NVDA
Dominant GPU supplier for AI training and inference; agentic AI multiplies compute demand per task.
Conviction breakdown
The case for this chipmaker remains exceptionally strong, backed by massive AI infrastructure deals and a dominant position in the space, with recent price softness seen as a pause rather than a problem.
Initial read this run: 93 → ranked to 99.
Run-by-run detail
The fund's full reasoning for each scoring run, in its own words. More technical than the plain-English summary above.
Best-in-batch forward setup: dominant thesis, exceptional catalyst stack (OpenAI $250B, SK $500B, NAVER), and solid win-probability. Momentum is the only mild concern, reading as consolidation rather than breakdown. This is exactly the winning pattern (thesis >=97, AI infra theme, sustained conviction) our learnings flagged. Sizing is the Sizer's call.
Top thesis rank in the batch with exceptional catalyst flow ($500B SK Hynix, $10B Naver deals, Blackwell ramp). Win-probability at 72 is strong, momentum at 56 shows healthy consolidation not extension. Risk quality 75 solidifies the forward case. Best stock in the batch from here — sizing is Sizer's problem, not mine.
Top of the batch on forward setup: highest thesis rank, strongest catalyst cluster today, and a healthy 56 momentum reading that isn't extended near the 52wk high (86% of high, not 99%). Win-Prob at 72 reflects clear path to a 10%+ move on the next earnings cycle. Concentration is the Sizer's problem, not mine — as a forward view of the stock, this is our highest-conviction name.
Top thesis rank in the batch with corroborating win-prob and a rich catalyst tape (Fort Worth GB300, Nscale, Blackwell ramp). Momentum is confirming without being extended (86% of high, mid-range). Forward setup remains best-in-class AI infrastructure; sizing/concentration is the Sizer's call, not mine.
Top-of-batch on every forward metric. Thesis at the 100th percentile, exceptional catalyst density today (Vera Rubin mass production, 300+ partners, 1000 racks/day target), and win-probability confirms continued upside with Blackwell ramp. Momentum at 65 shows healthy consolidation near highs — not extended. Risk quality 75 is strong for a mega-cap AI leader. The Sizer will handle concentration; my job is to score the stock, and from here NVDA remains the highest-conviction name in the batch.
Top-of-batch thesis and catalyst convergence. NVDA remains the dominant agentic-AI infrastructure name with multiple T1 catalysts landing today (Vera Rubin, Grace servers, Spectrum-6 networking). Win-probability of 72 is solid, momentum confirming at 70. Trading at 86% of 52-week high is elevated but PE of 31 is reasonable given growth. Forward setup remains the best in the batch — sizing decisions are the Sizer's problem.
Top-of-batch forward setup: thesis at the ceiling, direct positive catalysts today (Blackwell ramp, Nemotron 3, hot-water cooling), and mid-range momentum at 51% keeps it out of the extended zone that burned us on MRVL/ARM. Win-probability 72 is solid though not spectacular given PE at 31 vs growth. Risk quality high. This is the anchor AI infrastructure name — score the stock high, let the Sizer worry about concentration.
Top-ranked thesis in the batch with a direct positive news cluster today — TSMC record quarter, Blackwell ramp, ecosystem partnerships all reinforcing. Momentum mid-range at 52% means not extended near highs (86% of 52wk high but consolidating), which fits our winning pattern. Win-Probability at 72 is solid though not exceptional. Forward setup remains best-in-class AI infrastructure. Sizing is Sizer's call.
Best-in-class thesis rank with supportive win-prob, strong catalyst flow, and quality profile. The forward setup remains the strongest in the batch — AI datacenter demand and Blackwell ramp intact, with a clear breakout trigger above $236. Concentration is a Sizer concern, not a conviction concern; from a pure forward-return view NVDA is a top-tier hold.
Top thesis rank in the batch with multiple T2 catalyst confirmations today (SoftBank AI-RAN, Vera Rubin, Blackwell ramp). Win-probability solid at 72 with breakout setup near 52wk high. Risk quality strong for a mega-cap. This is the highest-conviction forward setup in the group; sizing concerns are the Sizer's problem, not mine.
Top-of-batch thesis rank with strong catalyst confirmation across multiple product launches (robotics AI, Nokia AI-RAN, Blackwell ramp). Win-probability solid at 72 with breakout setup near 52wk high. Risk_quality respectable at 75 for a mega-cap. Forward view remains best-in-class; concentration is the Sizer's problem, not mine.
NVDA leads the batch on thesis and has the cleanest forward setup: Blackwell ramp intact, dense positive catalyst coverage, and a breakout structure near 52wk high with supportive momentum. Win-probability at 72 confirms the 'will go up' side. Concentration is a Sizer concern, not a conviction concern — from a pure forward-return view, this is a core name.
Top-of-batch thesis strength with strong forward win-probability and multiple direct catalyst mentions today. Blackwell ramp, cooling alliances, and continued datacenter demand all confirm the bull setup. Momentum is only mid-range because it's consolidating below $236 resistance — a breakout there unlocks continuation. This is a core forward view; concentration is the Sizer's problem, not mine.
Top-of-batch on thesis with strong forward setup: Blackwell ramp intact, three+ direct positive catalyst articles today (Valar nuclear, FBI supercomputer, Kazakhstan $10B), momentum confirming. Win-prob at 72 reflects breakout-in-progress needing $236 clear. Best-in-batch AI infrastructure exposure. Concentration is Sizer's problem, not mine — stock quality from here is elite.
Top-of-batch on thesis and catalyst breadth with confirming momentum and solid risk_quality. Blackwell ramp intact, media/analyst convergence high, and the setup near breakout of $236 offers real optionality. Forward view is the strongest in this group — sizing is the Sizer's call.
Top-percentile thesis, dominant AI accelerator franchise, PE ~30x is reasonable for the growth profile, and multi-source catalyst news (Firmus, Gradium) reinforces the demand narrative. Momentum confirming, quality high. Best forward setup in the batch. Sizer handles the concentration question — I'm scoring the stock.
Top thesis rank in the batch with confirming momentum and a modest PE for the growth profile. Multiple agentic AI catalysts landing this week reinforce sustained compute demand. Win-probability is only 66, which keeps this out of the 90+ zone, but the forward setup remains the strongest in the group. Sizing decisions belong to the Sizer; my job is to say this is still a high-conviction forward hold.
Top-of-batch thesis rank, durable AI accelerator dominance, and PE of ~30x on 50%+ EPS growth make the forward setup attractive. Momentum confirming at 76 with the stock holding above mid-range. Today's 'trillion-dollar slide' headline is a real negative catalyst but doesn't break the thesis. Forward conviction remains strong; sizing is the Sizer's problem.
Best-in-class forward setup: dominant AI accelerator franchise, reasonable multiple for the growth, and catalysts landing daily. Win-prob only 66 reflects the -18% from highs and macro overhang, not a thesis crack. This is the highest-quality forward story in the batch. Sizing/concentration is the Sizer's problem, not mine.
Top-of-batch thesis with dominant AI accelerator position, cheap-for-growth 30x PE, and today's catalyst flow spanning Kazakhstan, Indonesia, and next-gen Vera/Rigel architecture. Win-probability at 66 is the governor — not exceptional — but every specialist is aligned bull and no deterioration signals. Momentum at 56 shows the stock is digesting rather than breaking. Forward setup remains strong; sizing is the Sizer's call.
NVDA remains the AI infrastructure anchor with the strongest thesis in the batch and a rich catalyst backdrop today. Win-probability is only moderate because the stock is mid-range and has been consolidating, but at 30x PE with 50%+ EPS growth the setup from here is asymmetric to the upside. Forward view is clearly top-tier.
Dominant AI accelerator franchise trading at 30x PE with 50%+ EPS growth is a rare combination. Thesis strength is unambiguous at the top of the watchlist and catalyst flow is directly confirmatory. Momentum has cooled with the pullback from $236, which is why I'm not pushing higher. From here, forward setup remains one of the strongest in the book. Sizing concentration is the Sizer's problem, not mine.
Best-in-class AI infrastructure thesis with the strongest fundamental setup in the batch and a heavy positive news cycle today. Momentum is middling — we're not at the breakout the playbook loves — but PE 29 for the category leader plus Blackwell ramp keeps the forward setup strong. High forward conviction; sizing concerns belong to the Sizer.
Top thesis rank in the batch with a still-reasonable valuation and clear Blackwell/data-center demand. Win-probability is only 65 and momentum has cooled to mid-pack, so this isn't the +30% breakout setup — but the forward setup remains constructive from here. Core hold; sizing decisions are the Sizer's problem.
Best thesis in the batch by a wide margin — dominant AI infra provider at a shockingly reasonable 29x PE with 19% room back to highs. Win-probability of 65 is solid, catalyst flow is loud and direct (revenue-sharing AI cloud, US manufacturing, data center demand). Momentum is only mid-pack which is the one hesitation — not a breakout setup like our best MU/MRVL/AMAT winners were. But forward setup from here is clearly the strongest in the batch. Sizing is the Sizer's problem.
Top thesis percentile in the batch with the strongest catalyst flow — Blackwell Ultra/Anthropic/Azure validates the agentic-AI infrastructure narrative directly. Win-prob at 65 is good not great, reflecting 19% gap to highs and macro/tariff overhang, but PE 29x on the dominant AI infra franchise leaves room. Momentum at 55 means we're not catching a breakout, we're holding the benchmark. Forward setup remains the strongest of the group.
Dominant AI infrastructure franchise with the strongest thesis in the batch and richest catalyst flow (IREN $4.4B, Firmus, full-stack dominance). Win-probability is moderate at 65 reflecting near-term consolidation, and momentum has softened, but PE of 29 on this growth profile is striking. Forward setup remains clearly constructive — this is a hold-and-let-it-work name.
Top-of-batch thesis at 100th percentile with a reasonable 32x PE for the growth profile and strong win-probability of 70. Multiple direct positive catalysts today across sovereign AI, biotech deployment, and analyst comparisons. The weak momentum print (37) is the one yellow flag — NVDA is consolidating while peers like TSM and AMD break out — but the forward setup remains the cleanest in the AI infrastructure complex. Concentration is the Sizer's problem, not mine.
Best-in-class AI infrastructure thesis with the broadest catalyst surface — BioNeMo, Ethernet switching dominance, sovereign AI, DGX Spark. PE of 32x is genuinely cheap for the growth profile, suggesting earnings revisions still ahead. The weak momentum print is the one yellow flag, but it reflects digestion after a massive run rather than a broken setup. Forward expected return remains strong; concentration concerns belong to the Sizer.
Top-percentile thesis with reasonable PE for the growth trajectory and a steady drumbeat of supportive AI infrastructure catalysts. Win-probability is solid at 70 with 11% room to the 52w high. Momentum is the one soft point — the stock has lagged while peers ripped — but the forward setup remains the strongest in the AI chip complex. Position sizing is the Sizer's call.
Forward setup is the strongest in the batch. OpenAI 10GW partnership, KKR-Kuwait $10B venture, Stargate Texas, and 409% datacenter revenue growth represent a once-in-a-cycle catalyst stack. Thesis is top-percentile and win-probability is solid at 70 with valuation actually supporting further upside. Momentum print of 34 is a transient consolidation under the 52w high, not a breakdown — the catalyst flow will pull price. Top conviction in the batch.
Best-in-class thesis with an exceptional cluster of directly-named T1 catalysts today (KKR $10B, OpenAI Ohio, 35 European supercomputers). PE of 32x is unusually reasonable for the growth profile. Momentum is the only soft spot — stock is 11% off highs while news flow rips — but that's a setup, not a breakdown. Forward setup remains top-tier.
Top-of-batch on thesis and catalyst — Blackwell ramp, Halos robotics launch, Hut 8 $4.25B deal, and agentic AI telecom expansion all directly confirm NVDA's central role. Win-probability solid at 70 with PE only 32x and 11% below 52w high giving room. Momentum is the one soft spot but the forward setup from here is the strongest in the batch. Sizing is the Sizer's call.
Top thesis percentile in the batch, with a still-reasonable PE and 11% room to 52w high. Win-probability is solid not spectacular, but the forward setup — Blackwell ramp, sovereign AI, hyperscaler capex — remains the cleanest in the universe. Catalyst flow is directly supportive. Sizing is the Sizer's problem; from a forward-return standpoint NVDA is still a core conviction name.
Top-of-stack thesis, reasonable PE for growth trajectory, and concrete named-partnership catalysts in the last 24h. Momentum is only mid-range as it bases around $200 but the forward setup remains the cleanest in the batch. Score the stock high; concentration is the Sizer's problem.
Forward setup remains best-in-class: top thesis percentile, reasonable PE for growth profile, multiple concrete near-term catalysts (HPE AI Factory, Blackwell ramp, sovereign AI). Momentum is only middling at 60 as stock bases 13% below highs, which tempers urgency but doesn't break the thesis. Win-Probability solidly constructive at 72. Sizer owns the concentration call.
Top-decile thesis, reasonable PE for growth profile, and a wall of concrete same-day catalysts (HPE, Coherent, CoreWeave). Stock 13% off highs giving room to run. Forward setup remains best-in-class; sizing is the Sizer's problem.
Top thesis rank in batch with a stack of concrete name-specific catalysts landing today (IDC #1 ethernet, Vultr/HPE, Oracle GPU surge, Equinix-Cisco alliance). Win-prob solid at 72, momentum constructive at 67 with base forming. Forward setup remains best-in-class AI infrastructure exposure. Sizing concerns belong to the Sizer.
Top-of-batch thesis with cheap valuation for the growth profile, multiple concrete catalysts (Blackwell ramp, sovereign AI, Jensen Korea visit), and steady analyst revisions. Momentum is mid-pack but the forward setup is the strongest in the group. Sizing decisions on concentration belong to the Sizer.
Top-of-batch thesis percentile with multiple concrete catalysts hitting today (SK Telecom partnership, home AI server, FT enterprise/consumer expansion). PE of 31x is genuinely cheap for the growth profile, and the $200 base looks constructive ahead of Blackwell ramp. Forward setup is as strong as anything in the AI-infra complex. Concentration is the Sizer's call, not mine.
Best-in-batch forward setup: top thesis percentile, reasonable valuation for 100%+ EPS growth, Blackwell ramp + sovereign AI + multiple direct positive catalysts today. Momentum confirming with base formation around $200. Win-probability solid at 72. Concentration is the Sizer's problem, not mine — the stock itself remains the highest conviction AI name.
Top-of-batch thesis percentile and supportive global expansion catalysts (Vera CPU, LG robotics, Australia DC) outweigh the China headwind. Pullback off highs at 31x PE on >50% EPS growth keeps the forward setup compelling. Win-prob is only moderate due to near-term selling pressure, but the AI infrastructure cycle remains the strongest call in the book.
Best-in-class AI infrastructure name with top thesis ranking and confirming catalysts (Vera CPU, Australia data center, LG robotics). Pullback to $205 with PE 31.5x is historically attractive given EPS trajectory. Forward setup remains strong; China headwind is real but offset by global expansion. Position management is the Sizer's call.
NVDA prints top thesis score with strong catalyst convergence and reasonable win probability. Down only 2.8% from entry and 13% off highs with PE 31.5x is historically attractive. Risk agent flags concentration but that is a Sizer decision, not a conviction haircut. Core hold.
NVDA is the highest-thesis name in the batch with a torrent of concrete catalysts (Naver gigawatt factories, Sharon AI 40k GPUs, Vertiv alliance, 85% revenue growth). Win-probability is moderate at 65 due to short-term selling but valuation at 31.5x PE is historically cheap given growth. Held position roughly flat — protect the core. No trim warranted.
NVDA prints the highest thesis score in the batch with strong catalyst convergence (SpaceX, SK AI data center, Blackwell cycle). Win-prob 65 is solid given PE 31.5x is historically cheap for >50% EPS growth. Position is modestly underwater but the AI infrastructure cycle thesis is intact and specialists are aligned bullish. Hold core position; sizer will manage concentration.
NVDA remains the highest thesis-ranked name with PE of 31.5x historically cheap given EPS trajectory. Pullback to $205 looks like opportunity not breakdown — three of four specialists constructive. Held position only modestly underwater (-4%), no breakdown signal. Maintain core conviction; sizing concerns are for the Sizer agent, not me.
Core AI infrastructure holding with top-decile thesis and a stacked positive catalyst day (Reuters guidance beat, SK Telecom Blackwell, LG/Naver AI factories). Down 4.6% from entry is noise, not a thesis break. Win-prob is moderate (65) reflecting near-term pullback, but PE 31.5x with >50% EPS growth is historically cheap. Hold at current weight; sizer can decide concentration trim.
Top thesis score in the batch with an exceptional catalyst cluster (NAVER, SK Telecom, LG gigawatt-scale partnerships). Win-prob only 65 reflects near-term technical softness, but Blackwell cycle and hyperscaler capex remain intact. Position is held essentially flat — protect core AI infrastructure leader. No reason to trim; sizing concerns belong to Portfolio Sizer.
NVDA is the highest-thesis name in the batch with a dense direct-catalyst cluster (Blackwell ramp, GB300, sovereign AI) and quality risk profile. Slight underwater PnL but near 52wk highs and consolidating. Core holding — protect it. Concentration is a portfolio-sizer problem, not a conviction problem.
NVDA remains the cleanest expression of the AI infrastructure thesis with top thesis percentile and landmark catalyst signals. Slight underwater PnL is noise — consolidation near ATH with intact bull setup. Core hold, do not trim. Sizing concentration is a Portfolio Sizer concern, not a conviction issue.
NVDA sits at the apex of the thesis ranking with broad-based catalyst support (Blackwell, sovereign AI, hyperscaler capex). Near all-time highs with healthy consolidation. Slight PnL drawdown is noise. This is the anchor AI infrastructure position — high conviction hold. Concentration concerns are for the sizer agent, not me.
NVDA is the strongest thesis name in the batch with overwhelming catalyst flow (LG Blackwell, Mistral, Apple, Lightmatter, Navitas). Held at modest +3% but specialist consensus is bull across the board. Win-prob 65 is restrained only by near-ATH levels. Protect and hold; sizing decisions belong to the portfolio sizer.
NVDA is the cleanest agentic AI infrastructure name in the book — thesis at 100th percentile, catalyst broadly confirmed across multiple T2 sources today (Vera Rubin, MGX/DGX, Microsoft stack), and win-probability supportive at 65. PnL +2.13 means we're not yet in protect-winner territory but the thesis is fully intact. Concentration is a Portfolio Sizer problem, not a conviction one. Core hold at high conviction.
NVDA sits at the top of the thesis ranking with exceptional catalyst breadth — multiple direct partnership and product announcements today. Win-probability of 65 is constructive with Blackwell ramp and sovereign AI spend. Held position up modestly, but the franchise is the centerpiece of the AI infrastructure thesis. Protect and hold; sizing concerns belong to the portfolio sizer.
NVDA is the highest-conviction name in the batch: top thesis percentile, exceptional catalyst breadth (Vera Rubin, RTX Spark, DGX Station, Foxconn healthcare), strong momentum, and held with meaningful gains. Win-prob a touch lower (65) only because near ATH, but pattern of confirmed AI-infra leader breaking out is exactly the winning setup. Protect and hold.
NVDA remains the anchor of the AI-infrastructure thesis with top-percentile thesis score and a dense, ticker-specific catalyst cluster today (Jensen parabolic comments, PC chip expansion, photonics). Held position modestly up. Win-Prob of 68 is the only soft spot but reflects already-large weight rather than thesis erosion. Protect the core.
NVDA remains the anchor AI infrastructure name with top thesis rank and the densest direct catalyst flow in the batch (photonics $6.5B bet, Taiwan backbone, Groq dynamics). Win-probability is moderate at 68 reflecting near-52wk consolidation, but the thesis convergence and catalyst density justify protecting this as a core hold. PnL only +4.9% so winner-protection rule doesn't apply at full force, but no reason to reduce. Sizer agent handles concentration.
NVDA remains the AI infrastructure anchor with top thesis percentile and exceptional T1/T2 catalyst breadth today (Blackwell, IREN $3.6B, SoftBank sovereign cloud, 20X supply demand). Win_prob 68 is moderate but reflects mature setup not breakdown. Held position up modestly - protect and let it work. No add given concentration, but conviction stays high.
NVDA tops the thesis ranking at 100 with three direct catalyst articles today (Vera CPU $200B TAM, analyst upgrade, Lisa Su collision piece). Win-prob 68 is solid, risk_quality 75 strong for a mega-cap. Held position modestly in the green. This is the anchor of the AI infrastructure thesis — protect it. Concentration is the sizer's problem, not mine.