Watchlists

Snapshot of a publicly documented experiment in which Claude managed a real $50K equity book. Weights are as of May 1, 2026, just after the ARDX position was opened; the model rebalances on its own schedule and this page is not updated when it does. A reference for what a large language model picks when given a real book and a wide universe.

Ten growth names across themes, listed on April 25, 2026 as candidates to compound over five years: AI cloud and connectivity (NBIS, CRDO), photonics and edge AI silicon (SIVE, BZAI), e-commerce, fintech and payments (MELI, SEZL, DLO), telehealth and GLP-1 (HIMS, NVO) and subsea robotics (KRKNF). A starting point for research, not a forecast.

Companies positioned to benefit from Trump-administration policy: government ownership stakes, revenue-sharing arrangements, Golden Dome defense and other policy-favored themes. A map of policy exposure, not a list of recommendations.

A record of the AI buildout's biggest 2026 gainers through May: the 15 names of a year-to-date leaderboard posted on X on May 2, 2026, plus ARM, STX and CIEN, added on May 31 (+140% to +213% year to date by then). Not updated since. Heavy concentration in optical transceivers (AAOI, LITE, VIAV), compound semis (AXTI, NVTS), memory (SNDK, WDC, MU), semi test (AEHR) and datacenter power (BE, VRT): the picks-and-shovels of the capacity bottleneck.

Holdings of the new Tema Space ETF (US-listed names only). Excludes SpaceX (private), OHB (Frankfurt-listed), FPLSF (likely OTC EU).

Hardware names next to the photonics and power themes, listed on April 25, 2026: display and RF chips (HIMX, QRVO), SiC and power conversion (WOLF, VICR), laser and electronics manufacturing (LPKF, SHMD) and defense electronics and test (NOC, TRT), among others.

The full-stack AI buildout mapped as 6 layers - silicon → systems → memory → optics → cloud → power. Bottleneck rank annotated where applicable: Memory is the #1 constraint today (HBM sold out through 2026), Networking/Optics is #2 (bandwidth demand exploding with inference), Power/Grid is #3 (multi-year structural). Capital rotates layer-by-layer as each bottleneck binds.

KOSPI exposure for US-based investors. Most names trade as OTC ADRs; KRX-direct names (.KS suffix) need IBKR or a similar broker with Korean cash-equity access. The case as of May 2026: the Korea Discount is narrowing (Value-Up Index +130% since Sep 2024, KOSPI +75% year on year, Macquarie projecting +48% earnings growth for 2026, bank payout ratios mandated to rise from 36% to 50%+ by the end of 2026), and the holdco names (HYMTF, SSGQF, 402340.KS) trade at 40-50%+ discounts to NAV. The memory leg overlaps with the AI memory bubble.

PEG-ratio snapshot of 21 popular AI / mega-cap names, ranked high → low. PEG > 2 is the editorial "danger zone" (rich vs. expected growth); PEG < 1 flags potentially mispriced growth. Notable: NVDA, AMD, AVGO, MU all sit at PEG < 1 despite being the most-watched names in the AI buildout - growth still outpaces the multiple. PEG moves daily; this is a one-day snapshot, not a live tracker.

10 popular AI / mega-cap names trading below the S&P 500's 2026 forward earnings multiple (~22-23x). Ranked cheapest first: MU at 6x, SNDK at 8x, then ADBE, TTD, PATH, CRM, QCOM, META, OKTA, NOW. Convergent with prior screens - MU and SNDK keep showing up across YTD performers, the DRAM ETF, and the PEG snapshot. Multiples move daily; this is a one-day snapshot, not a live tracker.

The entire photonics value chain mapped end-to-end, from MOCVD reactors to AI-cluster optical switches. Front-end fabrication (layers 1-4) covers equipment, photonic ICs, discrete components, and substrates. Back-end fabrication (layers 5-8) covers transceivers, packaging, test, and systems. Each layer flags one "top pick" (★) per The Analyst's framework. Companion to the $LYTE ETF coverage - same universe, different lens: the ETF wraps the thesis, this watchlist breaks it down by capability layer.

39 publicly-traded space enablers across 4 layers - Orbital Launch, Manufacturing & Supply, Satellite Operators, Space Infrastructure. Editorial thesis: 2026 is the "summer of space" with SpaceX IPO as forcing function (~92% implied prob, rumored $2T valuation), Golden Dome missile-defense buildout, and rising launch + satellite + defense demand. SpaceX itself is private and excluded; ULA exposure is via its parents BA and LMT (placed in M&S). Heavy in defense primes (RTX, LMT, NOC, BA, LHX, GE, HON) and pure-play sat ops (RKLB, ASTS, PL, IRDM, BKSY).

49 momentum-ranked stocks across three editorial tiers - Market Leaders (go-to names), Next-Gen Leaders (young high-growth), and Speculative Leaders (story stocks with momentum). Sourced from fullstackinvestor.co's leaderboard, snapshot taken 2026-05-03. YTD performance ranges from +47% (Aeluma) to +458% (AXT). Heavy Memory + Power + Optical/Photonics representation in the top tier - convergent with our DRAM bubble + Datacenter Power theme.

The five stocks @FIREDUpWealth named on X on May 3, 2026 for the next decade of the AI cycle: semiconductors (NVDA AI GPUs, AVGO custom chips, TSM foundry), cloud AI (MSFT Azure) and next-gen healthcare (LLY GLP-1). The selection and framing are the source's, listed here for reference, not as advice.

30 semiconductor stocks with > $50B market cap ranked cheapest first on next-fiscal-year forward P/E. Memory + Asian names dominate the cheap end (SK Hynix 4.3x, Samsung 5.1x, MU 5.4x - all PEG <0.3); Arm sits at the top at 98.5x / PEG 4.67. NVDA, AVGO, AMD all sit at PEG <0.5 - convergent with the 038 PEG snapshot. Includes 9 foreign tickers (OTC ADRs + 2 China-domestic) so PEG and growth comparison spans the full global semi universe; foreign tickers do not get auto-populated earnings dates. PEG = fwd P/E ÷ next-year EPS growth; N.M. when growth is flat/negative.

18 semiconductor names ranked by 1-month return as of 2026-05-02. Editorial signal: a sector beta catch-up month - laggards led (INTC +107%, CRDO +92%, ALAB +91%, AMD +72%) while YTD winners cooled (NVDA last at +13%). Memory mid-pack (MU +47%) despite cheapest on PEG. Design tools rallied (SNPS +23%, CDNS +22%, ARM +36%). 1-month windows roll daily; the date is in the slug.

Roundhill's full ETF lineup grouped by sector: single-ticker baskets spanning memory, AI, big tech, robotics, gaming, luxury, and factor strategies. Click any ticker for the live ETF page.

The 11 publicly traded chip suppliers inside an iPhone 15 Pro, mapped to the three boards they sit on (Logic / Memory / Radio Frequency). Sourced from Quartr's teardown infographic of the device. Apple-revenue concentration matters here: CRUS is ~80% Apple-dependent, SWKS ~55%, QRVO ~30%, while diversified names like QCOM/AVGO/TXN have Apple as one of many customers. Useful for read-throughs from iPhone unit-volume data into single-name supplier earnings.

Companies Donald Trump has personally praised or championed on the public record - investment-commitment shout-outs, White House events, named CEO endorsements. Rhetorical-signal basket, not a 13F.

Companies NVIDIA CEO Jensen Huang has publicly praised or named as flagship partners - GTC keynote shout-outs, earnings-call name-drops, joint AI-factory announcements. The "Jensen on stage with them" book.

Public-equity map of the humanoid + industrial robotics stack. Two layers - Brain (compute/software/semis) and Body (actuators/sensors/connectors/automation) - sourced from @randgroup. 50 picks across 13 categories. Underlying thesis: $725B AI capex in 2026, Tesla Optimus mass-production ramp, Hyundai humanoids already in service.

State-of-the-market scan across the robotics narrative: humanoid pure-plays, industrial automation incumbents, actuator/motion suppliers, sensor/vision component makers, enabling semis, and China/Asia exposure. Companion to the CCXI/AGLT (Agility Robotics) and robotics bubble refresh.

Pure-play AI-driven drug-discovery platforms - generative biology, computational chemistry, antibody design, and protein-structure prediction applied to therapeutic pipelines. Companies where AI is the differentiator, not a buzzword bolted onto a traditional biotech. Seed members are 2026's 4 cleanest public listings on this thesis; the basket will grow as new IPOs land and Insitro-class private names go public.