Skip to content
FrontierPicks

Forecasts · 6 bets · 39 refused

The model's bets, stated in advance.

When the radar surfaces a cluster, the model shapes it into a bet — naming the core names and stating a win-probability — or refuses it. What is fixed before the outcome is the date, the names and the probability; the bet then runs until a machine-owned exit closes it, and resolves into a binary result plus a categorical verdict against the mechanical baseline and the market. Unlike a dossier, a theme bet carries no kill line of its own — no price level is set for any of its names — so it is a dated, pre-committed forecast, not a falsifiable-on-a-close one. The exit rules that do end it are the same for every bet and are printed on every card below, thresholds included. This is an experiment at full transparency, not a proven edge.

Open bets · 6

v5 theme-bets the model has entered, each with a stated win-probability up front. Distinct from the names held on the live book and the research theses scored on the track record.

  • ATEX theme lead MEDIUM p 62% entered
    IRDM core ATEXMRCY

    Sovereign and critical-infrastructure connectivity is re-rating as utilities, governments, and defense buyers harden and modernize their networks: satellite communications (IRDM), private-wireless spectrum (ATEX), and secure defense electronics (MRCY) are bid together on rising grid-security and defense capex. Broad participation across the group plus heavy volume confirm a genuine theme rather than a single-name squeeze.

    How it resolves No bet-specific kill line: no price level is set for any name in this bet. Every theme bet runs on the same pre-registered mechanical exits — the basket down 50% against its entry cost, a single name down 70% against its entry price, the theme sleeve down 55% from its high-water mark, or a name breaking structurally (10 consecutive sessions under its 100-day average), the last of which becomes reachable only once the bet has been held its pre-registered minimum of 60 trading sessions. That minimum is a hold requirement, not a resolution date: nothing closes a bet because 60 sessions have elapsed. Fixed in advance: the date, the names, and the stated probability.

  • SIMO theme lead MEDIUM p 66% entered
    MU coreSTX core SIMOWDCUMC

    The AI data-center buildout is driving a genuine memory and storage supercycle: DRAM/HBM and NAND pricing are recovering while nearline HDD demand from hyperscalers stays tight. Micron and Seagate are the highest-quality, cash-generative center of this theme, with SIMO's NAND controllers and WDC riding the same storage upcycle and UMC providing broader foundry participation.

    How it resolves No bet-specific kill line: no price level is set for any name in this bet. Every theme bet runs on the same pre-registered mechanical exits — the basket down 50% against its entry cost, a single name down 70% against its entry price, the theme sleeve down 55% from its high-water mark, or a name breaking structurally (10 consecutive sessions under its 100-day average), the last of which becomes reachable only once the bet has been held its pre-registered minimum of 60 trading sessions. That minimum is a hold requirement, not a resolution date: nothing closes a bet because 60 sessions have elapsed. Fixed in advance: the date, the names, and the stated probability.

  • DELL theme lead MEDIUM p 64% entered
    DELL coreNTAP core HPEOKTASNX

    Enterprise AI-infrastructure buildout: hyperscaler and enterprise capex is pouring into AI-optimized servers, storage and IT distribution, and Dell is the volume leader turning that demand into record backlog. NetApp, HPE, the channel (SNX) and identity/security (OKTA) all ride the same datacenter refresh cycle — real earnings, real breadth (4 peers +30% in 3m, 1.82x volume), not a single squeeze.

    How it resolves No bet-specific kill line: no price level is set for any name in this bet. Every theme bet runs on the same pre-registered mechanical exits — the basket down 50% against its entry cost, a single name down 70% against its entry price, the theme sleeve down 55% from its high-water mark, or a name breaking structurally (10 consecutive sessions under its 100-day average), the last of which becomes reachable only once the bet has been held its pre-registered minimum of 60 trading sessions. That minimum is a hold requirement, not a resolution date: nothing closes a bet because 60 sessions have elapsed. Fixed in advance: the date, the names, and the stated probability.

  • SEZL theme lead MEDIUM p 64% entered
    SEZL coreDAVE core HNGEXMTRTXG

    Broad breadth (7 peers +30% in 3m) and 1.8x volume participation confirm real theme demand, not a single squeeze.

    How it resolves No bet-specific kill line: no price level is set for any name in this bet. Every theme bet runs on the same pre-registered mechanical exits — the basket down 50% against its entry cost, a single name down 70% against its entry price, the theme sleeve down 55% from its high-water mark, or a name breaking structurally (10 consecutive sessions under its 100-day average), the last of which becomes reachable only once the bet has been held its pre-registered minimum of 60 trading sessions. That minimum is a hold requirement, not a resolution date: nothing closes a bet because 60 sessions have elapsed. Fixed in advance: the date, the names, and the stated probability.

  • OSCR theme lead MEDIUM p 63% entered
    CNC coreHUM core OSCRHNGE

    Managed-care and ACA-marketplace insurers are re-rating together on a margin-recovery narrative, with sector-wide breadth and heavy volume rather than a single-name squeeze.

    How it resolves No bet-specific kill line: no price level is set for any name in this bet. Every theme bet runs on the same pre-registered mechanical exits — the basket down 50% against its entry cost, a single name down 70% against its entry price, the theme sleeve down 55% from its high-water mark, or a name breaking structurally (10 consecutive sessions under its 100-day average), the last of which becomes reachable only once the bet has been held its pre-registered minimum of 60 trading sessions. That minimum is a hold requirement, not a resolution date: nothing closes a bet because 60 sessions have elapsed. Fixed in advance: the date, the names, and the stated probability.

  • RXO theme lead MEDIUM p 62% entered
    RXO coreXPO core ARCBLSTRJBHT

    A genuine freight-cycle upturn after a multi-year trucking recession: brokerage, LTL and truckload volumes and spot rates are re-accelerating, lifting asset-light and asset-based carriers together. RXO leads as the highest-beta brokerage play; XPO's LTL execution is the durable compounding anchor.

    How it resolves No bet-specific kill line: no price level is set for any name in this bet. Every theme bet runs on the same pre-registered mechanical exits — the basket down 50% against its entry cost, a single name down 70% against its entry price, the theme sleeve down 55% from its high-water mark, or a name breaking structurally (10 consecutive sessions under its 100-day average), the last of which becomes reachable only once the bet has been held its pre-registered minimum of 60 trading sessions. That minimum is a hold requirement, not a resolution date: nothing closes a bet because 60 sessions have elapsed. Fixed in advance: the date, the names, and the stated probability.

What the model refused · 39

Clusters the radar flagged but the model declined to bet — co-movement that isn't a real theme. A refusal is a call too: curation only earns its keep if it beats betting every cluster.

  • OPI Refused
    OPIDFTXORKAAPGEBWIN

    Corporate-action artifact. The trigger and every corr-to-leader are synthetic. Remaining members are unrelated binary-readout biotechs (DFTX Ph3, ORKA, APGE) plus an insurance broker. No theme.

  • PAYS Refused
    PAYSHNSTUPAYCMGNIFAVACDCTH

    Index/ETF-flow (common-factor) artifact. Eight names, eight unrelated sectors: plasma-donation card processor, diapers, game engine, payroll SaaS, ad exchange, background screening, timeshare, oncology device. No shared narrative — leader correlations cap at 0.67 and cluster near 0.5, the signature of a high-beta small-cap rebound on risk-on flows. Any AI-software sub-story excludes the leader.

  • TBI Refused
    TBIQLYSZETAMTRN

    Uniform ~0.41 corr to leader = market/small-cap momentum beta, not a theme; breadth only 2/3 and MTRN already rolled (+11% 3m).

  • AMLX Refused
    AMLXXNCRDTILHAEWEAV

    Binary-event biotech lottery + sector-bucket artifact. XNCR and DTIL are independent readout/royalty events, not co-movers. HAE (blood-management devices) and WEAV (practice SaaS) share only the GICS Health Care bucket; corr 0.41-0.55, breadth 2.

  • URGN Refused
    URGNTVTXBLMNMBXTGTXDYNMTRNHQL

    Sector-beta/fund-flow artifact plus binary biotech. HQL is a closed-end biotech FUND holding this sector — a mechanical index twin. MBX (+109% 3m) and DYN are clinical-stage readout lotteries. URGN/TVTX/TGTX share no driver beyond XBI beta (bladder cancer vs IgAN vs MS). BLMN (restaurants) and MTRN (materials) prove the link is a risk-on factor. Corrs 0.41-0.54.

  • OPI Refused
    OPIDFTXORKAAPGEBWIN

    Corporate-action artifact + binary biotech. The only truly correlated names (APGE 0.90, DFTX 0.81, ORKA) are clinical-stage I&I biotechs — readout lotteries — and BWIN (insurance broker) shares no narrative with an office REIT. Co-movement, not a theme.

  • LIFE Refused
    LIFEXMTRGDRXSOPH

    Single low-float post-IPO squeeze masquerading as breadth. Peers share no narrative: XMTR (industrial marketplace), GDRX (Rx discounts), SOPH (genomics) each rallied on their own Q2 prints; corr just 0.42-0.44.

  • ZIP Refused
    ZIPUHUMTWLONSITBNEDSPT

    Factor/beta artifact, not a theme: managed care (HUM), IT distribution (NSIT), college retail (BNED), game engine (U), CPaaS (TWLO), social SaaS (SPT), job board (ZIP) share no demand driver. Tell: uniform ~0.40-0.46 correlations across unrelated sectors = one beaten-down/high-short-interest mean-reversion factor in a broad up-tape. An "AI software" story needs discarding 3 of 7 members.

  • URGN Refused
    URGNMTRNBLMNTGTXACRSDYN

    Binary-event biotech basket with no narrative link. Four unrelated-indication biotechs (URGN bladder cancer, TGTX MS, ACRS derm, DYN muscle) plus a beryllium/specialty-materials name (MTRN) and a steakhouse chain (BLMN). Correlations 0.40-0.55 are drift, not theme; moves are idiosyncratic FDA/readout and single-name catalysts. No theme to compound.

  • CDNA Refused
    CDNAMANATAIBAX

    Correlation artifact — no shared narrative. Members span four unrelated industries: transplant diagnostics (CDNA), temp staffing (MAN), clinical-stage psychedelics (ATAI), hospital products (BAX). Correlations to leader are only 0.40–0.53 — coincident beaten-down value/rebound factor beta, not a theme. ATAI is additionally a binary-readout biotech lottery.

  • CMPS Refused
    CMPSBBGHRSATAIERAS

    Binary-event biotech. CMPS/ATAI/GHRS are clinical-stage psychedelics driven by Phase 2b/3 readouts and FDA-path news (COMP360, GH001); ERAS is another pre-revenue oncology binary. ATAI is equity/deal-entangled with CMPS, so not independent breadth, and BB (BlackBerry) is an unrelated software name at 0.40 corr padding the count. Readout lottery, not a compounding theme.

  • TBI Refused
    TBIFTKZETAQLYSMTRNSHOPTALK

    Factor/beta artifact, no shared narrative: staffing, oilfield chemicals, adtech, cybersecurity, beryllium, e-commerce, telehealth. All catalysts idiosyncratic — TBI is one +28.7% post-Q2 earnings gap, FTK a Puerto Rico power contract. Corr-to-leader caps at 0.50 (all 0.41-0.50 = market beta). Low-price microcap 3x off its low masquerading as breadth.

  • PAYS Refused
    PAYSHNSTVACMGNIFAPAYCASPNU

    Factor-flow artifact, not a narrative: 8 unrelated industries (PAYS plasma/pharma payments, HNST diapers, VAC timeshare, MGNI adtech, FA screening, PAYC payroll, ASPN aerogels, U game engine). Corr 0.42-0.68 = high-beta/high-short-interest small-cap risk-on rotation, not a shared driver. Leader PAYS has zero business overlap with any member; standalone parabolic at its 252d high.

  • CRSR Refused
    CRSRPUBMSTLNQNSTOMDAAMNNTRAMOV

    Index/ETF-flow artifact: a beaten-down small-cap value / high-short-interest factor rotation, not a theme. Members span gaming peripherals, adtech, lead-gen, luxury watches, healthcare staffing and diagnostics — no shared catalyst, customer or supply chain. Leader correlations of only 0.40-0.55 are generic small-cap beta, not one narrative.

  • CLYM Refused
    CLYMDTILATICLDXAMLXEOLSEWTXDCTH

    Binary-event biotech basket. 7 of 8 members are clinical-stage readout lotteries (CLYM, DTIL, CLDX, AMLX, EWTX largely pre-revenue); the 8th, ATI, is aerospace/specialty metals with zero thematic link. Corr to leader is only 0.40-0.55 = XBI sector beta, not a shared narrative. Leader is pre-revenue, pinned at its 252d high after an ~8.7x, with a coin-flip trial readout as the next catalyst.

  • CORT Refused
    CORTMCSPLPCAGXBTSGBTSGUCORZILMN

    Share-class twin + no common narrative. BTSG/BTSGU are the same issuer (BrightSpring common vs units), padding breadth. The rest span Cushing's pharma (CORT), cinemas (MCS), utility hardware (PLPC), power EPC (AGX), bitcoin/AI datacenters (CORZ), sequencing (ILMN) — no shared driver. Corr 0.40-0.46 = market beta; CORT's move is a single-asset relacorilant catalyst.

  • PRCH Refused
    PRCHMCSSTGWEXTRFFIVDSP

    Sector-incoherent SMID beta basket, no shared narrative. PRCH's +148% is idiosyncratic (reciprocal-exchange restructuring) and shares nothing with theatres/hotels (MCS), agency+ad-tech (STGW, DSP) or networking (EXTR, FFIV) — three unrelated sub-themes. Uniform 0.41-0.46 correlation is generic small-cap beta, not theme co-movement; only 2 of 5 peers participate over 3m, EXTR flat.

  • URGN Refused
    URGNMTRNBLMNTGTXACRSDYNSSRMIBRX

    Binary-event biotech lottery plus unrelated sector strays. Five members are unconnected drug stories (URGN, TGTX, ACRS, DYN, IBRX) sharing no end market; SSRM (gold), MTRN (beryllium), BLMN (steakhouse) overlap with nothing. All peer correlations 0.42-0.54 = generic small-cap risk-on beta, not thematic co-movement. No narrative center to weight.

  • GHRS Refused
    GHRSCMPSATAINKTR

    Binary-event biotech (readout/FDA lottery). GHRS, CMPS, ATAI are pre-revenue psychedelic-CNS names moving on Phase 2b/3 data and scheduling headlines, not a compounding theme; ATAI/CMPS are further linked by atai's equity stake, so breadth is one event book. NKTR (corr 0.43, 3m -8%) is unrelated immunology. All carry post-data ATM dilution risk.

  • VSTS Refused
    VSTSAMBQMGNXPLSEARMKPACSUNFSPHR

    Correlation artifact, not a theme: grab-bag of unrelated sectors — uniform services (VSTS/UNF/ARMK), edge semis (AMBQ), biotech (MGNX), medtech (PLSE), nursing (PACS), Vegas Sphere (SPHR) — sharing only risk-on beta (corr 0.41-0.57). Leader VSTS is M&A/event-driven, pinned near its 252d high (capped), MGNX is a binary biotech, breadth thin (2 peers).

  • SEZL Refused
    SEZLDAVETWSTTXGLFSTSNEXXMTROPRA

    Spurious-breadth / momentum-factor artifact, not a narrative theme. 8 members span ~6 unrelated industries (SEZL BNPL, DAVE neobank, TWST DNA synth, TXG genomics tools, LFST mental-health, SNEX commodity brokerage, XMTR mfg marketplace, OPRA browser), weak corr 0.42-0.60. Only SEZL/DAVE share a story (consumer fintech), too narrow to carry breadth; rest is unrelated small-cap momentum co-movement.

  • DVA Refused
    DVAEVCCVGIFLEXTGTXGEOLTHJAZZ

    Spurious cross-sector momentum grab-bag — no shared theme. Members span dialysis (DVA), Spanish media (EVC), micro-cap truck parts (CVGI), EMS electronics (FLEX), MS biotech (TGTX), private prisons (GEO), gyms (LTH), pharma (JAZZ). The 0.43–0.73 corrs are bull-market beta, not a narrative; each name has its own catalyst. CVGI (+183%/6m, +7%/3m) is a spent squeeze; DVA is -1% off its 252d high.

  • LQDA Refused
    LQDAMPLTDRTSNBTXPTGXNESRIXBZH

    Correlation artifact, no shared narrative. Leader LQDA is a binary FDA/litigation pharma (Yutrepia vs United Therapeutics); All peer corrs ~0.41-0.48; breadth is spurious.

  • EVC Refused
    EVCDVAFLEXGEOTGTXNBTXLTHDOC

    no shared narrative driver, just bull-market beta. Also carries binary-biotech risk (TGTX/NBTX).

  • DFTX Refused
    DFTXOPIAPGEKYMRCGEMBCAXLILAKBHVN

    Binary-event biotech basket / FDA-readout lottery: DFTX, APGE, KYMR, CGEM, BCAX, BHVN are heterogeneous clinical-stage biotechs in different modalities with no shared commercial narrative — XBI sector beta, not a theme. OPI (+8900%, 0.81 corr) is a spurious squeeze artifact and LILAK an unrelated LatAm telecom padding fake breadth.

  • DNTH Refused
    DNTHMGNXCRVSELVNDRTSTRVIPVLAMGTX

    binary-event biotechs (FDA/readout lotteries): 8 unrelated clinical-stage, pre-revenue names — complement (DNTH), oncology antibodies (MGNX), ITK (CRVS), CML kinase (ELVN), radiation (DRTS), cough (TRVI), rare-derm (PVLA), gene therapy (MGTX). No shared narrative; low cross-corr (~0.45) = shared XBI sector beta, not a theme. Each is an idiosyncratic trial-readout lottery.

  • PTRN Refused
    PTRNDDOGFTNTCLDTLFSTXMTRHNSTCCRN

    Correlation artifact — spurious co-momentum/beta cluster, no shared narrative. Members span unrelated sectors (FTNT cybersecurity, DDOG cloud-monitoring, CLDT hotel REIT, LFST mental-health, XMTR mfg marketplace, HNST baby products, CCRN healthcare staffing) and leader correlations are only 0.42-0.54, so the >100% breadth is a risk-on beta wave, not a theme.

  • HNGE Refused
    HNGEPANWTENBRELYTBIOSCRCRWDOMDA

    Momentum-factor basket masquerading as breadth — no shared narrative. Members span 4 unrelated sectors: cybersecurity (CRWD/PANW/TENB), digital health (HNGE/OSCR/OMDA), remittances (RELY), staffing (TBI); only 0.40-0.56 corr to leader. Co-triggered by a common risk-on/momentum factor, not one story — no narrative makes a staffing agency, CrowdStrike and Hinge Health co-beneficiaries.

  • PLSE Refused
    PLSEVSTSARMKHLIORAL

    Correlation artifact — no shared narrative. Five unrelated sectors: PLSE bioelectric medical device, VSTS uniform rental, ARMK food/facilities, HLIO hydraulics, RAL test instruments. Peer corr only 0.41–0.44, breadth peers=2. VSTS is a standalone M&A/takeover story; PLSE a Duggan-controlled single-name device squeeze near its 252d high. Spurious co-movement, not a theme.

  • ELDN Refused
    ELDNIBRXTRVICADLSLDBRAPPROIV

    binary-event biotechs (FDA/readout lotteries): seven unrelated clinical-stage names (transplant, bladder-cancer IO, cough, oncolytic virus, DMD gene therapy, epilepsy, holdco), each driven by its own binary readout. Low cross-corr (0.40-0.52) = no shared driver; 6m move is small-cap biotech beta and 3m breadth is only 2 names. No narrative theme.

  • ATEX Refused
    ATEXIRDMATENMBX

    Single-name re-rating masquerading as breadth. Members span four unrelated sectors — 900MHz spectrum leasing (ATEX), LEO satellite (IRDM), application networking (ATEN), endocrine biotech (MBX) — with only weak pairwise corr (~0.43–0.49) and the leader (+364%) far outrunning peers. No shared narrative driver; MBX is a binary-event biotech. Correlation-artifact grab-bag.

  • DFTX Refused
    DFTXAPGECGONKYMRBHVNACA

    binary-event biotech (FDA/readout lottery) basket — APGE/CGON/KYMR/BHVN are unrelated clinical-stage names (atopic derm, bladder cancer onco-virus, protein degradation, neuro) with no shared modality/indication; co-movement is XBI risk-on beta, not a narrative theme. Weak corr (0.48/0.48/0.44/0.41), only 3 peers >30% in 3m, and a lone +235% leader = single readout run masquerading as breadth.

  • CORT Refused
    CORTACHCBTSGDNLIRPRX

    Correlation artifact: healthcare grab-bag, no shared narrative. Leader CORT (cortisol pharma, idiosyncratic ROSELLA ovarian-cancer catalyst) carries it; peers are unrelated models — ACHC behavioral facilities, BTSG home-health services, DNLI binary neuro biotech, RPRX royalty aggregator. Uniform low corr ~0.40 = sector/factor beta, not a theme; breadth weak (2/4 >30%).

  • SEZL Refused
    SEZLTWSTDAVEDDOGSNEXVACLFSTXMTR

    factor/beta momentum artifact, not a narrative theme. Members span 6 unrelated sectors (SEZL/DAVE fintech, SNEX brokerage, TWST synth-DNA biotech, DDOG SaaS, VAC timeshare, LFST mental-health, XMTR manufacturing) with only 0.40-0.58 corr to leader. Co-movement is risk-on high-beta ripping together, no shared story. Shaping would require dropping 5 of 8 names and manufacturing conviction.

  • DVA Refused
    DVAFLEXTGTXGEOLTHDOCJAZZPFGC

    Momentum-factor beta basket, no shared narrative. 8 unrelated sectors: dialysis (DVA), contract electronics (FLEX), private prisons (GEO), MS biotech (TGTX), gyms (LTH), healthcare REIT (DOC), pharma (JAZZ), food distro (PFGC). Moderate ~0.5 pairwise corr = common momentum/beta, not a driver; each top name has its own story. No theme center to curate.

  • DFTX Refused
    DFTXAPGECGONKYMRBHVN

    binary-event biotechs (FDA/readout lotteries) — grab-bag of clinical-stage biotechs across unrelated indications (APGE immunology, CGON bladder-cancer, KYMR protein degradation, BHVN neuro), no shared compounding narrative. Co-movement is XBI sector-ETF flow beta plus idiosyncratic readout spikes; sub-0.5 corr on 3 of 4 peers = sector twins, not a theme.

  • SYRE Refused
    SYREORKARVMDMKSIAPGECGNXARWESI

    binary-event biotech (FDA/readout lottery): leader SYRE and the two top-return narrative-center peers ORKA/APGE are pre-revenue clinical-stage I&I antibody names; remaining "breadth" (MKSI, CGNX, ARW, ESI) are unrelated profitable industrials sharing only risk-on beta (corr 0.41-0.54). Two incompatible sub-themes glued by beta, binary-biotech core.

  • PTRN Refused
    PTRNDDOGFTNTLFSTXMTRWRBY

    index/ETF-flow twins — heterogeneous high-beta growth basket (DDOG observability, FTNT cybersecurity, LFST behavioral health, XMTR manufacturing marketplace, WRBY eyewear) sharing no end-market or narrative; weak ~0.41-0.47 leader correlations = co-movement is momentum/growth-ETF factor flow, not a real theme.

  • DVA Refused
    DVAFLEXTGTXGEOVECODOCLTHVCYT

    spurious cross-sector correlation / factor-beta grab-bag — no shared narrative theme. Members span 6+ unrelated industries: DVA dialysis, FLEX electronics EMS, TGTX MS biotech, GEO private prisons, VECO semi equipment, DOC healthcare REIT, LTH gyms, VCYT diagnostics. Moderate corr (0.40–0.66) reflects common market beta, not a compounding theme. No coherent bet to shape.

The honest frame

An experiment, scored in the open.

Every mechanical harvest of this pattern lost to simply owning the market. The one untested part is whether the model choosing the names, the membership, and the timing adds anything — and that can only be answered forward. So it ships as a forecast log with a stated probability on every bet, feeding the public Brier score, under pre-registered kill gates that shut it down if it doesn't work.

Common questions

What is a theme forecast?
When the weekly radar surfaces a cluster of stocks moving together, the model either shapes it into a bet — picking the core names, and stating a probability that the theme plays out — or rejects it. The date, the names and the probability are fixed before the outcome, which is what makes the bet scoreable. What a theme bet does not carry is a level chosen for it: no price is set for any of its names in advance. It ends on uniform machine-owned rules instead — a minimum hold that has to pass before a sustained trend break can close a name, plus fixed drawdown stops on the name, the basket and the whole theme sleeve. Every card publishes those thresholds, so a theme bet is a dated pre-commitment rather than a falsifiable-on-a-close claim the way a per-ticker dossier is.
How is a bet scored?
Two ways, both categorical and non-monetary. Against the raw mechanical cluster (equal-weighting every name the scan flagged) — did the model's curation add anything? And against matched-window SPY — did the bet beat the market? Each resolved bet shows beat or trailed by a whole number of percentage points; the underlying returns are never published.
Why publish the rejections?
Because a refusal is a call. The radar is mechanical and flags plenty of co-movement that isn't a real theme — index twins, merger arb, a grab-bag of biotech readout lotteries. Showing what the model declined, and why, is part of the honest record: the curation only earns its place if it beats simply betting every cluster.
Does a theme bet have a kill line?
Not one of its own. No level is ever chosen for a particular bet. Every theme bet is governed by the same pre-registered mechanical rules, published in full on each card: a minimum hold that must pass before a sustained trend break can close a name, plus fixed drawdown stops on the single name, on the basket, and on the whole theme sleeve. One of those — the single-name stop — is a price rule, which is why its threshold is published rather than the site pretending no price matters anywhere. What does not exist is a level set for this bet in advance and graded against on the close. Per-ticker picks are different: those publish a price kill line up front and are graded against it on real closes, which is why they carry a Brier score and these do not.
Is this investment advice?
No. These are research forecasts, published to be scored in public — not recommendations, not a trade to copy. No position sizes, prices, or account details are ever shown.