Full Scan Market Brief — May 1, 2026

Issue #17 · The 30% Parasitic Loss Edition · ~14 min read

1. Three Things

1. Mega-cap tech went vertical, all at once — INTC RSI 87.5 (+97.7% 30D), GOOGL RSI 82.1 (+30.4% 30D), AMD RSI 79.4 (+65.8% 30D), AMZN RSI 77.3 (+27.9% 30D), MRVL RSI 76.7 (+54.1% 30D). Five trillion-dollar-class names simultaneously parabolic is not normal — it's a regime, and it's stretched.
2. NuttyCLD's 6-piece AI Power Crisis series rewrites our AI thesis — the 30% parasitic energy loss between grid and GPU is a whole second floor under Aschenbrenner's watts-per-rack. Surfaces a new VRM/PMIC/wide-bandgap/equipment cohort (MPWR, AOSL, VICR, IFNNY, NVTS, POWI, BESIY / AMS:BESI, MRAAY) we don't track. New child perspective ai-power-delivery opened today.
3. Defensives are failing the defensive role — VZ RSI 27, T RSI 29, MCD RSI 32, gold weak-down (GLD -17% from ATH and not bidding on anything), VXX -64% from peak. When defensives can't defend, the only hedge left is cash. Pelosi (+12.5% YTD via AMZN +13.8%, COIN, KKR) and Druckenmiller (+6.2% via BE +194%) are eating the index.

2. The Big Picture

Act 1 — The parabolic concentration. Five mega-caps are simultaneously ripping into 52-week highs: INTC, GOOGL, AMD, AMZN, MRVL all RSI 76-87, all 30D returns +27% to +98%. SOXL +50% in 30 days. SMH at all-time highs. QQQ within 0.04% of ATH. Six months ago you wanted "AI infrastructure exposure" and the trade was NVDA + AVGO + ANET. Today the tape is voting every layer of the AI buildout — semis, foundries, optical, networking, packaging — and rewarding them all simultaneously. INTC's RSI 87.5 is the most extreme single-name reading we've ever logged. That's a comeback story that became a momentum chase in the span of one quarter. Trim signal, not entry.

Act 2 — The new floor under the AI thesis. Our ai-power-bottleneck perspective (Aschenbrenner's watts-per-rack, BE the headline) was already paying — BE +89% 30D, RSI 67, +1,667% 1Y. Today's NuttyCLD capture changes the framing. Six articles published March 30 to May 1 walked the entire voltage stack from power plant to transistor: HVDC at the grid, 800V DC at the rack, VRM/VPD topologies on the board, BSPDN at the die. Same problem repeats four times: shorten the path, move the loss, waste less. NVIDIA's own analysis: 30% of all power lost to conversion before reaching the GPU. At a 1 GW data center that's 300 MW of "parasitic energy" turning into heat. The investible names are the ones we don't track — MPWR (projected ~70% Vera Rubin VRM share at +60% ASP), AOSL ($900M cap, OpenVReg co-leader on GB300), VICR (Q4 record backlog $176.9M, Phil Davies $200-400/XPU), BESIY / AMS:BESI (hybrid bonding sits at the BSPDN+HBM convergence). New child perspective ai-power-delivery opened. Coverage gap — surfaced for next session's deep dive queue.

Act 3 — The bid that isn't there. While risk assets rip, the supposed defensives are broken. VZ RSI 26.9 with -9.5% 30D. T RSI 29.2. MCD RSI 31.8 (first proper oversold print in months). HRL strong-down -33% from high. GLD weak-down RSI 44, -17% from ATH, not bidding on Iran or oil or rate cuts. GDX -12% in five days. VXX -16% in seven days, -64% from peak. The only "defensive" trades that worked this quarter are XLU (utilities) and the discount-retail compounders (COST, TGT, TJX, WMT) — and those aren't really defensive plays, they're just well-managed. When defensives fail to defend, the asymmetry shifts: capital that should hedge is forced to hold cash or chase the rally. That makes the parabolic mega-cap concentration in Act 1 more brittle, not less.

3. Focus List

Take, in one paragraph: the focus list is a full-throated risk-on book. Eight of ten near 52-week highs. TSLA is the lone broken name. NVDA cooled from blow-off RSI 91 in early April to a healthy 71 — the best tape in the cohort because the trend is intact and consensus has digested. BE is the highest-conviction position currently held in any persona (Druckenmiller: +194.4% open). COHR May 13 earnings is the next binary catalyst on the focus list.

4. Sector Scorecard

5. Market Vibe

The tape is in a strange shape: simultaneously broad-and-stretched, with the conspicuous absence of any working hedge. SOXL +50% 30D and IWM +11% says breadth is real. INTC RSI 87 says the leaders are climaxing. LULU RSI 29.6 and ZS RSI 32 say there's also real selling, just selectively. And gold/VXX/TLT all flat-to-weak says nobody is hedging. That's the puzzle: how can you have parabolic concentration AND broken defensives AND no fear bid simultaneously? The answer is positioning. Vol has been crushed for weeks; systematic vol-target funds have re-leveraged; the chase is mechanical. When this kind of structure breaks, it tends to break fast — the Wile E. Coyote moment is when leaders break trend and there's no hedge bid to absorb the unwinding.

The single biggest insight from today is the AI power-delivery thesis captured from NuttyCLD. We've been long the generation layer (BE, CRWV, CORZ, IREN, EQT) for three months and it's working. Today's catalyst rewrites the framework: the delivery layer (everything between the grid and the transistor) has a clean, identifiable supply chain that the EV industry already built — Porsche/BYD/Hyundai 800V SiC + GaN ecosystems are now being consumed by AI as a second demand curve. The names that capture the value (MPWR/AOSL/VICR/IFNNY/NVTS/POWI/BESIY/MRAAY) are not in our watchlists. That's a coverage gap, and it's the cleanest "catch up" trade on the board if you don't want to chase INTC at RSI 87.

The third pulse: Pelosi's portfolio is the single best-performing book we track at +12.54% YTD, ahead of every algorithmic persona we run. AMZN +13.84% open is doing most of the work. The fact that disclosed congressional trading beats our infrastructure of CLIs, scoreboards, and personas suggests our edge is in the thesis surfacing (which we just got fresh ammo for via NuttyCLD), not in stock-picking discretion at the margin. Lean into the surfacing, not the picking.

6a. Scan Dashboard

Correction 2026-05-18: this dashboard has 25 scan rows. Only 4 rows persisted as scan files on disk this run; the other 21 rows were captured from validated daily summaries and issue context, which remain the source of truth.

6b. Expanded Dispatches

monster-scan: 187 monsters, breadth widening. AXTI +346% above SMA200 (the most extreme small-cap parabolic we've ever tracked). AAOI +257%. LWLG +233% (NuttyCLD CPO validation). Five mega-caps just joined the >100% vs SMA200 club. Critical signal to watch: when monster count rolls over (currently 187 vs 173 Apr 28), that's the rotation tell. → [Full data in monster-scan.json]

defensive-scan: defensives can't defend. VZ RSI 26.9 with -9.5% 30D. T RSI 29.2. MCD RSI 31.8. HRL deathcross -33% from high. PPC -27% 3M. The only "defensives" working are XLU (utilities) and discount retail (COST/TGT/TJX/WMT). When the safe-haven bench is broken, the implied vol regime is fragile — the unwind, when it comes, has nowhere to flow.

bargain-bin: defense triplet is the entire bin. Strict filter (RSI<30 + vs SMA20<0): LMT (RSI 23), NOC (RSI 22), LHX (RSI 29), LULU (RSI 30). Defense primes are quality monopolies in golden cross + extreme oversold = textbook accumulation. LULU is the contrarian brand bet (-61% from high). Skip ZS/SNOW/WDAY/VEEV/TEAM — those are busted growth, not bargains.

optical-supply-chain: every name is a monster, four-wave framework holds. LITE Wave 1 cooling into COHR May 13 earnings (clean setup). TSEM (Wave 3) +51% 30D. LWLG (Wave 4 CPO) +112% 30D — speculative tier. NuttyCLD Part 4 explicitly validated CPO joining the AI fabric story from the power-delivery side: light moves in when copper can't carry the I²R loss. Cross-thesis convergence on COHR is now four paradigms deep.

ai-power-bottleneck: thesis got a second floor. BE +89% 30D / RSI 67 (Aschenbrenner trade fully paid). NuttyCLD adds: 30% parasitic energy loss frame surfaces a whole new VRM / wide-bandgap / equipment / MLCC supply chain (MPWR, AOSL, VICR, IFNNY, NVTS, POWI, BESIY / AMS:BESI, MRAAY) where consensus hasn't priced in. New child perspective ai-power-delivery opened today — this is the most important new framework in months.

7a. Perspective Dispatches

Gold Crash (Recovery Dead, Pivot to Silver). GLD lost another $19 since Apr 26, GDX cracked -12% in 5 days. Recovery thesis dead. Pivot to silver: SLV held flat while gold fell, SIL +5% week. Status: ACTIVE, working. → 2026-05-01-gold-crash.md

Oil $200 Scenario. Demoted to tail risk only. USO -10% 7D, fertilizer cascade reversed (CF -11% 30D), VXX -64% from peak. Trigger never fired. If nothing fires by June 1, retire entirely. → 2026-05-01-oil-200-scenario.md

Crypto-Geopolitics. BTC $76,315 trigger held barely (RSI 56 cooled below 65 threshold). Miner equities exploding — RIOT +22% 7D, MARA +41% 30D, CLSK +44% 30D, MSTR +36% 30D. AI compute-pivot story crystallizing. → 2026-05-01-crypto-geopolitics.md

War Ends Playbook. Concentrated to mega-cap tech + Asian semis: AMZN +29% 30D RSI 73, GOOGL +28% RSI 71, NVDA +29% RSI 71, MSFT +20% RSI 65, EWY +33% RSI 62, EWT +27% RSI 73. 13 failed branches retired (housing/India/consumer). Renamed in spirit to "risk-on tech-rotation." → 2026-05-01-war-ends-playbook.md

Gulf Infrastructure Strike (Helium Clock). Partial hit, narrative bifurcating. APD/LIN/LNG working US-side. Asia-fab-pain leg gone quiet — TSM/EWT/EWY all strong-up, market ignoring helium tail risk. Aug/Sep depletion deadline still binds in 90-120 days. → 2026-05-01-gulf-infrastructure-strike.md

Protein Economy (HIMS Pivot). HIMS pivot still right but entry elusive — RSI 58 (we want <55) after +49% 30D. Correction (post-scan, 2026-05-02): LLY actually prints $963 RSI 56 strong-up with +6.4% on the week — NOT the "$874 RSI 38 weak-down" originally reported here. Cohort weakness on MRK/JNJ/GILD/AMGN/MDT is real and confirmed, but the leader bounced. Channel-beats-SKU thesis partially weakens if LLY holds. CVS +12.7% 30D RSI 63 still the regulatory-safer adjacent vehicle. → 2026-05-01-protein-economy.md

Optical Supercycle (Wave 4 Validated). Wave 1 consolidating into COHR May 13 earnings. Wave 3 ripping (TSEM +51% 30D, ALAB +83%). Wave 4 quietly validated by NuttyCLD's CPO endgame — power-budget play, not just bandwidth. Cross-thesis on COHR now 4-paradigm convergence. → 2026-05-01-optical-supercycle.md

Food Security Cascade. Stage 3 (commodities) is where the action is — DBA RSI 74, CORN/SOYB strong-up, ADM at new highs. CF cooled -11% 30D but +30.8% 3M intact. MOS still locked in death cross — Stage 4 (policy response) STILL hasn't fired. → 2026-05-01-food-security-cascade.md

US Energy Dominance (Late-Cycle Divergence). HOLD with conviction. Canadian premium cleanest expression: ATH.TO +156% 1Y, BTE.TO RSI 70 +207% 1Y. Internal split: integrateds cooling, independents holding, services outperforming both — textbook late-cycle divergence we explicitly named in the prior watch-for. → 2026-05-01-us-energy-dominance.md

AI Power Bottleneck (Watts Per Rack + Last Micrometer). ESCALATED. BE +89% 30D paid out. NuttyCLD AI Power Crisis series adds NEW dimension: power-delivery loss (30% parasitic = 300 MW at 1GW data center). Sister perspective ai-power-delivery covers the delivery side. Bloom Energy "side story" reinforces BE position vs SMR cooling. → 2026-05-01-ai-power-bottleneck.md

AI Power Delivery — The 30% Parasitic Loss (NEW). Six orders of magnitude voltage drop, redesigned at every layer. Nutty's framing: EV supply chain (Porsche/BYD 800V SiC + GaN) crossing into AI as second demand curve. Validated by Vertiv +152% organic orders Q4, Delta passing Foxconn as Taiwan #2, Intel 18A Panther Lake ramping 6-7x in Q2. Coverage gap: MPWR, AOSL, VICR, IFNNY, NVTS, POWI, BESIY / AMS:BESI, MRAAY all need triage to watchlist. → 2026-05-01-ai-power-delivery.md

7b. Monster Watch (Top 8 most-interesting)

Theme cluster: AI infra (optical + power + semis) is 80%+ of the parabolic tier. AI Power leg (BE/POWL/AGX/FIX) is now caught up to optical in concentration. When 187 → 173 = first breadth contraction signal.

7c. Deep Dive Spotlight

No fresh deep dives this scan. TASKS.md has a queue of 9 deep dives lined up for next session: NuttyCLD VRM cohort (MPWR, AOSL, VICR, IFNNY, NVTS, POWI, BESIY) + RDDT + PLTR. Run /deep-dive TICKER per name when ready.

8. Key Signals

Gold / Commodities — Pivot to silver. GLD weak-down RSI 44, GDX -12% in 5 days. SLV held flat, SIL +5% week. Gold:silver ratio compressing toward 6.0. Theo's industrial-demand silver thesis is the only piece of the precious-metals complex showing real signal. Entry: SLV >$71 confirms; gold:silver ratio <6.0 = silver leading materially.

Insider Buying — Defense triplet is the cleanest play. LMT/NOC/LHX all at RSI 22-29 with golden crosses still intact and positive 1Y returns. Quality monopolies in deep oversold. GD broke up +9.9% 7D on submarine-budget catalyst chatter — first prime to decouple. The bargain-bin's strictest filter pulls only these four names + LULU.

Crypto — Speculation returning at the edges, regime unchanged. BTC +14.7% 30D bounce but death cross intact, RSI 56 cooled. DOGE RSI 73 (canary firing). Miners parabolic (CLSK +44%, MARA +41%, RIOT +22% 7D). MSTR +48% 30D. None of it cracks SMA200. The April rally is a vote that the bottom is in; the SMA200s say "prove it."

AI Power Delivery (NEW) — The 30% parasitic loss frame. NuttyCLD's six-piece series captured today opens a sub-thesis we didn't have a perspective for. EV supply chain crossover (Porsche/BYD 800V SiC + GaN ecosystem) crossing into AI as second demand curve. Companies that capture value: MPWR (~70% projected Vera Rubin VRM share at +60% ASP), AOSL ($900M cap, OpenVReg co-leader), VICR (record backlog, $200-400/XPU framing), BESIY / AMS:BESI (hybrid bonding), MRAAY (GB300 = 30,000 MLCCs, 13-15x traditional server). Coverage gap → triage candidates for next session.

The Late-Cycle Energy Divergence — HAL/SLB/OIH outperforming producers. Services lead producers when capex is finally being deployed. RSI 72 OIH already overbought. Two ways this resolves: (a) services keep running, producers catch up = bullish for whole complex; (b) services peak first, producers flatten = textbook end-of-cycle. Trim signal on OIH if it stays >70 for another two weeks.

9. The Wild & Whacky

• INTC RSI 87.5 + +98% 30D + +383% 1Y — comeback name became momentum name. Asymmetric downside risk if rotation breaks.
• BE +194.4% open in Druckenmiller-13f mirror — single best open position across all 12 personas we track.
• Pelosi mirror beating every algo we run at +12.54% YTD. Disclosed congressional trades > our personas.
• AXTI +346% above SMA200 — most extreme small-cap parabolic ever logged.
• LMT RSI 23 / NOC RSI 22 / LHX RSI 29 — three quality monopolies simultaneously in extreme oversold with golden crosses still intact. Has not happened in 18 months.
• MTUM RSI 73 + GLD weak-down + VXX -64% from peak — momentum leading while every classical hedge fails. Implied vol regime fragility.
• Pizza franchise breaking — DPZ -32% from high RSI 35 strong-down. YUM strong-up RSI 57 +3% 3M. Same category, opposite tape — brand positioning beats category.
• 30,000 MLCCs per GB300 server (per Murata, via NuttyCLD) — 13-15x traditional server. The MLCC bottleneck just became core AI infra; we don't track Murata or Samsung Electro-Mechanics. Coverage gap.
• TSLA the lone broken mega-cap — RSI 39, -29% from high, while INTC/AMD/GOOGL/AMZN/MRVL all near 52wk highs.
• Gold AND dollar both weakening simultaneously — atypical. Suggests positioning unwind, not flight to safety.

10. Paper Trade Report Card

Today's exits (Phase 5a auto-execute):

• claude-momentum SOLD LLY @ $963.33 — +6.62% / +$199 (RSI recovery exit at 57.8)
• claude-momentum SOLD GD @ $345.84 — +10.42% / +$417 (RSI recovery exit at 58.0)
• claude-trader SOLD GOOGL @ $385.69 — +35.88% / +$1,794 (profit target +30% hit)

GOOGL was the play of the scan. Bought at $283.84, sold at $385.69 — five-week hold, 35.88% return.

Current leaderboard (vs DCA Larry +0.26%):

Best open: druckenmiller-13f BE +194.4% (+$2,783). Worst open: druckenmiller-13f RH -34.74% (-$122).

Danger zones: ABT bench-signals -11.8% RSI 28.8 (urgency 25 — proper capitulation). NUGT yolo -10.3% RSI 38.7 down trend (urgency 25). RBLX ark-innovation -11% strong-down. AMZN pelosi RSI 77.3 — sell signal candidate at next RSI > 80 print.

Best trade thesis working: BE in Druckenmiller mirror. Aschenbrenner watts-per-rack thesis fully validated, position +194% open. Worth a ladder trim if it pushes RSI > 80 again.

Worst position: PLTR in Burry-13F mirror at -16.76% (-$11,065 unrealized). The contrarian voice held against the AI infra rally.

11. What I'd Tell a Friend

If you woke up tomorrow with $50K and asked me what to do, here's the ranked list:

1. LMT $510-515 entry, stop $475, target $620 — RSI 23, -26% from high, golden cross intact, quality monopoly, DOD budget intact. The cleanest setup on the board. Sized 5%.

2. NOC $560-575 entry, stop $540, target $685 — RSI 22, -27% from high. B-21 program is the catalyst. Sized 4%.

3. CF $120-123 entry, stop $108, target $145 — Strong-up + golden cross + RSI 50.6 sitting on SMA20. Cleanest food-security cascade entry; if natgas firms or any geopolitical shock rolls through, CF benefits immediately. Sized 3%.

4. MCD < $295 entry, stop $270, target $330 — RSI 31.8 first proper oversold print in months on a quality franchise. Sized 2-3%.

5. HIMS waiting at RSI < 55 (currently 58) — protein-economy pivot vehicle. +43% 30D run still has room but want a deeper entry. Sized 2% on entry trigger.

What to watch this week: COHR Q3 earnings May 13 — Wave 1 demand confirmation, primary near-term catalyst. INTC's RSI 87.5 — does it climax to 90+ or break trend? BTC reclaiming $84,019 SMA200 — regime change trigger. Pelosi's next disclosure window — copy if there's anything new. AMZN above $265 sustained = breakout to ATH; below $250 = trim.

12. Active Perspectives Table

13. Scan Summary

• 25 dashboard scan rows analyzed (4 persisted as markdown files this run; 21 captured in summary JSONs + this brief)
• 11 perspectives refreshed (10 existing + 1 new — ai-power-delivery)
• 3 paper exits executed (LLY +6.6%, GD +10.4%, GOOGL +35.9% — all RSI-recovery or profit-target rules)
• Active deep-dive queue: 9 names (NuttyCLD VRM cohort + RDDT + PLTR)
• Data fetch: all 666 unique symbols across 60 watchlists refreshed
• Source date: 2026-05-01 (Friday EOW)
• Previous brief: Issue #16, Apr 28

---

The dominant editorial signal this issue: AI Power Delivery is the new frontier. Generation thesis (BE, CRWV, EQT) still working; delivery thesis (MPWR, AOSL, VICR, BESIY / AMS:BESI, MRAAY) is the next layer the market hasn't priced. NuttyCLD's six-piece series captured today is the catalyst.

Sources

Price, RSI and trend figures read from the desk's validated daily scan summaries (research/market-engine/data/summaries/). No number in this note was computed in prose.