The Index
400 dossiers
A research screener for every company we cover. Search a name or ticker, then sort and triage dossiers by coverage freshness, our conviction and trading relevance.
agtech12 names· avg TR 2 | ||||||||
| A $2B SPAC zero — the tech was real, the business model never was; the only live residual is the IP estate now inside privately-held Confluence Genetics, not a tradeable Benson Hill. | — | 1 | ||||||
| The clear commercial leader in laser weeding with a real data flywheel and a fresh CFO-led pre-IPO posture — but penned into the narrow, cyclical specialty-crop niche with a ~$370M-marked-up-to-~$550M valuation that already prices the broadacre dream it has not yet earned. WATCHING, not buying, until the AutoTractor RaaS line and a credible row-crop path prove the TAM is bigger than vegetables. | — | 1 | ||||||
| The category-defining digital-ag platform is now a non-core software unit inside a balance-sheet-stressed Bayer that gives it away to sell chemicals — great product, captive owner, no clean way to own it. | — | 1 | ||||||
| A wide-moat seed/chem duopolist priced for steady compounding, sitting on a triple inflection — royalty flip to net out-licensor, the Bayer settlement unlocking corn/cotton licensing, and a 2H26 split that hands the clean-growth Seed business to SpinCo and dumps every legacy liability on Crop-Protection-Co; own the seeds, scrutinize the chem. | — | 1 | ||||||
| Best-in-class ag-equipment compounder at the bottom of a brutal large-ag cycle, already re-rated to ~32x on a trough-earnings recovery — the moat and the precision-ag option are real, but the price has front-run the recovery, so this is a high-quality WATCH, not a fresh entry. | — | 1 | ||||||
| A real ag-distribution disruptor that just admitted the hard part — the Oct-2025 GCS spin-off makes FBN an asset-light marketplace whose thesis now lives or dies on take-rate and lending, not on owning crop-protection inventory; an eventual IPO is a story, not a date, and the entry round you'd actually clear is a flat-to-down one. | — | 1 | ||||||
| A George Church / Flagship gene-editing platform whose entire "freedom-to-operate" moat — edit a rival's already-GM trait and call it your own — is precisely what Corteva is suing to invalidate; a Jan-2026 ~33% site layoff one year after a "robust" $144M raise says the commercialization clock is running faster than the cash. | — | 1 | ||||||
| The cautionary tale of the agtech bubble — a $3.5B "biology-will-rewire-farming" thesis that survived only by shrinking into a soil-carbon broker; the equity is impaired by ~94% and the surviving business is a thin-margin MRV middleman in a market with an additionality problem. WATCHING, not buyable — and not investable until there is a real revenue/margin disclosure, not a credit-tonnage press release. | — | 1 | ||||||
| A counter-cyclical compounder mid-re-rate — the destocking trough is behind it and the FY26 earnings snapback is real, but at ~¥2,900 / +81% off the lows the easy money is made; tariffs and a structurally weak NA ag market cap the upside from here. | — | 1 | ||||||
| A dead company — Zimeno/Monarch shipped 500 real tractors but never a real business; collapsed into an assignment-for-creditors and an IP-only Caterpillar acqui-hire. Uninvestable; value here is the post-mortem, not a position. | — | 0 | ||||||
| World's #1 potash producer + #1 ag-retailer, mid-cycle and cheap (~11x fwd P/E, 1.8x levered), self-help via portfolio simplification — a high-quality cyclical you buy near trough, not a compounder; BULLISH-tactical, not structural. | — | 8 | ||||||
| The category-defining gene-edited nitrogen-fixation company with real revenue (>$100M, 2023) and a tier-1 syndicate — but the product only displaces 15-30% of synthetic N, independent university trials largely fail to find a yield benefit, and it has not raised primary capital since its July-2021 $430M / $2B Series D; WATCHING a private name living off a five-year-old round whose terminal value hinges on an unresolved efficacy question. | — | 8 | ||||||
ai7 names· avg TR 2 | ||||||||
| The best business in AI on every metric that isn't valuation — enterprise-led, multi-cloud, ~$30B+ run-rate doubling on weeks — wrapped in a ~$965B mark that already prices a decade of flawless execution and rests on a circular ring of vendor-investors funding ~$180B of compute the company hasn't yet earned the margins to pay for. WATCHING, bull on the business / bear on the entry price. | — | |||||||
ai-bio8 names· avg TR 7 | ||||||||
| A 20-year, ~$1.7B-deficit cautionary tale of synbio that hit the same wall every fermentation pure-play hits (scale-up economics never closed the gap to chemistry), capped by an SEC revenue-recognition charge and a Chapter 11 that wiped equity to zero — the emerged "Amyris 2.0" is a smaller, Doerr-controlled private with no public equity to buy and no edge proven; AVOID / un-investable, watch only as the canonical synbio-failure comp. | — | |||||||
bci9 names· avg TR 6 | ||||||||
| The softest neural interface on the planet with zero chronic human data — a genuine materials-science lead wrapped around an unproven clinical thesis and a behind-the-pack regulatory position; a WATCH, not a bet, until a Fleuron probe records cleanly past 90 days in a living brain. | — | 1 | ||||||
bci-neuroscience6 names· avg TR 1 | ||||||||
| The most-implanted BCI on Earth and the worst-capitalized relative to its peers — a 2008-era electrode bet now owned by a stablecoin issuer, where the moat is two decades of human data and the kill-switch is glial scar tissue. Watching, not buying, until a De Novo/PMA submission and a clean secondary mark say the Tether money actually bought commercialization rather than just runway. | — | |||||||
biopharma12 names· avg TR 3 | ||||||||
| The best-funded biotech ever, reorganizing toward AI after four years and zero disclosed clinical assets — a platform bet whose lead rivals just beat it into the clinic; WATCHING, not ownable, until a named program and a runway figure exist. | — | 1 | ||||||
critical-materials14 names· avg TR 6 | ||||||||
| The only Western "mine-to-magnet" pure-play levered specifically to HEAVY rare earths (Dy/Tb) — whose ex-China prices have ~doubled in 2026 (Dy ~US$930/kg, Tb ~US$4,028/kg) — via China's own low-cost ionic-clay geology; but it is a pre-revenue developer that has already re-rated ~5x to a ~CA$1.1B cap, owns a Chilean asset 99% of the local community voted against, faces MP Materials commissioning its own US HREE separation by mid-2026, and must fund a >US$1.5B integrated build off a ~US$40M cash balance with no binding offtake. Asymmetric optionality, not yet a position. | — | |||||||
crypto2 names· avg TR 1 | ||||||||
| A Bitmain/Antalpha-controlled reverse-merger shell wearing a Bitcoin-miner costume and now an AI-compute costume — high-cost fleet, ~50% related-party voting control, a NYSE sub-$1 delisting clock, and a balance sheet half-built on Antalpha loans; avoid the equity, watch only as a Bitmain-proxy / forced-restructuring special situation. | — | |||||||
electrification8 names· avg TR 1 | ||||||||
| ALB is a high-quality, low-cost lithium asset wearing the price tag of a leveraged call option on the lithium spot price — own the resources, not the P&L; the stock is the commodity in a cyclical costume. | — | 1 | ||||||
energy1 name· avg TR 1 | ||||||||
| The best nuclear asset base in America, now bolted to a gas fleet and priced as an AI-power call — at ~26x forward the upside is no longer in the price for free, and the Dec-2025 FERC co-location order is the binary the bulls are underweighting. | — | 1 | ||||||
genomics2 names· avg TR 1 | ||||||||
| The base-editing platform with the cleanest in-vivo clinical proof-of-concept yet (BEAM-302 corrected a disease-causing mutation in living patients) and a near-term commercial shot (risto-cel BLA by YE26) — a genuine BULLISH-but-binary genomics name where the science de-risked faster than the commercial model. | — | |||||||
hardware6 names· avg TR 11 | ||||||||
| The most dangerous competitor bulls underestimate — the named beneficiary of NVIDIA's own inference-share vulnerability, growing AI +143% — but priced at a PREMIUM to NVIDIA (45× vs 30× EV/EBITDA) on one-third the ROE, with lower earnings quality (acquisition-amortization add-backs, $64B debt) and the same hyperscaler concentration. The most expensive, lowest-quality way to play the custom-silicon thesis. NEUTRAL / WATCHING / MEDIUM. | WatchingMed | |||||||
optical-computing17 names· avg TR 3 | ||||||||
| A genuine top-4 global optical-component IDM riding the AI interconnect wave — but it is the low-margin generalist (~8% net) trading at a specialist multiple after a 5x run, with consensus targets implying ~55% downside. Real franchise, priced for a margin structure it has never earned. | — | |||||||
robotics11 names· avg TR 1 | ||||||||
| A genuine engineering bet (tendon-driven safe humanoid + in-home data flywheel) wrapped in a marketing fiction — NEO ships "autonomous" but runs on stranger-in-a-VR-headset teleop, and the company most likely to make it work, OpenAI, just became its competitor. | — | 1 | ||||||
space7 names· avg TR 1 | ||||||||
| A real #2 to SpaceX with anchor demand and a flight-proven heavy lifter — but a single-pad, single-stage company that just lost its only launch site and its only customer's confidence in the same month, forced into its first outside raise at the exact moment SpaceX vacuums up the capital. | — | 1 | ||||||
| First public print HARDENS the bear-leaning thesis — the OpenAI "diversification" is now a quantified circular liability ($1B customer loan on the balance sheet + warrants) and the Q2 gross-margin guide-down to 36–38% (from 47%) plus FY26 op-margin of −28/−32% prices the OpenAI ramp as expensive-before-profitable. FY26 revenue guide $855–865M comes in BELOW my $1.0B forecast. WATCHING, bear-leaning on valuation + margin trajectory. | — | 8 |
| The lab that broke the cost-of-intelligence narrative and now anchors China's sovereign AI stack — a frontier-class open-weight model trailing the closed leaders by 3-6 months at ~1/50th the unit cost, but state-captured, un-investable to US capital, and monetizing almost none of the value it creates. | — | 1 |
G | The crown jewels and the brain already left for Nvidia for ~$17–20B; what remains is a cash-rich but architecturally-cornered inference cloud whose flagship Saudi demand evaporated — a melting-ice-cube neocloud, not the LPU disruptor the headlines sold. | — | 1 |
| Europe's only credible frontier lab — a sovereignty-and-infrastructure bet, not a model-quality bet; the moat is regulatory/political, the risk is that "good enough, in Europe, on-prem" gets out-priced by Chinese open weights before the €20B mark is earned. | — | 1 |
| The revenue leader that turned into the cash-burn leader — a $1.4T compute book financed against a P&L losing ~$14B a year, where the only question that matters is whether the IPO window opens before the funding treadmill stalls. | — | 1 |
X | A frontier lab fused into Musk's compute-and-distribution empire — Grok is a fast-following #4 model bankrolled by a $30B/yr capex furnace; the real instrument is now SpaceX equity at a $1.25T mark, not a standalone xAI bet. | — | 1 |
| The credible enzymatic-DNA-synthesis survivor — a real fidelity moat (1,005-base record, 50 kb clonal, ~99.9% stepwise yield) now distributed through Danaher/IDT — but it is a sub-$25M-revenue tools shop selling a faster picks-and-shovels commodity into a brutal synbio funding winter; WATCHING as a private until an IPO path or an IDT buyout crystallizes the value. | — | 8 |
| The only AI-bio company building a proprietary, legally-clean training dataset 100x deeper than UniProt — own the data layer of TechBio and it is a generational compounder; if foundation labs (ESM, AlphaFold) reach "good enough" on public data first, it is a beautifully-sourced commodity. Watch for an EDEN-derived therapeutic deal with real economics and a priced Series C. | — | 8 |
| "The platform pioneered AI drug discovery and was real; the public-market vehicle and the unit economics were not — a £1.5bn SPAC that destroyed ~96% of value and was taken private by its founders at a ~€40-57m carcass. UNINVESTABLE as a public name (delisted); WATCH only as a private/strategic-acquisition tell for the ai-bio thesis. The lesson, not the ticker, is the asset." | — | 8 |
| A real lab-in-the-loop antibody-engineering shop with blue-chip pharma validation (AbbVie/Lilly/Merck/Amgen) and now its own clinic-bound ADC — but it is dramatically out-capitalized ($145M vs. $1B+ peers) and is becoming a Lilly satellite; WATCHING as a private, with the IND-into-clinic 2026 prints as the only catalyst that re-rates it. | — | 8 |
| The single most exciting private name in AI-bio — Chai-2/3 is the first AI antibody platform with a wet-lab-validated, repeatedly-improving hit rate AND signed pharma revenue (Lilly + Pfizer); the bet is whether a portable, lab-light software platform can defend value capture against Isomorphic's deeper engine and the platform-vs-asset value trap. WATCHING for an IPO that is 2-3 years away. | — | 8 |
| A de-risked computational-discovery option on AstraZeneca's rilvegostomig — the last surviving TIGIT bet — wrapped around a $135M cash shell and a binary 2027 ovarian readout; you are buying the royalty stream and the platform, not COM701. | — | 1 |
| The cleanest pure-SaaS bet in AI protein engineering — Google-pedigreed, wet-lab-backed, already billing 6 of the top-25 pharma — but it has deliberately disarmed itself of the one thing that creates equity value in this field (it owns no molecule, takes no royalty, signs no IP), so it is selling a fast-commoditizing copilot whose best customers (Novo Nordisk) are simultaneously funding the open-source models that erode its moat. | — | 8 |
| The only at-scale, FDA-cleared, revenue-bearing non-invasive BCI franchise heading to a public listing — but it is a Chinese military-civil-fusion controversy with a confidential HK IPO, no public financials, and a US national-security cloud, so it is un-investable on a US/UK book until it trades and discloses. |
| — |
| 1 |
| A non-invasive, AR-first AAC headset with a real FDA-Breakthrough wedge and two distribution partners — but at a ~$25–46M valuation after 12 years and ~$25–32M raised, it is a mission-driven research-tools / assistive-device niche player, not a venture-scale BCI bet; the asymmetry is acquisition-by-a-giant, not an IPO. | — | 0 |
| The only European bidirectional-BCI in US human trials, backed by the BioNTech family office (Santo/Strüngmann) — a Breakthrough-Device, stroke-rehab-first asset that is structurally early and structurally under-capitalised; WATCHING for the funding round that proves the Strüngmanns will write a BioNTech-scale cheque, not just an option. | — | 8 |
| The most-cited consumer-EEG brand in research, now repricing itself as "cognitive AI infrastructure" — but 23 years in, ~$10M ever raised, and a Chilean supreme-court delete order make it the BCI category's pioneer that capital forgot, not its winner. WATCHING — a be-early IPO-readiness name, not yet investable. | — | 8 |
| Forest is a nonprofit research instrument, NOT an equity — the tradeable BCI bet it incubated walked out the door as Merge Labs (Altman/OpenAI, $850M); track Forest as the science-de-risking front-end, position via Merge or Butterfly (BFLY), never Forest itself. | — | 8 |
| The profitable, un-hyped grandfather of non-invasive BCI — a 27-year Austrian bootstrap that quietly owns the research-EEG install base and is turning recoveriX neurorehab into a franchise annuity; a real cash-flowing business but NOT a tradeable IPO candidate (no venture capital, ~$24M est. value, founder-owned), so it is a WATCH-the-category tell, not a position. | — | 8 |
| The best-positioned European BCI pure-play and the only graphene-native one — but it is a Series-B, pre-revenue spin-off two full funding rounds behind Neuralink/Precision, and its material moat is unproven precisely where its closest competitor (Precision's Layer 7) already has FDA clearance; a watch-and-wait private, not yet investable, gated on a Series C + a graphene-specific regulatory de-risking event. | — | 8 |
| A genuinely good optical-imaging instrument company wearing three failed business models — capital-starved ($158M total, no priced round since Dec-2023) and IPO-irrelevant while the rest of BCI raises billions; WATCHING, not investable, until a real revenue line or a strategic acquirer appears. | — | 8 |
| A blueberry-sized, battery-free epidural stimulator with a Medtronic-grade team and a clean FDA IDE — but it is a single-asset, sub-$30M minnow in a field of nine-figure whales, and its closest peer (Inner Cosmos) already has the clinical lead. | — | 1 |
| The category's brand and bandwidth leader, but the bet is binary on durable signal-at-scale and CMS reimbursement — not yet a tradeable security, and the moment a Synchron or Precision reaches a pivotal trial first, Neuralink's "lead" is narrative, not regulatory. | — | 1 |
| The highest-bandwidth intracortical BCI just put its wireless device in a human brain (17 Jun 2026) — but it is two patients into an EFS, ~3-5 years from revenue, and out-capitalized ~3-4:1 by Neuralink, so the bet is "bandwidth wins the speech-prosthesis category" against a far better-funded field. | — | 1 |
| The only FDA-cleared, commercially-shipping cortical BCI — but it is selling a 30-day surgical-monitoring tool, not the chronic implant the $500M valuation is priced on; Medtronic is the real tell, IP overhang from Rapoport's Neuralink past is the real risk. | — | 1 |
| The safest, most surgically scalable path into the brain and the only BCI Apple made a native input — but it bet the company on "good enough" 16-electrode bandwidth, and its own third-gen "whole-brain" pivot is a confession that the moat it is famous for may be the ceiling it has to escape. | — | 1 |
| Alphabet's $2.5B/12-year longevity moonshot just lost its only pharma partner — AbbVie walked after the lead asset flopped — and the surviving thesis is now a 3-asset rare-disease biotech (ADPKD antibody the crown jewel) wearing an aging-research halo; not a 2026 tradeable, watch the ADPKD Phase 2 readout as the sole de-risking event. |
| — |
| 1 |
| The best operating asset in pharma trading at a price that already imputes a $150B+ obesity TAM — own it on the franchise and the orforglipron/retatrutide optionality, but size for a multiple that compresses from ~30x forward as the price war and TAM-skeptic camp grind on the narrative. | — | 1 |
G | The single-partner-concentration bear point is dead — Gero now has TWO pharma validations, and the bigger one (Chugai/Roche, exclusive worldwide antibody rights, up to ~$250M milestones + royalties) is disclosed on a listed Roche subsidiary's own corporate filing, the most credible source in this dossier. Plus an open-sourced sequence-to-molecule model (ProtoBind-Diff). Still WATCHING (private, no public security, no named owned IND asset), but the watch is materially higher-quality — conviction on the platform-validation leg rises from "one partner, unfalsifiable" to "two top-tier pharma partners + a verifiable open-source artifact." The asymmetric event is now a partner advancing a target to the clinic or an institutional/crossover up-round. | — | 0 |
| The first company to put epigenetic age-reversal into a human body — a binary, single-asset bet where a clean Phase 1 safety read on OSK reprogramming is worth more than the whole longevity field's $4B+ of pre-clinical promises, and a single inflammation or tumor signal ends it. | — | 0 |
| The first company poised to win an FDA label for "lifespan extension" in any species — a genuinely novel regulatory beachhead on a large, growing owner-pays pet-pharma market — but the bet rides on one not-yet-read-out 1,300-dog trial, a surrogate-style ("reasonable expectation") approval, and a single CMC section still outstanding; WATCHING (privately ownable conviction, no tradeable instrument) with the STAY lifespan readout as the binary. | — | 8 |
| Metsera is no longer an investable equity — Pfizer closed the buyout at $65.60 cash on 13-Nov-2025; the only live instrument is the non-transferable CVR (up to $20.65/sh on three obesity-approval milestones), realistically worth a fraction of par, and the durable trade is the read-through to PFE/NVO/LLY and the amylin thesis, not MTSR. | — | 0 |
| A platform-quality data moat (1,000× the field's reprogramming-screen data) wrapped around an unproven biology, priced at $3.1B pre-IND for a 2027 first-in-human — the bet is the engine, not the liver drug, and the engine won't be falsifiable for 24+ months while a rival is already in the clinic. | — | 8 |
| A fallen-angel GLP-1 franchise trading at a generic-pharma multiple on a still-growing branded book — but the de-rating is a verdict on durability (Lilly's tirzepatide is structurally better, semaglutide's US patent cliff is 2031–32), and the cheapness is only a buy if the oral-Wegovy ramp and the amycretin pipeline can defend share before exclusivity breaks. | — | 1 |
| A billionaire-seeded Swiss "hallmarks-of-aging" holdco that quietly went dark — rebranded to Centenara in Sep-2024, lead eye asset terminated, holdco comms silent ~21 months; the ONLY live value driver is a separately-financed Belgian portco (Rejuvenate Biomed RJx-01, Ph2 COPD-sarcopenia, topline end-2026). Un-investable as a private until an IPO path or a single de-risked asset emerges; WATCHING the RJx-01 readout only. | — | 8 |
| A genuine builder's longevity platform de-risking faster than its peers, but the $1.8B mark is a 64% haircut off the $5B it was chasing six months ago — own it only after the August RTR242 readout converts "no tox" into a real biomarker signal, not before. | — | 8 |
| A cash-gushing CF monopoly priced for a successful four-engine diversification (pain, renal, gene therapy, diabetes) that is, so far, only one-third proven — own the moat, underwrite the pipeline at a discount. | — | 1 |
| 1 |
| A binary supply-shock proxy — the only Western-aligned tungsten scale producer ramping into a 550%-China-driven price spike, but priced at ~150x trailing sales on a single-mine ramp; own the commodity thesis, not this multiple — a quarter of Sangdong slippage or one China de-escalation re-rates it 40%+. | — | 1 |
| The highest-grade tin mine on earth, throwing off >$600M annualised EBITDA at a 6.7x P/E — and it sits 180km from an active M23 front line in the eastern DRC, now 56%-owned by Abu Dhabi. The discount is the country, not the company; you are paid ~6.5% to wait, but a single security headline can halve it overnight. | — | 0 |
| A three-engine critical-materials conglomerate whose 2025–26 earnings are flattered by a transient antimony windfall while its real long-duration call options (European LiOH at Bitterfeld, spent-catalyst vanadium recycling, the Saudi Supercenter) are still pre-cash-flow — own the structural story, but underwrite the cycle, not the print. | — | 0 |
| A copper pure-play being born inside an iron-ore company — own the re-rate that completes when Anglo Teck closes, not the diversified miner that's disappearing. | — | 8 |
| A best-in-class, Luksic-controlled Chilean copper pure-play priced for the 30% volume ramp it has not yet delivered — own the asset, not this multiple; the entry is a copper-price dip or a Centinela-2 commissioning wobble, not 14x EV/EBITDA on a name that just missed guidance again. | — | 8 |
| A 38-year, $1.7B-NPV NdPr asset that the equity market is pricing as a serial dilution machine — the call is on financing-close + China's price floor, not the ore body. WATCHING into FID-completion; the project is fundable, the share count is the bear case. | — | 8 |
| A self-help margin re-rating priced as a doomed smelter — the $0 TC/RC tape masks that Aurubis already earns more from downstream premiums, recycling and a finished $1.7B capex cycle than from the concentrate spread the bears fixate on; watching, not yet a buy, because the metal-price tailwind that lifted FY25/26 guidance twice is the same lever that breaks on a copper pullback. | — | 8 |
| A best-in-class integrated Nordic miner-smelter whose precious-metal-rich, e-scrap-fed smelter chain is the rare structural winner of the negative-TC era — but at ~21x P/E and a price above the average analyst target, the stock already prices the gold-silver windfall, not the smelter-margin reset risk and post-fire/post-acquisition execution debt. | — | 8 |
| World's #2 nickel reserve at first-quartile cost with state backing — but a US$2.5bn funding gap on a ~C$400m shell means the entire bet is binary on the 2026 financing close, not the geology. | — | 8 |
| A leveraged, pure-play Americas copper call trading at a mid-tier P/NAV discount — own it for the deficit and the MV-Optimized/Santo Domingo growth stack, but the entry price is now a bet that record copper holds, not that the assets are cheap. | — | 8 |
| The world's largest rare-earth producer is a state policy instrument, not a shareholder vehicle — at P/E ~66 / ~$25B cap on ~7% ROE you are paying a growth multiple for a margin that its own parent (Baogang) taxes away one concentrate hike at a time. Own the NdPr theme through a price-taker, not this price-maker-that-isn't. | — | 8 |
| World's #1 cobalt + a top-tier DRC copper grower trading at a deep state-owned/jurisdiction discount (~4-9x P/E vs 12-22x peers) — the value gap is the thesis; the DRC quota regime and Beijing's hand on the till are the reasons it persists. | — | 8 |
| A genuinely strategic Western heavy-rare-earth asset wrapped in a serial-promoter, pre-revenue, dilution-machine equity — the deposit is real, the metallurgy and the 2029 cash-flow date are not yet, and the stock already prices a permitting+offtake fairy-tale at ~$1.5B with $0 revenue; WATCHING, would only own on a proven-metallurgy or fully-funded BFS catalyst. | — | 8 |
| A levered, sub-NAV Ethereum proxy run by the best operators in the trade — own the discount-closing buyback, not the "treasury premium" that already died; the bet is ETH plus a 0.83x→1.0x mNAV mean-reversion, fighting a 95% drawdown and a structurally bigger rival. | — | 1 |
| The world's lowest-cost EV/battery manufacturer is now margin-bleeding in a brutal domestic price war it helped start; the entire bull case has migrated to a high-margin export ramp that tariffs and a Brazil/Hungary regulatory cloud are built to throttle — cheap on forward P/E only if you trust earnings that GMT says are flattered by ~¥300B of supply-chain-financed hidden debt. |
| — |
| 1 |
| The cheapest dominant company in the world — a 40%-share, 27%-ROE compounder priced at ~22x because it carries a Chinese-passport discount and a US-blacklist overhang the market refuses to underwrite away; bullish, but the re-rate needs a geopolitical thaw that may never come. | — | 1 |
| The cheapest electron-storage on the grid at the worst round-trip efficiency on the grid — a genuine multi-day monopoly whose moat (iron at $0.10/kWh, a US steel-mill factory, ArcelorMittal upstream) is only worth owning if regulators keep paying for firm capacity rather than cheap throughput; the Google $1B deal proves demand exists, the sub-40% RTE proves the bet is on policy and capacity value, not arbitrage. | — | 1 |
| A great battery operator strapped to a collapsing demand curve and a U.S. subsidy that is its only profit — the franchise is real, but the 2027-2032 order book just lost ~$9B of contracted Ford/FBPS volume and AMPC repeal is an existential tail risk; WATCHING, not buying, until the ESS pivot proves it can replace lost auto cells at a real margin. | — | 1 |
| A premium-cell incumbent being squeezed from below by Chinese LFP scale and above by its own Tesla over-concentration — but the unit now has a genuine second engine (data-center backup, ~80% share) that the parent's 46x-PER stock does not yet price as a battery turnaround. | — | 1 |
| A loss-making #9 EV-cell laggard re-rating on an ESS/data-center pivot the market is pricing as a turnaround — the call is whether US grid-storage demand and 2027 solid-state outrun Chinese LFP deflation before the balance sheet (rights issue + Display-stake sale) runs out of runway. | — | 1 |
| A structurally sub-scale #6 battery maker whose survival now rests on a US-policy bet that just inverted — the Ford anchor is gone, EV demand cratered, and the entire growth pivot (ESS/LFP) is a margin-thin race into China's home turf; only the SK Group balance sheet keeps it solvent, and there is no tradeable security to express a view on. | — | 1 |
| A real long-read/native-modification monopoly inside a structurally hard razor-and-blade model — but the market has stopped paying for "deep tech that takes 10 years"; down 73% from IPO, near 52-week lows, it is a show-me story where 2027 EBITDA breakeven (twice-pushed) is the entire thesis, and the new ex-Danaher CEO either professionalizes the commercial engine or the optionality stays un-priced. | — | 1 |
| 12 |
| The toll-booth on chip yield — a ~58%-share process-control near-monopoly whose revenue scales with chip *complexity*, not wafer volume, so it compounds even through capex air-pockets; the only real debate is the price you pay, not the quality of the business. | — | 1 |
| "NEUTRAL / WATCHING — the business genuinely transformed (clean books, $22B prepaid take-or-pay raising the floor, data-center 61% of revenue), but at ~$1,133 (~8x PEAK EPS, upside capped at peak-price SCA ceilings, ~60% of revenue spot-exposed to CXMT, EV/EBITDA premium to the HBM leader) the bull case is in the price. Quality confirmed; entry unattractive — own the dips, not the rip. Diverges from the prior web-only dossier's BULLISH on the ENTRY, not the business." | BullishMed | 18 |
| The cheapest way to own the memory supercycle — a $63B-net-cash colossus that just clawed back into Nvidia's HBM4 supply chain at ~7x forward earnings, but the market's discount is a verdict on memory cyclicality and Samsung's execution, not a mistake. Bullish on the re-rate; the kill-switch is a 2027 oversupply that turns peak earnings into the top tick. | — | 1 |
| Captures the LARGEST slice of the HBM scarcity rent (~62% share, ~2/3 of NVIDIA's HBM4, 72% op margin, ~61% ROE) yet is valued like a Korean memory-cyclical (~6–9× forward) — and management's own $14B US listing is an explicit re-rating bet. Best HBM rent-capture in the cluster. BULLISH / MEDIUM-HIGH / 1Y. | BullishMed | 12 |
T | The chokepoint that captures the foundry + CoWoS-packaging rent for the ENTIRE AI complex (NVIDIA, AMD, Broadcom, Apple all depend on it), trading at ~22× forward — BELOW every customer it supplies — with the only real risk exogenous (Taiwan geopolitics, the source of the discount). The cleanest, highest-quality, lowest-multiple way to own AI infrastructure. BULLISH / HIGH / 1Y+. | BullishHigh | 22 |
| The purest public-market-bound bet on optical I/O — strategically anchored by Nvidia + AMD + TSMC, but priced at $3.75B into a thesis (third-party CPO chiplets inside the GPU package) that its own anchor investors are simultaneously building in-house. | — | 1 |
| Path-to-tradeable already resolved — Marvell bought Celestial for $3.25B upfront (+$2.25B earnout) and closed Feb 2, 2026; the only way to own this thesis is MRVL, and the bet is whether photonic scale-up I/O hits a $1B run-rate by FY29 or stalls at a science-fair demo. | — | 1 |
C Coherent Corpcatalyst in 43d | An II-VI/Coherent merger debt-mule reborn as the picks-and-shovels supplier of the AI-optics supercycle — backed out to 2028, Nvidia-anchored, and de-risking the balance sheet — but priced at ~49x forward EPS that already embeds flawless CPO execution it has not yet proven. | — | 1 |
| A 20-year serial cash-burner with a real PLC fab and a going-concern stamp, re-skinned as a CPO/AI-optics pure-play — the FR8 order is genuine but the balance sheet, not the technology, decides this stock. | — | 1 |
| The cheapest, highest-margin way to own the 800G/1.6T AI-interconnect ramp — but it is a no-CIK Chinese A-share whose 47% margins, 96% offshore revenue and ~52x P/E all rest on a Broadcom-DSP / Lumentum-EML leash it doesn't own and a hyperscaler order book it can't diversify. | — | 1 |
| The best modulator in AI optics — but a modulator-only company in a market where the indium-phosphide laser, not the modulator, is the scarce asset; the $80M Series C buys it a foundry-scale shot at owning the TFLN layer before silicon-photonics incumbents fold thin-film into their vertically integrated stacks. | — | 1 |
| A near-death epiwafer monopolist just got bought out of bankruptcy by its own customer — own the InP photonics call, not the company; the AI-optical TAM is real but the float is a recovery option on a chronically capex-starved, margin-thin foundry whose two largest customers now sit on its board. | — | 1 |
| World's first listed pure-play photonic-computing stock — and a ~640x-sales SOE-backed China name with $15M revenue, a $185M annual loss and Huawei owning 98% of its market; the technology is real, the equity is a momentum trade priced for a future that is 5 years out. | — | 1 |
| The purest public-markets-bound bet on AI's copper wall — but it is racing NVIDIA, Broadcom and TSMC to commercial CPO from outside the package, and "2027 volume shipments" means the $4.4B mark is paid entirely on a roadmap. | — | 1 |
| A genuine top-tier Chinese optics merchant (top-3 passive, #9 global components, real InP/SiPh chip + 1.6T/CPO roadmap) trapped behind a 6-year private wrapper — the only tradeable expression today is the STAR-Market IPO it has been "about to file" since 2023; not investable until the S-1 (招股书) lands, and the coverage ticker 0877.HK is the WRONG company. | — | 8 |
| A genuinely differentiated all-optical scale-up fabric with a real CEO and a real AMD/ARIA anchor — but a $93.5M micro-cap fighting $1.2B–$4.4B peers and a hyperscaler-backed open-standard tide; the bet is a strategic acqui-buy (à la Marvell/Celestial), not a standalone IPO. | — | 8 |
| The best-capitalised, hardest-physics bet in quantum — $4B+ in private + sovereign money betting that silicon-photonic fusion-based computing leapfrogs straight to a million error-corrected qubits by ~2027, with the one fact that matters still unproven (a utility-scale machine running) and a freshly installed operator-CEO signalling the company is now an execution problem, not a science one. | — | 8 |
| The most credible analog-photonic-compute bet in Europe — a vertically-integrated TFLN coprocessor with a real fab, a first paying cloud customer (IONOS) and an ARM-founder-grade board — but at $80M raised it is a seed-stage David swinging at a $4B-valued field (Lightmatter) and a just-acquired peer (Celestial AI → Marvell, $5.5B); WATCHING for IPO-proximity, not yet tradeable. | — | 8 |
| The best-engineered way to lose to PsiQuantum — a genuine deterministic-single-photon moat and real machine sales, capitalized at ~1% of its photonic rivals; a sovereign-Europe call option, not a standalone winner. | — | 8 |
| A credible Oxford/Münster silicon-photonics team that pivoted from a losing optical-compute bet to the right one — all-optical circuit switching — and shipped a real 32-port product, but it is a ~£137M-valued, sub-$50M-raised minnow standing in a lane Nvidia, Lumentum and a $4.4B Lightmatter are all paving; WATCHING, not investable, until a named hyperscaler design win proves the silicon-photonic-OCS approach beats incumbent MEMS in production. | — | 8 |
| The picks-and-shovels king of the AI buildout — #1 in optical transceivers with >50% of Nvidia's wallet and the lowest forward P/E of any peer — but it is a margin-taker squeezed between US chip suppliers it cannot displace (Broadcom/Marvell DSP, Lumentum/Coherent EML) and Nvidia's own CPO/in-house ambitions, and the price already discounts flawless 1.6T execution. | — | 1 |
| The only humanoid actually billing paying customers for real warehouse work — but it is being squeezed between Figure's $39B AI-narrative capital and Unitree's $16K Chinese hardware, and SoftBank already tried to buy the whole thing for under $1B. |
| — |
| 1 |
| The credible #2 Western humanoid — NASA-pedigree hardware + Google DeepMind's brain + the Mercedes/GXO/Jabil go-to-market — but priced for a labor-replacement future it has not yet shipped, into a market where Unitree is collapsing the price floor from below. | — | 1 |
| The most physically capable humanoid in the world, wholly bankrolled by its own customer (Hyundai) — which is exactly why the standalone economics are unprovable until a third party pays full price in 2027. | — | 1 |
| The category's pace-setter and best-funded vertically-integrated humanoid bet — but the $39B mark prices a fleet that, on the public record, is still a pilot; WATCHING for the BMW-pilot-to-multi-customer-scale proof that converts the story into a tradeable S-1. | — | 1 |
| The strongest research bench in robot learning, priced like a product company while still being a research project — own the index (NVIDIA/data), not this single ticket, until a paying-fleet number or a π1 zero-shot result exists. | — | 1 |
| World-class robotic hands and the #3 humanoid patent estate on Earth, attached to a balance sheet that already had to sell its best asset (the Apptronik stake) to make payroll — this is an IP carcass walking, far likelier to be acqui-hired or stripped for patents than to IPO independently. | — | 1 |
| The leading robot-foundation-model pure-play by capital and strategic syndicate, but its $14B mark prices a software-layer monopoly that the Foxconn/Blackwell line has yet to prove and that Physical Intelligence + NVIDIA's own GR00T are both armed to contest. | — | 1 |
| A genuinely differentiated case-handling robotics moat wrapped around an ungovernable accounting-and-concentration core — adverse ICFR opinion + active SEC whistleblower probe + >90% Walmart revenue means the multiple is pricing a clean compounder that the filings say does not yet exist. | — | 1 |
| A profitable EV maker priced as a solved-autonomy robotics company — the car business is shrinking, the GAAP profit prop (reg credits) is going to zero, and the entire ~190x multiple now rents on robotaxi + Optimus execution that is real but years behind the price. | — | 1 |
| The only profitable humanoid maker and #1 by volume — but the float is a Shanghai-only embodied-AI bet whose revenue is 74% lab demos, not labor replacement, and whose Western TAM is being legislated to zero. | — | 1 |
| A real lunar/launch franchise wrapped in a serial-failure rocket, a money-furnace P&L, and a ~32x-sales price — the Golden Dome/SciTec defense pivot is the only thing that justifies the multiple, and Eclipse first-flight is the binary that settles it. |
| — |
| 1 |
| "The 'Maxar' instrument has dissolved — Lanteris (the satellite-builder) was sold to Intuitive Machines (LUNR) in Jan 2026, leaving the private residual = Vantor, an Advent carve-out repricing 30 cm imagery as defense-software ahead of a ~$14.5B exit. WATCHING the Vantor IPO/strategic-exit; the only tradeable Maxar today is LUNR's space-systems half." | — | 1 |
| No longer a rocket startup — an Eric Schmidt-controlled option on Terran R becoming the launch arm of an orbital-AI-datacenter thesis, gated entirely on an unflown late-2026 maiden flight and carrying a stale $4.2B (2021) mark that almost certainly no longer holds. | — | 1 |
| A genuinely great company and a genuinely terrible price — the only Western full-stack launch+satellite pure-play, compounding at ~50%, but trading at ~64x EV/sales with the entire Neutron thesis still un-flown. Own the business, fade the multiple. | — | 1 |
| A genuine launch-and-connectivity monopoly wrapped inside an unprofitable $2T+ aspiration stock — Starlink is the real business, but at ~110x sales the market is paying for Mars, orbital AI data centers, and a $60B Cursor bet that aren't earnings yet. | — | 1 |
| The only credible bet on FULL (both-stages) reuse besides SpaceX — a metallurgy/physics moat the others ducked — but it is a single-vehicle, zero-revenue, zero-flights company whose entire value is gated on one un-flown second stage surviving reentry; WATCHING, not investable, until Nova reaches orbit and the upper stage comes home intact. | — | 1 |