Curator Forwardtest (CFT)

Rationale: - **rockets_spacecraft**: SpaceX (SPCX) is trading below its $135 IPO price after a scrubbed/aborted Starship-13 attempt (news.bitcoin.com, cnbc.com, arstechnica.com, 2026-07-16/20), and while the company is real, a newly-IPO'd stock sliding on a launch abort is a live-catalyst-risk story, not a clean buildup signal -- and current holdings RKLB and AVAV already give the wave real exposure via a name proven through a full quarter, not a week-old post-IPO air pocket. China's reusable Long March 10B landing (bbc.com, 07-10) is a state-program milestone with no US-investable ticker. No swap warranted. - **nuclear**: Steady incremental buildup news (Centrus DOE contract expansion past $1B, four US reactors hitting criticality milestones, Hyundai/FANCO SMR tie-up, Kola siting licence) confirms the wave is intact but none of it names a fresh US-listed ticker outside the CCJ/BWXT/OKLO/SMR/VST already held -- this is confirmation, not a new catalyst. - **quantum**: QUBT and a 'quantum stocks set for upside' listicle (finance.yahoo.com, 07-09/17) are pure momentum pieces with no concrete new commercial catalyst, and none of the current holdings sit in this bucket already -- IonQ partnership news (Jane Goodall Institute tie-up) is a PR-grade research collaboration, not a revenue signal. Not strong enough to displace a held name. - **robotics**: Humanoid (Europe) Series A at $1.35B and Mitsubishi/Highlanders humanoid MOU are real buildup signals but both are private/foreign, no clean US-listed line; Pudu Robotics ranking and the Forbes 'first humanoid strike' piece are colour, not catalysts. Nothing investable surfaced. - **AI**: Google's near-doubling of AI infra capex and the Z.AI domestic-chip milestone are dominant-wave confirmation, not new-wave discovery -- profile rule is to trim AI before the crest, not chase further; no action indicated for NVDA/AMD/GOOGL, which the optimizer already weights on trailing performance. - **geo_defense**: Hypersonic-weapons listicle and Golden Dome shipyard contract are the closest things to a fresh name (Hanwha Philly Shipyard), but Hanwha's US listing exposure is thin/foreign-parent and the piece is a single vessel contract, not a re-rating catalyst; KTOS, LMT, AXON already cover this wave. - **geo_drones**: Ukraine's Lasar Group loitering-munition strikes and India's Tata/Nibe loitering-munition tender are real proof-of-concept and procurement news, but Tata Advanced Systems is not US-listed and no US pure-play attritable-drone ticker surfaced in the pool this window; noted for future coverage, no action now. - **geo_tankers**: Renewed US-Iran strikes over the Strait of Hormuz (BofA tanker-play piece, Guardian, Morningstar, 07-09/13) keep the tactical tanker thesis alive per the profile's own caveat that low global oil reserves sustain freight rates even without full closure -- but no ticker is currently held in this bucket and the pool named no specific investable vehicle beyond a vague BofA mention, so nothing actionable to swap in. - **geo_reconstruction**: EU's $1B Gaza reconstruction fund and Saudi Vision 2030 SDG review are directional confirmation of the wave's existence but neither names a specific engineering/construction/materials ticker -- pure macro colour, no catalyst to act on. - **demographics**: Alzheimer's blood-test and Sully diagnosis coverage plus health-care REIT background piece are wave-relevant but generic/backgrounder in nature, no company-specific inflection point named; LLY already anchors this bucket among current holdings. - **verdict**: nothing in this cycle clears the bar of a concrete, fresh, US-listed catalyst strong enough to justify displacing one of sixteen full slots -- SPCX is the closest call but its post-IPO volatility argues for watching another cycle rather than swapping out a proven name; watchlist composition and thematic shape are unchanged this period.

Rationale: - **rockets_spacecraft**: SPCX (SpaceX) dominates coverage -- Starship Flight 13 (first since its IPO) partially succeeded (booster engine failure on re-entry), stock whipsawing from all-time lows to a bounce; RKLB won a $266M Space Force suborbital contract and popped after-hours; Skyroot's Vikram-1 debut in India is not US-listed and not investable. RKLB is already held and its contract win reinforces the existing position; SPCX is a real catalyst-rich name but its extreme volatility and single-digit-week-old post-IPO price history argue for watching another cycle before displacing a steady incumbent. - **nuclear**: OKLO and X-Energy jumped on a Trump administration program pairing them with Microsoft/Nvidia for AI-data-center power plants -- a fresh, concrete catalyst, but OKLO is already held, so this is reinforcing, not a new-name case. The Saudi civil-nuclear agreement is a structural tailwind for the sector broadly (CCJ, SMR, BWXT already held) but names it deal from (US firms generally) rather than one specific new investable ticker. NuScale (SMR, held) has an August 5 earnings catalyst pending -- no action yet. LEU (Centrus, uranium enrichment) got a bullish comparison piece but is not a strong enough single-article case to displace a current holding. Nuclearelectrica's SMR proposal rejection and Brazil/Korea SMR items are foreign/non-investable noise. - **AI**: Google's 'Frozen v2' efficiency chip, Qualcomm hyperscaler chip deals, Nvidia/AMD data-center competition, and 'AI bubble fear' pieces are all coverage of names already held (GOOGL, NVDA, AMD) or non-actionable macro chatter -- no new catalyst-backed ticker surfaced. - **quantum**: Heavy volume of quantum news (IBM/HRL acquisition, Infleqtion Chicago hub, D-Wave laptop-match story, Quantinuum/SoftBank paper, Rigetti NSF award) but every distinct catalyst points to a different small-cap name (IBM, RGTI, IONQ-adjacent, D-Wave) with no single one standing out as a clear buildup leader over a full watchlist slot, and the profile's quantum wave already has no representative -- worth tracking but nothing here rises to displacing a current holding this cycle. - **robotics**: Samsung's new RX robotics division and Deutsche Bank's humanoid-market-acceleration note are notable, but Samsung is a foreign mega-cap conglomerate (no pure-play robotics investable line), and Humanoid (UK startup) and Pudu Robotics (China) are private/foreign, not investable. No US-listed pure-play surfaced. - **geo_defense**: LMT (held) had an enormous run of contract news -- new low-cost Patriot interceptor, $58.62B seven-year PAC-3 MSE award, $1.4B hypersonic work -- all reinforcing an existing position, not a new name. Hanwha Philly Shipyard's $2B missile-tracking vessel deal is a Korean-parent OTC-adjacent line, not a clean liquid US add. - **geo_drones**: Hypersonic-threat and counter-drone commentary (Thales, China hypersonic display) referenced only foreign or already-covered primes; KTOS (held) is the existing attritable-autonomy exposure and nothing here displaces it. - **geo_reconstruction**: Middle East reconstruction commentary (cranes/energy grid rebuild pieces) is thematic scene-setting with no specific investable US ticker named yet. - **geo_tankers**: Only a generic tanker-news index page surfaced, no ticker-specific catalyst. - **demographics**: No coverage this window. - **verdict**: every catalyst this cycle reinforced an already-held name (RKLB, OKLO, LMT, KTOS, GOOGL, NVDA, AMD, SMR/CCJ/BWXT) or pointed to a non-investable/foreign/too-early name (SPCX volatility, Samsung robotics, Humanoid, quantum small-caps scattered across too many tickers to pick one). No candidate cleared the bar of beating the weakest current holding paired with a citable fresh catalyst, so the watchlist stays unchanged this biweekly window.

Rationale: - **AI**: Nothing new -- Qualcomm hyperscaler deals and Nvidia/AMD/Intel share-of-market pieces are old-news rehashes about names not on the watchlist; AMD and NVDA already held and no fresh catalyst for either beyond routine competitive chatter. No action. - **rockets_spacecraft**: SpaceX (SPCX) dominates the pool -- Starship Flight 13 booster failure, moon-crash debris, first post-IPO earnings, and a stock down >50% from its highs. This is a crested/de-rating story, not a buildup one, and SPCX is not on the watchlist anyway. Skyroot's India test launch is a private, non-US-listed company -- not investable. RKLB (held) gets no direct news in this pool. Voyager Technologies (VOYG) posted a record-revenue, guidance-raise quarter -- a genuine positive catalyst, but it is a brand-new IPO (first quarterly report just posted) with unclear listing depth and no second data point yet; not compelling enough to displace a current holding this cycle. No action. - **nuclear**: Mixed and mostly noise -- Nuclearelectrica's SMR study was rejected by its own ministry (a negative for a non-US, non-held name), LEU vs SMR is a comparative fluff piece, and Mangoceuticals/Nuclea Energy is a micro-cap reverse-merger microreactor SPAC-style deal, too speculative and too small to be a liquid, real catalyst. OKLO, SMR, CCJ, BWXT, VST all held already with no fresh news this window. No action. - **quantum**: IBM's HRL Laboratories acquisition and CEO comments about 2028-2029 revenue impact are IBM-specific (not on watchlist, not a pure-play), and the 'stocks that can soar' / listicle pieces (Rigetti, photonic quantum roundups) are generic content-mill roundups with no dated concrete catalyst for a specific ticker. No name here clears the bar for a fresh, citable catalyst. No action. - **geo_defense**: Lockheed Martin (LMT, held) has a genuinely strong run of news -- a $10.5B SOCOM logistics IDIQ, a fresh $58.62B multiyear PAC-3 MSE award, and a Pentagon deal with LMT and Northrop to quadruple THAAD interceptor production. This is buildup-continuation for an existing holding, not a new-name catalyst, and per the profile's guidance on already-re-rated primes, this is confirmation to let the optimizer's weighting handle it rather than a reason to touch the watchlist. No action. - **geo_drones**: Only a generic loitering-munition market-research report surfaced, naming no specific investable US-listed ticker. No action. - **geo_tankers**: Iran's reported attack on two tankers in the Strait of Hormuz pushed WTI/Brent higher -- a reminder the tactical tanker thesis is still live, but the pool names no specific tanker-operator ticker to evaluate, and none is currently held. No action. - **geo_reconstruction**: Middle East reconstruction commentary (cranes, energy-grid rebuilding, Gaza infrastructure) is all thematic prose or non-US-listed small caps (e.g., an ASX gold miner mentioned only incidentally) -- no liquid US-listed pure-play surfaced. No action. - **demographics**: No relevant coverage in this pool. No action. - **verdict**: This window's pool is dominated by noisy, already-crested news (SpaceX's post-IPO stock collapse) and confirmatory-but-not-new news for names already held (LMT's contract wins). No candidate cleared the bar of a fresh, concrete, pre-run-up catalyst strong enough to displace a current holding, and the watchlist is full, so per the profile's guidance an empty slot beats a weak swap. Holding the current 16 unchanged this biweekly cycle.

Rationale: - **rockets_spacecraft**: Sector broadly re-accelerating -- ASTS, RKLB, RDW, LUNR all named as leaving SpaceX behind on satellite launches, defense contracts and lunar milestones; Musk called the space boom 'much bigger' than Goldman's $1.8T forecast (2026-08-19), and Stifel upgraded LUNR on accelerating delivery. Already hold RKLB and AVAV/KTOS-adjacent exposure; no swap needed since the wave is broadening rather than concentrating in one new name. - **nuclear**: Split tape -- BWXT ($8.65B backlog) and CCJ (uranium spot $88/lb, +34% YoY) keep monetizing the AI-baseload story, while NuScale (SMR) reported $75K Q2 revenue and a new $750M dilutive raise, down 35% YTD, and a Yahoo/Simply Wall St piece flagged that only 2 of 5 hyped nuclear names (CCJ, LEU) actually sell fuel today -- SMR and OKLO sell 'construction timelines.' Removed SMR as the weakest-conviction holding; kept OKLO since its SMR technology progress this week was framed positively (ETF inclusion angle) versus SMR's outright revenue miss and dilution. - **quantum**: IonQ's 287% revenue surge and SkyWater acquisition keep dominating headlines, but multiple pieces this week specifically flag RGTI as the highest-upside name (81% analyst upside, $100M CHIPS Act LOI, $541M cash/zero debt) -- a fresher, less-run buildup name than IONQ, which has already re-rated hard. Added RGTI, paired against the SMR remove. - **AI**: NVDA ($82B revenue, +85% YoY), AMD (+121% YoY, doubled data-center revenue), and the Musk/SpaceX-NVDA 10GW moonshot dominate coverage -- all names already held (AMD, NVDA) or adjacent (GOOGL). Nothing here rises to a swap-worthy fresh catalyst versus current holdings; noise-filtered names like Credo, Nebius, Vertiv, Eaton are supply-chain plays with no current watchlist analog trimmed for them this cycle. - **geo_tankers**: VLCC/Hormuz rates near $500K/day, Brent flirting with $100, DHT and International Seaways posting record quarters -- tactical tanker thesis still live per profile note, but no tanker name is currently on the watchlist and none rose to displacing an existing holding this round; filtered as noise for now (watch for next rebalance if reserves stay tight). - **geo_defense**: LMT ($3B+ Pentagon deal with NOC on munitions, Kodiak live-fire test), KTOS (raised 2026 outlook +9.4%, new Javelin/missile-defense work), RTX ($22.9B Navy Tomahawk deal) all show continued buildup -- LMT and KTOS already held and performing; no swap triggered. - **geo_drones/hypersonics**: DARPA RFI for next-gen hypersonic cruise missile (Oct 2026 responses due) is early-stage procurement news, not yet a tradeable equity catalyst for any name not already covered by LMT/KTOS exposure; filtered as noise. - **geo_reconstruction**: Gaza Board of Peace issued its first construction contract (Arkel International, a private Baton Rouge firm with no public ticker) -- no investable public equity signal yet; filtered as noise. - **demographics**: American Healthcare REIT (AHR) and CareTrust REIT show strong senior-housing M&A and FFO growth, but neither is on the current watchlist and the news is more sector-confirming than a fresh discovery catalyst strong enough to displace a held name this cycle; filtered as background confirmation, not actioned. - **verdict**: One swap only -- RGTI in for SMR -- reflecting the biweekly cadence's bias toward inaction: SMR's revenue miss and dilution is a clear thesis-quality deterioration (not just maturation) while RGTI is a fresh, policy-backed buildup name in a wave (quantum) already represented but under-diversified within the watchlist. All other waves showed continuation of already-held themes (rockets, defense, AI) or lacked an investable, watchlist-displacing name (tankers, reconstruction, demographics), consistent with the profile's preference for an open/rotated slot over marginal padding.

Rationale: - **AI**: NVDA rallied 6.8% on a strong post-earnings guide and AMD/GOOGL keep drawing 'still not too late to buy' coverage alongside $735B of 2026 hyperscaler capex -- both already-crested-adjacent mega-caps I hold rather than chase further; no change, let the optimizer size them. - **rockets_spacecraft**: Heavy volume of bullish coverage (Musk's 'much bigger than $1.8T' space-boom comments, Trump's 1,000-launches policy push, Falcon 9 wind-down opening share for RKLB/LUNR/ASTS, LUNR's $600M defense win and Stifel upgrade) but RKLB is already held and well-represented; filtered out SPCX (not separately investable alongside RKLB in this slot count) and China's Zhuque-3/Chang'e-7 (foreign, non-investable). No new add needed -- current RKLB exposure already captures the theme. - **nuclear**: Sharply mixed tape -- OKLO and NuScale (not held) both hit by a well-documented $2B short-seller profit and 'bubble deflating' narrative, then bounced on no news days later, while BWXT (held) got a fresh, concrete US Army microreactor contract win on Aug 26. This is exactly the buildup-vs-peak split the thesis cares about: BWXT's catalyst is real and current, so no change to nuclear holdings; NuScale's cash-burn/short-seller story confirms I'm right not to hold it. - **quantum**: Genuine dispersion emerged this window -- IONQ, QBTS, and QUBT all posted concrete revenue-growth catalysts (record revenue, 12x bookings, 9,000% YoY growth) while RGTI's edge is a not-yet-booked CHIPS letter of intent; removed RGTI to fund the tanker add rather than adding a peer, since the watchlist is full and none of the uncovered peers cleared the bar for a same-period double swap. - **robotics**: Unitree's 460% Shanghai debut and 'ChatGPT moment for robot brains' commentary, plus KOID's 42% supply-chain-fund return vs HUMN's 11%, are compelling but center on foreign listings (Unitree, XPeng's Dogotix) or a fund I don't hold; no clean US single-name catalyst displaced a current holding this cycle. - **geo_tankers**: Frontline's record quarter is the standout fresh catalyst of the whole pool -- concrete, dated, and squarely matches the profile's tactical thesis that low global reserves keep freight rates elevated even as the Hormuz rerouting premium unwinds. Added FRO. - **geo_defense**: RTX's $22.9B Tomahawk win, Castelion's $1B hypersonic raise, Ursa Major's SPAC, and Blackbeard interceptor awards show the wave still building, but none of these are currently open watchlist slots and existing defense exposure (AVAV, AXON, KTOS, LMT) already spans the theme; no change. - **geo_drones**: AeroVironment's backlog and swarm-drone partnership continues its buildup story but AVAV is already held; no new names cleared bar. - **demographics**: Senior-housing REIT coverage (AHR's $114M portfolio buy, CTRE's raised guidance) confirms a slow steady wave but no held ticker or open catalyst justified touching this bucket this cycle. - **verdict**: One swap only -- RGTI out (revenue-story lagging its quantum peers), FRO in (the cleanest, most concrete fresh catalyst in the whole pool, matching the profile's explicit tactical-tanker thesis) -- keeping the watchlist's thematic shape intact while trading a speculative pre-revenue name for a proven-cycle geopolitical name with a live catalyst.

Rationale: - **AI**: NVDA earnings beat, AMD rack-scale expansion, Micron surge -- all already-run mega-caps or already-held names, nothing early-buildup and un-owned; noise (Apple CEO change, chip-export loophole) filtered. - **quantum**: Pasqal Nasdaq debut, IonQ Superion 256 launch and investor day, MIT qubit design, IBM/Lockheed Swiss hub -- genuine acceleration but IONQ/QUBT-type names are flagged in the same pool as most-likely-to-fall-first in a September selloff (high multiples, no fresh catalyst beyond hype); not a clean early-buildup entry and watchlist is full, so passed rather than force a swap. - **nuclear**: Army $2.2B microreactor awards to five firms including BWXT (already held), NANO/Enveniam MOU, uranium at $95/lb, SMR/OKLO 'dead-cat' rotation bounce explicitly called out as 'no fresh catalyst' -- reinforces existing BWXT/OKLO/CCJ/VST holdings rather than surfacing a new name. - **robotics**: Figure-Nscale $3.5B GPU deal, Unitree STAR listing, Goldman's 6.5M humanoid forecast, Teradyne +229% -- Figure and Unitree are private/foreign, Teradyne has already re-rated hard (229% run), not an early-buildup entry. - **rockets_spacecraft**: SpaceX/Starship test 14 buzz (private, no ticker), Chinese STAR-market space IPOs (not US-listed) -- both filtered as non-investable; RKLB already held and unremarkable this window. - **geo_defense**: Patriot/THAAD 7-year framework deals (LMT, General Dynamics), Boeing F-15 $131B ceiling, Northrop $508M award, L3Harris PAC-3 propulsion win -- all reinforce already-held LMT; no new name clears the bar over existing holdings. - **geo_drones**: AeroVironment (AVAV, already held) +6% then +5% on laser/LOCUST awards; KTOS down 37% YTD despite awards -- read as thesis-intact but not-yet-inflecting, not a remove trigger; Red Cat/Ondas/Unusual Machines are smaller-cap read-throughs, not clearing my conviction bar to displace a held name. - **geo_tankers**: Supertanker rates to $800k/day, freight fund up 3,600%, FRO's own $659M record quarter and $2.61 dividend cited directly -- strong confirmation the already-held FRO thesis (tactical, reserves-driven) remains intact and undiminished; this is a hold signal, not a new add (FRO is already on the list, so no action needed here despite the validator flagging it as a duplicate last round). - **geo_reconstruction**: no material coverage this window. - **demographics**: AHR/Morgan Stanley senior-housing acquisitions continue but no liquid new eldercare ticker surfaced beyond what's already excluded from the list. - **verdict**: every wave the pool touched either reinforced an existing holding (BWXT, OKLO, CCJ, VST, FRO, AVAV, LMT, KTOS) or surfaced only non-investable/already-run/over-owned names (SpaceX, Unitree, Figure, IONQ-adjacent quantum peak-risk, Teradyne post-run robotics); with the watchlist full at 16/16, no candidate cleared the bar of displacing a current holding, so no_changes preserves the existing thematic spread across AI, nuclear, quantum, rockets, geo_defense, geo_drones, and geo_tankers.

Rationale: - **AI**: AMD topped $1T on record chip-export demand and NVDA/AVGO/MU all ran on AI-infrastructure optimism, but this is late-cycle confirmation of an already-crested rally, not a new buildup signal; also filtered out one 'doomsday trade' dip (chip stocks sold off on AI-safety commentary) as noise -- kept AMD, GOOGL, NVDA unchanged. - **nuclear**: Oklo, NuScale, Uranium Energy bounced 3-5% on 'no fresh company news,' and simplywall.st listicles kept recycling SMR/HALEU boilerplate; OKLO and CCJ stay as-is, no action warranted since the bounce lacked a name-specific catalyst. - **quantum**: IonQ's real-time quantum error-correction decoder is a genuine, fresh, dated technical milestone (industry-first, NVIDIA research-center tie-up, >10% pop) squarely in the early-surge phase the profile wants -- added IONQ, funded by removing the weakest current holding. - **rockets_spacecraft**: Starship's slip to Sept. 28 and SpaceX's Falcon-9 rideshare wind-down are shifting backlog to RKLB (record ~$2.36B backlog, Neutron bookings at $50-55M) which is already held; filtered SPCX (private, no ticker) and VOYG/LUNR/ASTS (no current position, thesis not yet strong enough to displace a holding) as noise for this rebalance -- kept RKLB, no new add here. - **robotics**: Multiple market-sizing press releases (humanoid robots to 2036, Goldman's 6.5M-unit forecast) are generic TAM stories with no US-listed pure-play catalyst (Tesla, Unitree, Figure, Boston Dynamics are private/foreign-heavy) -- filtered as noise, no watchlist name available to act on. - **geo_defense**: Lockheed's JATM and PrSM contract wins and the Navy's Blackbeard hypersonic award are real but incremental for an already-held prime (LMT) and a non-watchlist micro-cap (Castelion); kept LMT unchanged, no new name met the bar. - **geo_drones**: Ondas' fuze acquisition and loitering-munitions market growth are noted but involve no watchlist ticker directly; KTOS and AVAV held unchanged pending a name-specific catalyst. - **geo_tankers**: FRO-relevant tanker rates hit $1M/day on Hormuz strain, but Business Standard and other reports flagged fresh US-Iran reopening talks unwinding the rerouting premium -- exactly the crest-trim signal the thesis warns about; held FRO unchanged rather than adding more exposure into a peaking trade. - **geo_reconstruction**: Gaza/Lebanon reconstruction financing headlines ($2.45B Board of Peace plan, $5B Partnership for Allied Construction) remain government-financing announcements with no clean US-listed engineering/construction ticker surfaced -- filtered as noise, no watchlist candidate. - **demographics**: Senior-housing REIT strength (Welltower, Ventas, CareTrust NOI growth) is real but these are non-watchlist REITs with no name currently held or catalytic enough to displace an existing position -- filtered as background confirmation, no action. - **verdict**: One clean swap this cycle -- IONQ in on a dated, cite-able quantum breakthrough that fits the 'early buildup, fresh catalyst' mandate, IRDM out as the holding with the least recent evidence behind it. Net effect keeps the watchlist's thematic spread intact (AI, nuclear, rockets, defense, tankers, quantum) while sharpening quantum exposure right as that wave shows its first hard technical proof point, without chasing the AI/nuclear names that already ran.

Rationale: - **rockets_spacecraft**: Starship reached orbit (Flight 14, Sept 28) and SpaceX followed with a string of launches (ISS crew, Google AI satellite, military rideshare, FCC clearance for 15,000 more Starlink birds), but SpaceX itself is private -- the only ticker in the headlines is SPCX, which multiple sources flag as a speculative/unofficial proxy with no confirmed listing I can verify as the real SpaceX equity; discarded as noise rather than risk a bad add. RKLB (already held) got the real investable catalyst: a record Electron deal with Synspective pushing its backlog past 100 missions, plus a 5-7% pop on a separate rideshare win -- reinforces the existing position, no new name needed. - **nuclear**: Real catalysts this window: Google-Constellation 20-year, 890MW PPA (lifted OKLO/NANO/SMR), TVA's NRC construction permit, BWXT's BANR microreactor selected for a Canadian transportable plant, General Matter's enrichment license filing. BWXT is already on the watchlist and got its own concrete win (BANR/Canada), OKLO already captured the Google-driven rally. UBS downgrade and dilution/legal-probe stories on SMR, plus a bearish 'uranium at 19-year high but nuclear stocks falling anyway' piece, confirm I'm right not to chase SMR or UEC -- noise/churn, not a clean buildup signal distinct from what BWXT/OKLO/CCJ already cover. - **quantum**: Heavy volume of news (IonQ DARPA Stage C advance, BofA/analyst upgrades, Nvidia's $1B quantum research commitment, IBM/Atom/Diraq also reaching DARPA Stage C) but it's almost entirely about IONQ, already held, and IONQ is simultaneously down ~45% from its May high per one dated retrospective -- a volatile, already-owned name, not a new entrant. No emerging non-IONQ quantum ticker cleared a bar of a concrete fresh catalyst plus investable liquidity (QBTS, RGTI, QUBT mentioned only in generic 'top stocks to watch' listicles, which is template content, not news). - **robotics**: Humanoid shipment growth (+432%) is a China-dominated, pre-revenue story (95% China market share per one piece) with no clean US-listed pure play surfacing; Boston Dynamics shelved its listing and China is tightening humanoid IPO rules -- cooling, not buildup. No AVAV/AXON-caliber name with a fresh catalyst emerged here. - **AI**: AMD, NVDA, GOOGL (all held) dominate coverage (Nvidia's $1B quantum pledge, AMD's Foxconn/TSMC supply push, chip-stock selloff on OpenAI revenue jitters and AI-bubble warnings). This is maturation/volatility in an already-owned, already-dominant wave -- exactly the plateau phase where the optimizer should be left to size existing names, not where I add. - **geo_defense**: Boeing's $20B F/A-XX win and $14.7B PAC-3 seeker deal, Raytheon's $20.7B AMRAAM and $24.4B SM-6 and $6.3B SM-3 awards, Lockheed's (held) $871M F-35 mod and new PAC-3 Edge interceptor, Army's $7B counter-drone IDIQ -- a genuinely strong multi-year buildup, but all of it accrues to primes already on the watchlist (LMT) or to names (BA, RTX, NOC) that are large, already-re-rated primes rather than early-buildup vehicles, consistent with the profile's caution that defense primes have already run. - **geo_drones**: AVAV and KTOS (both held) both named directly in the Pentagon's new Autonomous Warfare Command initiative; Ondas got a fresh $56M ESAD contract and $270M in cumulative orders, but shares fell despite the news ('orders pile up without lifting shares') -- a thesis-confirming but not thesis-changing data point for a name I don't hold, and ARK's small KTOS/AVAV buys just reinforce existing exposure rather than argue for churn. - **geo_tankers**: FRO (held) is squarely in the money -- VLCC/suezmax rates hit fresh records ($1.3-1.4M/day), near-daily Hormuz strikes continue in October, and the profile's own reserves-driven thesis for staying tactically long is playing out in real time. No reason to trim; this wave is mid-buildup, not crested. - **geo_reconstruction**: Thin this window -- a Gaza recovery plan unveiled but stalled by disarmament/withdrawal disputes, Syria/Gulf construction-expo mentions only; no investable E&C or building-materials ticker surfaced with a concrete catalyst. - **demographics**: Senior-housing REIT deal volume is heavy (CareTrust's $1.4B UK buy, Omega's Florida SHOP transition, NHP, LTC, American Healthcare REIT) but none of these names fit the technology/early-wave profile this curator targets, and LLY (held) already covers healthcare exposure to aging demographics; no action warranted. - **verdict**: Nothing in this pool clears the bar of a fresh, concrete, non-noise catalyst for a ticker NOT already on the 16/16-full watchlist strongly enough to displace a current holding -- the real catalysts (Starship orbit, Google's nuclear PPA, DARPA quantum Stage C, F/A-XX win, record tanker rates) all accrue to names I already hold (RKLB, OKLO/BWXT/CCJ, IONQ, LMT/KTOS/AVAV, FRO) or to SPCX/BA/RTX/UEC/SMR names that are either unverifiable-as-listed, already-crested primes, or carrying fresh negative signals (dilution, legal probes, declines despite orders). Holding the line with no_changes this cycle.

Rationale: - **AI**: Chip stocks whipsawed hard (SOXX down 8 straight sessions, then bounced on NVDA/Intel dip-buying) on financing-risk headlines (Anthropic debt terms, CoreWeave 13% coupons, Meta capex-driven FCF collapse) -- noise for position-level action since AMD and NVDA are already held and this is a trade/valuation story, not a new-name discovery. No fresh non-mega-cap AI name cleared the bar. - **nuclear**: Real catalysts kept landing (Amazon/AEP/X-Energy 12-SMR Indiana deal, Westinghouse $1B SC fuel plant, NANO-IMPSA MMR agreement, Radiant microreactor picked by US Navy) but also sharp pain in pure-play developers (NuScale down to 52-week lows on fraud probe/downgrades, OKLO -5-7% repeatedly) -- the wave is real but the two tickers with earliest, most financed catalysts (X-Energy via AEP/Amazon PPA, Radiant via Navy) are either not yet investable at scale or too illiquid; CCJ and OKLO/BWXT/VST already give the book exposure to both the fuel and utility/developer legs, so no swap-worthy name surfaced. - **rockets_spacecraft**: RKLB had a genuinely bad two weeks -- Neutron delayed again to 2027, stock down ~13-15% intraday on the news, now 63% off its 52-week high -- a real thesis dent, but RKLB retains HASTE hypersonic contracts, NASA moon-race tailwinds, and a pending Iridium deal, so this reads as a stumble within the wave rather than a broken thesis; SpaceX (ticker SPCX) is the standout buildup name this cycle (161% revenue growth beat, AI/Starlink surge, analyst double-upgrade to Buy) but it is not in the current watchlist and would need a remove to admit -- not pulling the trigger on a volatile single print. - **robotics**: Heavy flow of humanoid catalysts (Agility/AGCO Kansas pilot, Figure, Samsung CES 2027 humanoid, NVIDIA Isaac Agile ONE reference humanoid, Eka Robotics $1.45B Series A) plus a bearish cross-current (IDC: US robotics revenue could fall 58% below baseline by 2030 on FCC China-robot restrictions). No single US-listed pure-play with a liquid line and a dated catalyst rose above the noise of startups, Chinese names (Unitree, Jodell), and components suppliers (Peraso, Nyobolt) -- nothing concrete enough to swap in. - **quantum**: IonQ's $2B equity raise plus Elisity/SkyWater deals, Google's 40x error cut, and Quantinuum's 10x error-correction record are all strong buildup signals, but IONQ is already held and insiders across the pure-play group (IONQ, RGTI, QBTS, QUBT) have sold >$1B in stock over five years per one report -- a dilution/insider-selling overhang that argues against adding a second quantum name on top of IONQ right now. - **geo_tankers**: VLCC rates pulled back from record highs (Oman-China index down 13% from peak) even as attacks continue and Goldman flags freight at $20/bbl vs $3/bbl a year ago -- this is the 'normalize as reserves ease' phase the thesis anticipated; FRO already captures this and the move is toward trimming, not adding. - **geo_defense**: Broad-based strength continues (LMT raised 2026 guidance, record $292B backlog, 850-jet F-35 framework; RTX/Northrop also raised profit outlooks; $729M Raytheon counter-drone export wins) -- this is a maturing, already-re-rated wave per the profile's own caveat; LMT and KTOS already held, no new name needed, let the optimizer's weighting do the work. - **geo_drones**: AUSA 2026 dominated by counter-drone and loitering-munition news (Anduril Steelfin, Raytheon Coyote/KuRFS, Epirus, STM Kargu) but nearly all the interesting names are private (Anduril, Epirus) or illiquid micro-caps (Mobilicom, Powerus/PUSA) -- AVAV and KTOS already anchor this bucket; nothing investable cleared the liquidity bar. - **geo_reconstruction**: Syria-World Bank START fund and Gaza rebuild tenders are still at the government-announcement stage with no clean US-listed vehicle named in the pool -- too early for a position. - **demographics**: Senior-housing M&A hit record pace (LTC, Brookdale, Welltower all active) but these are steady-state REIT consolidation stories, not an early-wave catalyst, and none are in the current watchlist or clearly superior to what is held -- no action. - **verdict**: A genuinely newsy two weeks (Neutron delay, Amazon's SMR bet, IonQ's raise, the AI chip-stock rout and bounce) but nothing cleared the bar of a concrete, dated catalyst in a name clearly stronger than the weakest current holding, and the watchlist is full so every add needs a decisive swap; holding the current 16 names unchanged and letting the optimizer's trailing-return weighting handle the RKLB stumble and the AI-selloff rotation internally.