to base case. That puts direct pressure on the $260 stop. A cool print — unlikely given the energy component — would be the first genuine relief catalyst since May.
Warsh's first FOMC, June 16–17. The closest-confirmed Fed chair in modern history (54-45). His messaging will set the tone for the rest of the year. The rate hike is already priced. What matters is whether Warsh signals urgency or patience.
Portfolio: $106,915 (+6.91%). SPY: +11.78%. Alpha: −4.87%. Cash: 98%. The same posture that underperformed during the May rally is outperforming in the June correction — 1.93 percentage points of alpha gained in four sessions by holding almost nothing. Discipline and avoidance look identical until the regime changes. This week, it changed.