The call: Tradable bottom — LONG quality compute into the hole (LRCX $195–210, AMD book $400–425 held), S&P target 7,780–7,820, 1–3 months, SPECULATIVE-HIGH.
Why now: 165K jobs (vs 59K expected) + Lam's $25B buyback + VIX 28 spike (68% hit-rate for +4.2% in 3M) beat the fear story.
The disagreement: Consensus says Blue Wave kills the data-center buildout; capex says 2027 is already funded — regulation bites 2028, if at all.
The level that changes everything: Daily S&P close below 7,300 — halve all longs, no debate.
Start with what actually happened. Thursday wiped out ~$2.9 trillion: S&P 7,353.73 −4.36%, Nasdaq 24,315 −5.52% into correction and below its 200-day, Dow 51,523 −3.44% (−1,833), SOXX −12%+ for its worst day since March 2020. All 11 S&P sectors finished red; tech −7.3%.
The trigger was not earnings — AMD had just printed a record quarter — it was politics repriced in one session. Democrats swept Congress plus Texas, Georgia, Ohio, Michigan and Nevada governorships on explicit anti-data-center platforms. Goldman via Hatzius, Jefferies ("worst-case near-term scenario"), Meta President Dina Powell McCormick ("very challenging to start new build-outs"), and Texas Governor-elect Gina Hinojosa's moratorium call all landed within 24 hours. NPR summed it up: voters revolted against water-and-power-hungry hyperscale projects.
My opinion: the market was right to reprice permitting risk and wrong to reprice 2026–27 revenue. JPMorgan's read is the correct one — no material change to 2027 capex ($800B+ from Amazon/Alphabet/Meta/Microsoft/SpaceX alone), possible drag in 2028+. Jefferies cut end-of-decade capacity only from 122.4 GW to 112.8 GW — an 8% haircut that cost the SOX ~30% from peak. That is panic arithmetic.
The October payrolls report Friday at 8:30am ET was unambiguously hot: +165,000 vs 59,000 Dow Jones consensus, unemployment 4.1% vs 4.2% expected, August revised +53K to 186K, September +31K to 60K, 3-month average 137K vs 45K 12-month average. Professional/business services +70K, healthcare +55K, manufacturing +19K, transportation +25K. Wages +0.4% m/m, +3.1% y/y.
The consequence is not dovish. CME FedWatch December hike odds jumped to ~73–80% (from ~60% Thursday), 10-year yield to ~4.88%, and the Dow actually lagged on the print. But for equities this is the good kind of hot: broad-based hiring rebuts the "AI capex is propping up a dying economy" thesis from Rosenberg and others. Schwab's Collin Martin put it cleanly: the economy is still absorbing labor comfortably, not buckling under tighter financial conditions.
Caveats I will not hide: U-6 underemployment rose to 7.9%, average duration jumped to 27.1 weeks (highest since Sept 2021), 1.96M long-term unemployed, and wage growth 3.1% still trails 3.7% CPI. Low-hire, low-fire persists. That is why the bounce favors quality balance sheets, not speculative neocloud.
Lam Research (LRCX) authorized a $25 billion buyback on Nov 5 — in the teeth of the rout — and rose 8.3% to $200.64 Friday intraday. Read the press release wording: CFO Doug Bettinger cites "strong positioning at key technology inflections and robust cash generation." WFE suppliers do not authorize $25B (roughly 15–20% of market cap scale) if 2027 tool orders are evaporating.
It is not alone. White House NEC Director Hassett said Friday the "Blue Wave won't stop AI build-out." Cathie Wood's ARK sold $119.7M Broadcom and bought $45.9M Nvidia on the Thursday close — a rotation toward the prime compute beneficiary, not an exit from AI. Pre-market Friday: SNDK +10%, APLD +10%, CRWV +10%, NBIS +9.5%, LITE +9.5%, MRVL +9.5%, AMD +8–9%, NVDA/AVGO/INTC +4–5%.
My take: follow the cash, not the commentary. Buybacks + hyperscaler reiterations + a hot labor market are two independent bullish signals. That meets my HIGH threshold only when combined with the sentiment washout below — hence SPECULATIVE-HIGH for a new LRCX trade, HIGH for the combined book.
Do not mistake an intraday bounce for an all-clear. Overnight, KOSPI fell 6–8.7%, triggering a 20-minute circuit breaker, with Samsung and SK Hynix −10% and TSMC hitting limits. The PHLX SOX is still down ~30% from peak. CNN Fear & Greed 24 (Extreme Fear), crypto Fear & Greed 28 (Fear), VIX 28.18 Thursday close, SMH traded below NAV with $1.5B of $1.7B options premium in puts.
Bank of America's Hartnett says wait for three stars to align before rushing back in, and Schwab notes high-yield spreads jumped 15bp Thursday — only matched twice in a year — plus the most net retail selling since Covid 2020. History helps but does not guarantee: Bespoke finds 19 VIX +5-from-under-20 spikes since 1990 delivered median +4.2% in 3 months (68% hit rate) and +4.8% in 6 months (79%), but average 1-month return was −0.4% thanks to 1990 and Feb 2020 tails. Translation: sharp VIX spikes mark tradeable bottoms 2 times in 3, and disasters 1 time in 5.
Second-order effects most coverage misses: (1) a permitting squeeze helps incumbents with powered land — shovels already in ground trade at a premium; (2) 10-year at 4.88% toward 5.10% kills long-duration neocloud first, hyperscalers last; (3) a Democratic Congress still faces a Trump veto — Edwards Asset Mgmt is right that a nationwide moratorium is unlikely, so the stranded-asset fear is overstated.
The flagship is selective dip-buying, not index-chasing. Yesterday's piece (AMD record print, panic sale) triggered DAILY-AMD-1105 BUY $400–425 at $404.40 close — now $442.35 (+9.4% intraday). Hold it. The incremental add is equipment over memory.
Why LRCX over MU here: MU's 17% plunge left a 3.72x forward P/E that screens irresistible, but DRAM is the most politically elastic link (power-hungry fabs, Taiwan halt headline, $7B DRAM ETF losses). Compute tools get paid whether the winner is Nvidia, AMD or a hyperscaler ASIC. LRCX: 33.4% revenue growth, 51% gross margin, 71.7% ROE, 27.4x TTM — a cyclical buying its own dip at scale.
Downstream consequences to position for: (a) utilities with contracted hyperscaler load (VST, CEG report Nov 10/9) stay bid even if chips chop; (b) gold/XLP/staples hedges from the book still earn their keep into CPI Nov 10; (c) Bitcoin $75,749 (−0.3%) and ETH $2,228 remain risk-off hostages — no new crypto risk until S&P reclaims 7,500.
| Trade board (flagship + live book) | Entry | Target | Invalidation | Horizon / Conviction |
|---|---|---|---|---|
| NEW 96. DAILY-LRCX-1106 — BUY WFE dip | $195–210 (last $200.64) | $250–265 | Daily close <$175 | 1–3 mo / SPECULATIVE-HIGH |
| 95. DAILY-AMD-1105 — HOLD (triggered $404.40, now $442) | $400–425 held | $580–620, stretch $750–800 | Daily close <$360 | 1–3 mo / HIGH |
| Book rule — risk governor | If daily S&P <7,300 or 10Y >5.10%, halve all longs | 1–3 d / HIGH | ||
Bull 30%: Payrolls + CPI cools to allow one-and-done hike, Hassett veto backstop + hyperscaler capex reaffirmed, SOXX reclaims 150 in two weeks. Trigger: CPI Nov 10 benign + CRWV/NBIS prints guide up. Action: press LRCX to full size, add NVDA $185–195.
Base 50%: Choppy 7,300–7,600 range into NVDA Nov 17. Chips +5–8% Friday, then two steps forward one back as permit headlines alternate with buybacks. Trigger: S&P holds 7,300 on daily closes, 10Y stays 4.70–5.00%. Action: hold AMD/LRCX halves, clip XLP/GLD.
Bear 20%: Hot CPI + second hike priced for 2027, KOSPI contagion + credit spreads widen, S&P breaks 7,300 and META breaks $600. Trigger: CPI >0.4% m/m or 10Y daily >5.10%. Action: halve all longs per book rule, LRCX stopped at $175.
Acknowledgement: call 89 was stopped out Nov 5 (−5% vs $95 trigger) and replaced by 95 — loss owned. Call 95 is working (+9.4% day one). No new memory longs until DRAM contract prices stabilize.
Contents: A. Data snapshot · B. Models & load-bearing assumption · C. What I read · D. Linked sources
| Metric | Level | Note |
|---|---|---|
| S&P / Nasdaq / Dow (Nov 5 close) | 7,353.73 −4.36% / 24,315 −5.52% / 51,523 −3.44% | Nasdaq correction, below 200-day |
| SPY / QQQ Friday intraday | $739.38 +0.87% / $658.72 +1.80% | Implied SPX ~7,417 |
| SOXX / NVDA / AMD / MU / LRCX | $136.22 +3.27% / $189.51 +4.36% / $442.35 +9.38% / $689.04 +4.97% / $200.64 +8.34% | AMD off −17.3% Thu; MU −17.5%; NVDA −9.3% |
| Jobs Oct | +165K vs 59K exp; U-Rate 4.1%; wages +0.4% m/m +3.1% y/y | 3-mo avg 137K; revisions +84K |
| Rates / Fear | 10Y ~4.88%; Dec hike 73–80%; VIX 28.18; CNN 24 Extreme Fear; Crypto 28 Fear | Bespoke: +5 VIX spikes +4.2% med 3M |
| Crypto | BTC $75,749 −0.28%; ETH $2,228; total cap $2.57T −5.27% 24h | Risk-off hostage |
| Fundamentals | NVDA rev +83.4% GM 74.7%; AMD rev +41.9% GM 53.7%; LRCX rev +33.4% GM 51.1% PE 27.4x | Via financial handler |
Model: 2027 hyperscaler capex is funded (cash + debt already raised) so permitting delays shift timing, not quantum, until 2028. Load-bearing input: Jefferies 112.8 GW vs 122.4 GW (8% cut). If actual 2027 capex guides cut >15% (CRWV/NVDA), fair value for SOXX falls ~20% and both AMD $360 and LRCX $175 stops hit — answer flips from buy-dip to sell-rallies.
Primary: BLS release via CNBC/CNN jobs pages (sector table, revisions, wage/workweek lines); Lam Research Nov 5 press release (CFO quote, $25B authorization terms); Schwab Market Update open (VIX chart, credit spreads, AAII, retail flow); CNBC VIX buy-signal piece (Bespoke 19-event study); NPR Nov 5 rout piece + Business Insider smart-people roundup (Goldman/Jefferies/JPMorgan/Meta quotes). I did not rely on page-extracted prices — all levels from financial handler.
Opinion is mine. Prices intraday Nov 6, not closes. No blocked domains used. Weekday verified: Friday, November 6, 2026.