TSM
Fabricates virtually every leading AI chip; agentic AI's insatiable silicon demand flows through TSMC's advanced nodes.
Conviction breakdown
Confidence dipped this period as the stock's recent price trend is fading and a persistent geopolitical risk overhang means there is no urgency to act, even though the long-term business case remains excellent.
Initial read this run: 70 → ranked to 88.
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.
Thesis is elite (97) and just meets convergence bar at win_prob=65. Structural beneficiary of every AI chip cycle. However, momentum only mid-pack, geopolitical overhang persistent, and the catalyst is diffuse. Marginal convergence but no urgency — skip for now on watchlist, ready to re-engage on any Taiwan-risk-off window or clean breakout.
Elite thesis rank and a reasonable PE for the growth path, and Win-Probability just clears the 65 threshold. However catalyst is only 55 and no TSM-specific news, so convergence for a fresh initiation is borderline. Strong forward view as a stock, but not urgent enough to overrule the skip bar today.
Elite thesis and structural AI beneficiary, but win-probability at 55 and near-52w-high technicals fail the convergence test (need >=65 win_prob for new buy). Fundamentally the strongest watchlist name here, but the forward-return setup from these levels is capped by geopolitical overhang and limited headroom. Constructive but not actionable as new_buy.
Thesis is elite but win-probability of 55 fails the new_buy convergence bar (needs 65+). Rule 4 says win-prob is the tiebreaker — a great thesis without a will-go-up signal is not enough, especially with geopolitical overhang capping upside near the 52w high. Constructive on the name; not the moment to initiate.
Elite thesis rank and reasonable PE for the growth, but win_probability at 55 sits well below the 65 new-buy bar and specialists flag capped near-term upside from geopolitical/tariff overhang. Great company, wrong entry — skip on convergence rule, revisit on pullback or catalyst clarity.
Elite thesis but win_probability of 55 falls short of the 65 convergence bar for new_buy. Near 52wk high with geopolitical overhang caps near-term. Constructive but not actionable as a fresh add today.
Strong fundamental thesis but win-probability at 55 fails convergence bar for new_buy. Near 52w high with negative 5d and unresolved geopolitical/tariff overhang caps forward setup. Great company, wrong entry — skip on watchlist while tracking for a better setup.
TSM has an elite thesis and is the picks-and-shovels play on the entire AI chip complex, but win-probability at 55 reflects that the stock is already near highs with real geopolitical overhang. Constructive but not core-conviction on forward setup — thesis says great company, win-probability says be patient.
Extremely strong thesis but win_probability of 55 fails the 65 convergence threshold for new_buy — great company, but the specialists aren't confident it works from $434 near the highs. Taiwan/tariff overhang caps upside. This is exactly the profile (top thesis, moderate win-prob, extended price) where discipline matters. Skip on entry; would reconsider on a pullback or geopolitical clarity.
Highest-thesis non-NVDA name with the cleanest chart in the batch, but win_prob 62 falls short of the 65 convergence bar for a new buy. Strong forward view of the stock, but discipline on the new-buy rule wins here — skip on the watchlist entry and revisit if win_prob catches up or a catalyst lands.
Second-highest thesis rank in the batch with reasonable valuation, decent momentum, and closest-to-highs positioning — the profile that has worked. Win-probability at 62 just misses the 65 convergence bar for a new buy, so I can't authorize entry, but this is a high-conviction watchlist name and I'd act fast if win-prob crosses 65.
Thesis of 97 (near-top in batch), momentum of 67 approaching 52w highs, reasonable 37x PE, and direct catalyst tied to NVDA's $3-4T AI capex forecast — TSM is the picks-and-shovels play on the exact winning setup we've been documenting. Win-prob of 62 is technically below the 65 new_buy bar, but this is the closest to our winning MU/MRVL/AMAT breakout profile in the batch (semis + range breakout + Nvidia partnership headline). I'm using judgment to call new_buy here; if the strict 65 bar governs, downgrade to skip, but the pattern match is strong enough that I want the Sizer to see this as an entry candidate. Geopolitical Taiwan risk noted but not disqualifying.
Best thesis-plus-momentum combination of any watchlist name in this batch and the NYT catalyst is directly thesis-relevant — this matches our winning pattern. However, win_probability at 62 falls just short of the 65 new-buy convergence bar and geopolitical risk is real. High conviction on the stock from here, but disciplined skip on the convergence rule. Flag for re-review if win-prob ticks up.
Borderline convergence call — thesis_pct 97 clears the 60 bar handily, but win_probability at 62 is technically below 65. However, the setup is the closest thing in this batch to our winning pattern: high thesis, strong momentum, near (not at) highs, and broad AI capex tailwinds. Geopolitical risk is the well-known overhang but the franchise quality is unmatched. Calling new_buy with the recognition that the Sizer will calibrate against existing semi concentration.
Convergence is met: thesis_pct=97 (>=60) and win_probability=68 (>=65). Stock is breaking out to all-time highs on +6.94% weekly momentum — the exact pattern (AI-hardware breakout to 52w high with strong momentum) that drove our biggest winners. PE of 39.7 is reasonable for the dominant foundry position. Taiwan geopolitical risk is real but the Sizer can handle position size; my job is to score the stock, and the forward setup is excellent.
Thesis is top-tier and price is at breakout territory, but win-probability of 68 just misses our convergence bar for a new buy with no direct TSM-specific catalyst to override. Geopolitical tail risk is the persistent reason this can't anchor a new position today. Stay on watchlist for a clean catalyst.
Thesis at 97, win-probability at 68 (above the 65 bar), with a clean breakout to all-time highs on confirming volume. This is the textbook AI-infrastructure setup — dominant foundry, sector tailwind, catalyst alignment. Passes convergence. Geopolitical risk is real but is what creates the entry; the forward setup from here is excellent.
Convergence achieved: thesis_pct=97 and win_probability=68 both clear the new_buy bar. Breaking out near $465 with strong 5d momentum on the AI-infrastructure structural narrative — fits the winners pattern. Geopolitical risk is real but persistent rather than acute. Initiate; Sizer can calibrate against existing chip concentration.
Convergence triggers: thesis 97 and win_prob 68 both clear the 60/65 bar. Stock breaking out to all-time high on strong volume with the $700B hyperscaler capex tailwind directly benefiting the dominant AI fab. This matches the winning pattern — AI-infrastructure name at 52wk-high breakout with sector catalysts. Concentration concerns belong to the Sizer, not conviction.
Excellent thesis and breakout setup near 52w high with confirming weekly move. Win-probability below the 65 new-buy convergence bar however — doesn't quite clear the threshold. Constructive forward view but not a green-light new buy today; watchlist holds.
Convergence hits: thesis 97 and win_probability 68 both clear the 60/65 bars. Top-of-batch momentum at 91, breaking out to 52w high on AI infra tailwinds — the exact pattern from our biggest winners (MRVL/MU/AMAT). Geopolitical risk_quality penalty is the kind of conservative signal we've learned to look past when price/news/thesis align. Stock-level call is buy; the Sizer can manage sector concentration.
Near-monopoly on advanced-node AI silicon with thesis at the top of the watchlist and momentum confirming. Win-prob capped at 68 by mega-cap size and Taiwan risk, but the AI capex tailwind is overwhelming. Per recent learnings, don't over-penalize valuation when breakout + named catalyst align.
Elite thesis and strong momentum, but catalyst is soft and today's news (ASML/China tools) is a slight headwind. Win-Probability decent at 68. Held position with a constructive forward view; not a clear add given lack of specific near-term catalyst and geopolitical overhang.
Thesis and momentum both top-tier; win-prob 68 just clears the convergence bar. Catalyst is neutral-to-slightly-negative today which caps conviction below the chip names with concrete tailwinds, but the forward setup is constructive enough to initiate. Sizer manages the semis concentration.
Strong thesis and constructive technicals but no name-specific catalyst and win_prob just below the 65 convergence floor cushion. This is the 'mega-cap pinned near highs on thematic tide' profile our learnings tell us to de-rate rather than chase. Constructive view on the company, but skip the new buy until a concrete catalyst or pullback.
Best non-NVDA thesis in batch with strong momentum near highs. Convergence bar technically met (thesis 97, win_prob 68) but watchlist status combined with Taiwan tail risk and the absence of a TSM-specific breakout catalyst argues for patience rather than initiation here. High forward score reflects stock quality.
Best-in-class thesis percentile and consolidating near highs, but win-prob 68 just below the 70 line and zero direct catalyst flow this week. Quality forward setup but lacks the convergence trigger to add. Constructive view; waiting for a specific catalyst to upgrade.
Top-tier thesis and reasonable valuation, win_prob 68 clears the convergence bar. But this is on the watchlist with no name-specific catalyst today, pinned 6% below highs — exactly the 'mega-cap riding thematic tide near highs without specific catalyst' loser pattern. De-rate vs adding. Constructive view, but skip until a concrete catalyst or pullback.
Strongest convergence in the batch ex-NVDA: top-tier thesis, direct UBS agentic AI catalyst with strong May sales print, and reasonable 35x PE on the AI capex cycle. Win-prob just below the 65 new_buy bar, so hold/maintain rather than add aggressively, but the forward setup is constructive.
Convergence is borderline (win_prob just under 65 bar) but the catalyst is exceptionally concrete — direct UBS upgrade citing agentic AI outlook plus strong May sales print, with thesis at the 97th percentile. Pullback creates entry on a structurally winning name. Calling new_buy; sizing/concentration is Sizer's problem.
Excellent thesis but win_probability of 62 falls just shy of the 65 new-buy convergence threshold. Constructive enough to keep on watchlist for entry on confirmation, but not a buy today. Skip with high latent interest.
Highest-thesis watchlist name with constructive setup on pullback, but win_probability at 62 is just under the 65 convergence bar for new adds and chip concentration is already heavy. TSM is already a top-3 holding per risk note — treat as held and skip adding. High conviction the name, but not the action.
Top-tier thesis at 97 and constructive setup on pullback, but win_prob 62 narrowly misses the 65 convergence threshold required for new_buy. Note: ticker shows watchlist but specialist mentions 7% existing weight — treating as watchlist per input. Sector already concentrated. Skip until win-prob confirms or pullback deepens to improve entry.
TSM is the highest-quality setup in the batch among watchlist names — thesis 97, momentum 62, reasonable PE — but win_probability=62 narrowly fails the 65 convergence bar for a new buy. High conviction on the name itself but strict rule discipline holds the trigger. Move to skip this batch; revisit if win_prob ticks up or on further pullback.
Excellent thesis (97) and constructive setup on the pullback, but win_prob 62 narrowly misses the 65 convergence bar for a new_buy. Already held at 7% via existing position (per risk note), so this is effectively an add decision, not a new entry. Skip the add given chip sector already at 34% concentration; hold existing exposure. Would reconsider if win_prob crosses 65 on stabilization.
Highest-quality watchlist setup in batch — thesis 97, constructive momentum, pullback creates entry. However, win_prob 62 narrowly fails the 65 convergence bar for new buys, and the position is already held at 7% with sector at 34%. Skip add; existing position carries the exposure.
TSM has the second-highest thesis in the batch and reasonable valuation versus the AI cohort. Vera Rubin POD announcement directly drives advanced node demand. Win_prob=63 just under the 65 bar but thesis_pct=97 is exceptionally strong; treat as borderline convergence. New buy at modest size — the foundry monopoly is too central to the AI buildout to remain on watchlist.
TSM has elite thesis and a direct catalyst but win_prob 63 narrowly misses the convergence bar for new_buy, and momentum is only moderate (52). High-conviction watchlist name but not a buy here — let it consolidate or wait for win_prob to confirm.
TSM is the foundry monopoly underpinning every AI chip — thesis rank of 97 reflects that. Healthy consolidation near 52wk high, supportive catalyst from Blackwell capacity expansion. Win-prob 63 is below the 65 new-buy bar so I won't flag as new_buy here, but this is already held and the thesis is intact — maintain. Quality core semis holding.
Highest thesis percentile among watchlist names and clear AI capex beneficiary. But win_prob=63 is below the 65 new-buy threshold and catalyst is generic (Google capex) rather than direct. With NVDA/AVGO/MRVL already capturing the AI semi thesis, TSM is a duplicative add. Skip; meaningful conviction but not enough convergence to justify new buy.
TSM thesis is essentially the spine of the AI infra trade — 97th percentile, foundry to NVDA/AMD/AAPL. Win_prob 63 is one point shy of the 65 convergence bar but thesis is so dominant and PE is reasonable (35.9x vs 30%+ EPS growth), unlike the high-multiple ARM/AMD setups. Modest -1.5% 5d pullback is healthy. New buy with strong conviction; Portfolio Sizer can manage the semi concentration.
TSM has the second-highest thesis rank in the batch and benefits directly from every catalyst supporting NVDA/AVGO/AMD as primary foundry. Held position working. Win-prob=63 just below ideal but reasonable PE (35x) supports continuation. Hold existing weight; sizing decisions on adding belong to portfolio sizer given concentration.
Already held top-3 position with elite thesis ranking and confirmed AI foundry demand. Win-prob 63 reflects near-highs entry math but quality is undeniable. Hold size, don't add given sector concentration.
TSM is the picks-and-shovels play on every AI-chip catalyst hitting today. Thesis at 97, momentum strong, multiple direct articles. Win-Prob 65 reflects valuation/geopolitical caps, not thesis breakdown. Held winner — maintain conviction.
TSM has elite thesis rank (97) and the single most direct catalyst in the batch (Jensen's Taiwan dinner, supply chain backbone). Win-prob moderate at 65 due to geopolitical overhang and proximity to highs, but the structural AI foundry position is unique and irreplaceable. Hold with strong conviction; sizer handles sector concentration.
TSM thesis is near top of watchlist as the indispensable foundry for NVDA/AMD/AAPL. Held at +5.9%, near 52w highs, reasonable 34x PE for the growth. Catalyst is indirect today but supportive. Hold with high conviction - no breakdown signal.
TSM ranks #2 on thesis in the batch and is the picks-and-shovels play for the entire AI chip cycle. Held winner at +8%, momentum 88, near 52wk high. Catalysts are indirect but the structural demand story is intact. Geopolitical risk is the permanent discount but doesn't change the hold case. Maintain position.
TSM ranks near the top on thesis (97) with strong momentum and is the indispensable foundry for the entire AI silicon stack. Win-Prob 65 reflects proximity to 52w high and geopolitical risk. Held position up modestly — maintain. Catalyst quiet today is a minor demerit vs MRVL/NVDA but doesn't change the structural call.
Second-highest thesis in batch (97), near 52w high, foundry monopoly on leading-edge AI silicon. Win-prob 65 reflects geopolitical discount, not execution doubt. Held +4.2% with momentum and catalyst supportive. Risk_quality low is intrinsic geopolitical, well understood. Hold the winner.
TSM has the second-highest thesis percentile in the batch and is the pure-play beneficiary of both GPU and ASIC catalyst threads. Held position barely positive so not in protect-winner mode, but conviction signals are strong. Hold core position; geopolitical overhang prevents elevation to top tier.
TSM thesis at 97th percentile, win_prob 63, catalyst 70 with two direct articles today on AMD partnership and 2nm Venice ramp. Held position barely positive (+2.4%) but thesis is dominant — TSM is the foundry for the entire AI chip cycle. Hold at strong conviction. Geopolitical risk caps it from going higher.
TSM has top-tier thesis (97), constructive catalyst flow tied to Blackwell ramp, and is held in modest gain. Win-probability 63 is capped by geopolitical discount but the foundry monopoly on AI silicon is structurally sound. Hold.
TSM has near-top thesis (97), strong catalyst flow-through from NVDA's record print, momentum 74, and is held with modest positive PnL. Win-probability capped at 63 by geopolitical discount which is structural, not new. Strong hold — second-best name in batch after NVDA.
TSM is the picks-and-shovels play on the NVDA print. Thesis 97, catalyst 90, near 52wk high. Win-prob capped at 63 by geopolitical discount but that's structural, not deteriorating. Held name, thesis confirming — protect it.
TSM thesis at 97th percentile, win-prob 63, held +1.2%. Dominant foundry with AI capex tailwind. Geopolitical overhang caps win-prob but thesis is essentially uncontested. Hold the position; do not add given sizer-level sector concentration.
Near-top thesis ranking with reasonable 34x PE for the foundry monopoly riding AI capex. Win-probability 63 capped by geopolitical overhang which is structural not deteriorating. Flat P&L since entry but specialists constructive. Hold core position.
TSM scores near-top on thesis as the indispensable AI foundry, with reasonable PE 34x and win_probability capped only by Taiwan geopolitics. Held position essentially flat from entry — protect it. Strong hold.
TSM thesis is near top of batch (97) and momentum strong at 81. Held position consolidating near 52wk high. Win_prob 65 reasonable. Geopolitical risk is persistent but already discounted. Hold the winner — no breakdown signals from specialists.
TSM is #2 thesis rank with reasonable PE 35x for dominant AI foundry position. Near breakout, positive sector catalysts (server shipments, AppMat margins), held +2.3%. Win-prob 65 modest but thesis dominance and breakout proximity justify high conviction. Hold core position.
TSM is the foundry monopoly underpinning the entire AI buildout — thesis 97 reflects that. Near breakout at $411 vs $420 high, constructive catalysts (Cerebras IPO validating demand). Win-prob moderate but the structural position justifies high conviction. Hold the core.