28 companies

One molecule, one trial, one quarter. HLP003 (deuterated psilocin) for adjunctive MDD reads out topline from APPROACH in Q4 2026; $166.4m of cash at 2026-06-30 funds the company past it with ~4.5 quarters of runway. The company's own Phase 2 in GAD is the warning: its sub-therapeutic 2mg control arm produced a 30% responder rate against 59% at 20mg, and APPROACH uses inactive placebo with a drug that produces 90 minutes of unmistakable acute effect. How the placebo arm behaves is the thesis.

Price $13.71→Target $28.00the APPROACH Phase 3 readout, Q4 2026

What would prove it wrong

APPROACH misses its primary endpoint, or hits on a margin the placebo arm's own response makes uninterpretable. +4 more

Watch for

APPROACH topline efficacy, Q4 2026, and specifically the placebo arm's own MADRS movement; the HLP004 next-study design, committed by end of Q3 2026; whether the unused $100m ATM is drawn at $13.71 rather than at June's $4.85; confirmation of the apparent shift from foreign-private-issuer to domestic reporting (first-ever S-8, Form 3s, Form 4s and 13Ds all in September 2026); a permanent CEO appointment - the company is interim-led into the readout.

Priced at ~12x management's FY28 EPS target: the market is discounting the AI revenue guides ($115bn FY27, $230bn FY28 AI). If the custom-accelerator schedule is delivered, the earnings arrive and the discount closes.

Price $364.54→Target $540FY2028 (Oct 2028)

What would prove it wrong

FY2027 AI semiconductor revenue below ~$115bn. +4 more

Watch for

Quarterly AI revenue against the $115bn FY27 guide; Singapore substrate capacity unlocking in FY27; customer-concentration disclosure (one distributor at 50% of Q3 FY26 revenue) not rising further.

13.4x LTM EPS and 9.2x EV/EBIT Adjusted (ex-leases) for a group whose two brands swapped roles: Miu Miu went from +34.8% (FY2025) to +2.5% (H1 2026) while the Prada brand re-accelerated to +6.3% in Q2 2026 on full-price like-for-like. Versace (USD 1,395m, closed 2025-12-02) is guided to flat losses in 2026 and took the H1 margin from 22.6% to 17.4%; the variant is that this is the trough, on a flat 2,558.8m share count and 0.43x pre-lease leverage. Governance discount is real: 80% family control, float at the 20% waiver minimum, a steady flow of real-estate deals with the founders' vehicles.

Price $37.38→Target $49.90FY2028 (Dec 2028)

What would prove it wrong

Prada-brand retail below 0% at constant FX in any reported half or quarter (H1-26 +3.3%, Q2 +6.3%). +4 more

Watch for

Q3 2026 revenue update ~late October 2026 (DERIVED from cadence: 2025-10-23) — Miu Miu's quarter first (forward growth dodged on the Q1 and Q2 calls), then Prada-brand growth against Q2's +6.3%; FY2026 results ~early March 2027 with Versace's FY2026 loss against 'not dissimilar' guidance; Pieter Mulier's first Versace collection, May 2027; any Milan prosecutors' action on the 2025-12-04 labour-exploitation document request naming Prada and Versace; any further connected-transaction property deal with a family vehicle.

0.76x its own book including AOCI and ~4.7x annualised H1 2026 adjusted operating EPS; the $23.71-a-share gap to book ex-AOCI is an $8.5bn unrealised bond loss that accretes back at par. The base case is LNC getting back to its own book, not a re-rating. The constraint nobody models: Bain's $44.00 floor sits above the market price and encumbers four-fifths of the $714m buyback authority.

Price $41.01→Target $53.002028 model line

What would prove it wrong

Reinsurance recoverable net of collateral above $8bn, or the collateralised share below 75%. +7 more

Watch for

Q3 results ~2026-10-29 — the annual assumption review (charged net income in two of the last three years) and the leverage ratio against the 25% target now that $437m of preferred left the stack and $500m of 2056 subordinated notes stayed in it; any Schedule 13D/A showing LNC buying Bain's stock at $44; the permanent CFO hire; the 2026-09-15 short-interest settlement published 2026-09-24 after +33% at 2026-08-31.

On management's own numbers it trades at ~14x FY28 EPS — the dossier calls it cheap, and says that is the tell: the market is pricing a real chance the FY28 guide misses because a rising share of demand is underwritten by NVIDIA itself.

Price $228.87→Target $295FY2028 (Jan 2028)

What would prove it wrong

Supply and capacity commitments fall. +5 more

Watch for

FY28 ~70% growth guide held or raised; supply commitments ($279bn) not falling; gross margin holding the 71-72% Q4 trough then 72-73%; receivables growth slowing.

Organic net new ARR accelerated 7% → 10% → 17% through FY26 while the FY27 guide embeds ~4%; if Security for AI and Agentic SOC land, 17% is a floor and the stock re-rates from ~7.7x EV/FY27 revenue.

Price $210.18→Target $260FY2029 (Jul 2029) — beyond two years

What would prove it wrong

Organic net new ARR growth below 10% in any two consecutive quarters +5 more

Watch for

Investor Day 2026-10-06 (FY28-30 targets, non-seat ACV mix); Q1 FY27 results ~2026-11-24; net new ARR and NRR (kill below 112%).

The only P&C name in the cluster at or below book (P/B 0.97). The base case is AIG keeps trading at book while earning 11-12% on it; the Corebridge exit is done and underwriting has started growing again.

Price $75.18→Target $93.002028 model line

What would prove it wrong

A third senior departure at Executive Chair / segment-CEO level before the Q3 2026 call, or a General Insurance CEO hired from outside. +7 more

Watch for

Q3 results ~2026-11-03; buyback pace ($2.6bn authority left) against the premium trend; Item 5.02 8-Ks after two senior departures in fifteen days.

10x forward earnings for a bank compounding book value; the acquired Capital One IRA deposits are cheap funding if they arrived, and NIM above 4.6% carries the bull case.

Price $89.13→Target $110FY2028 (Jun 2028)

What would prove it wrong

Net charge-offs to average loans exceed 0.50% for two consecutive quarters. +7 more

Watch for

Fiscal Q1 2027 10-Q (~2026-10-29): the deposit table (did the ~$1.3bn of IRA balances transfer?) and the purchased-loan marks; efficiency ratio.

~20x FY28 model EPS for the connector and cable supplier inside every AI rack; the bull case is AI content per rack compounding, 22% growth holding and a 21.5% margin.

Price $82.84→Target $98.00FY2028

What would prove it wrong

Book-to-bill falls below 1.0. +5 more

Watch for

Quarterly IT Datacom growth; the FY2026 10-K re-test of whether acquisitions are 'immaterial in the aggregate' (the SEC asked); buyback nearly exhausted ($411m left).

At 1.00x book HHH is no longer a pure land company: since 2026-06-04 it owns a five-year-old Bermuda specialty insurer (40% of assets) of which a Pershing Square fund's $1.0bn preferred takes up to 49%, whose float Pershing Square now manages without an extra fee and moved within weeks from bonds to a concentrated stock book (32% at 2026-06-30, ~40% by August, target >50%). S&P cut the insurer to BBB on the change of parent [search-summary]. The land bank (MPC EBT $349m → $476m, 2024 → 2025) is being liquidated on purpose to fund it; the base case is book growth, not a re-rating.

Price $66.09→Target $78.00mid-2028 (Q2 2028 book)

What would prove it wrong

Adverse prior-year development for Vantage in the FY2026 10-K loss-development disclosure. +4 more

Watch for

Annual meeting 2026-09-30; Q3 results ~2026-11-05 to -11-10 (DERIVED from cadence) — Vantage capital contributions and the equity sleeve against $1.55bn of net reserves; FY2026 10-K (Feb–Mar 2027) with Vantage's first loss-development table; any A.M. Best action; any takedown of the 2026-08-14 universal shelf; trailing MPC EBT against the 2024 $349.1m.

A levered equity on a backlog: $103.7bn of RPO and 4.2GW contracted against $35.55bn of debt on $5.02bn of book equity, with $10.6bn of principal due inside eighteen months at effective rates to 15%. The September package is the evidence that matters — $4.2bn of 2.875% converts plus a 35m-share ATM, against an existing stack paying 8.5-9.75%. A market writing 2.875% paper is not pricing the bear case.

Price $86.76→Target $97.00FY2028 (Dec 2028)

What would prove it wrong

RPO stops growing, or its front end stretches. +6 more

Watch for

Q3 2026 results, early November — whether the 5-10 point contribution-margin uplift management refused to decompose shows up in reported margin; RPO and its 41% 24-month share; the debt table's next-twelve-months column after the convert; any disclosed delay on a contracted gigawatt; whether the material weaknesses are still open in the FY2026 10-K.

8.6x 2026E earnings with book compounding and shrinking the share count into a soft market; underwriting remains excellent on a current-accident-year basis.

Price $94.95→Target $1052028 model line

What would prove it wrong

Long-tailed prior-year development exceeds $75m adverse in a single quarter, or exceeds the short-tailed release in any quarter. +6 more

Watch for

Q3 2026 10-Q (late Oct): Note 5 prior-year development by tail length — long-tail adverse development was $32m in Q2; buyback pace ($1.95bn in H1).

7.4x normalised 2028 EPS on peak-cycle personal-auto profitability; consensus Hold with targets ~$267-300. The bear column ($148) is larger than the upside — asymmetric the wrong way.

Price $229.50→Target $2492028 (normalised)

What would prove it wrong

Prior-year reserve development turns adverse in any quarter, or releases fall below 1.5 points of the combined ratio for two consecutive quarters. +6 more

Watch for

September monthly catastrophe release ~2026-10-15 (July-August already $1.43bn vs $558m for all of Q3 2025); Q3 results late Oct/early Nov; buyback pace.

Durable pieces (first-party cloud storage +30%, Keystone +65%, 93%-margin support) compound under a cyclical flash-price print; 100% of FCF returned. The stock has already doubled in six months.

Price $192.91→Target $209April 2029 (FY29) — beyond two years

What would prove it wrong

Product gross margin below 50% (non-GAAP) in two consecutive quarters +4 more

Watch for

INSIGHT 2026-09-29; Q2 FY27 call 2026-12-01 (product GM vs 54-55%, inventory, purchase commitments); NAND contract prices; June 2027 note refinancing.

Best-in-class E&S underwriter (57% loss ratio) at 16x earnings; the bull case needs the 32.7% Commercial Property decline to prove cyclical and ex-property growth to reaccelerate.

Price $351.83→Target $3812028

What would prove it wrong

The share-count half is dead. +7 more

Watch for

Q3 release ~2026-10-22: gross written premium ex-Commercial-Property and the current accident year ex-cat loss ratio; the first buyback in the company's history.

A concession with a tariff framework fixed to 2030 at ~14x earnings; the bull case is passenger traffic reverting to the 2023-25 6-8% growth as airline capacity returns.

Price $102.45→Target $1052027

What would prove it wrong

Passenger traffic goes negative for two consecutive quarters. +4 more

Watch for

Monthly traffic prints (the dossier's buy trigger: two at +4% y/y or better); the 2030 tariff framework staying unamended; the 2026-09-18 unexplained volume spike.

The EUV monopoly with a shrinking share count; at ~29x FY28 the base case is roughly priced in, and upside needs the second capacity step (+30% low-NA EUV, +30% more for 2028).

Price $1,747.90→Target $1,780FY2028

What would prove it wrong

The 2028 capacity step is dropped rather than confirmed. +6 more

Watch for

Quarterly orders and 2027 capacity commentary; whether the 'investigated' 2028 +30% is confirmed; TSMC capex discipline as the bubble indicator.

A frac company becoming a power producer (PROPWR): no return in the base case, a triple if 1.1GW is earning in 2028 — a position-sizing question, not a valuation one. The bear case impairs the equity.

Price $10.03→Target $10.002028

What would prove it wrong

Contracted PROPWR megawatts fall, or fail to reach 700 MW by the Q2 2027 release, on a reconciliation that shows gross additions AND capacity that came off contract. +4 more

Watch for

Q3 release (late Oct): the PROPWR contract book reconciled line by line (additions vs capacity coming off contract); EBITDA per MW, which management has not published.

A 14% yield at 1.06x book, and the yield is the whole base case. From 2021 to 2025 the company reported $12.71/share of EAD while a holder's economic result was about zero — the dividend was paid out of carry and the book absorbed the cycle — and the share count doubled. Since 2023 it has earned double-digit economic returns; the open question is whether that is the new resi-credit/MSR mix or the MBS basis tightening under the GSEs' $200bn purchase mandate.

Price $21.44→Target $21.20mid-2028 (Q2 2028 book)

What would prove it wrong

BVPS below $18.45 at any quarter-end (the Q2 2025 figure, the lowest quarter-end BVPS in the filings read). +3 more

Watch for

Q3 2026 results ~2026-10-21 (DERIVED from cadence) — book value per share after a 7.9% fall in the stock since 2026-08-20; EAD against the raised $0.75 dividend; funding of the $442.5m Series I redemption on 2026-10-01; the current-coupon MBS basis (94bp at 2026-06-30) and the GSE mandate's remaining capacity (~$45bn of $200bn settled through May).

Base lands essentially ON the traded price, so the multiple is not the argument — the share count is. Oracle exhausted a $20bn ATM inside one quarter (141m shares, +5.0%), bought back nothing, and funded $28.5bn of quarterly capex partly with $11.4bn of customer prepayments its own accounting calls a significant financing component. Every further $20bn raise at ~$150 is another ~4.4% of the company. The CFO declined to give a free-cash-flow date.

Price $149.20→Target $142FY2028 (May 2028)

What would prove it wrong

Remaining performance obligations stop growing, or their conversion schedule stretches. +6 more

Watch for

October 2026 Investor Day — the CFO deferred BOTH the gross-margin and free-cash-flow questions to it; Q2 FY27 results 2026-12-14; the Note 1 prepayment line ($11.4bn vs $0 a year earlier) and RPO ($664bn, 13% inside twelve months); a second ATM authorisation; the Ellison 10b5-1 plan for up to 50m shares running to 2026-10-24.

The base case is below the price; the upside is the data-centre option — the $5bn FY27 / $15bn FY29 targets are not yet visible in a reported revenue line. Bull: merchant-accelerator share after the mid-2027 launch and a Meta CPU ramp.

Price $198.27→Target $175FY2028

What would prove it wrong

December-quarter data-centre revenue does not appear. +7 more

Watch for

Q4 FY26 results 2026-11-04; the FY2026 10-K's accounting for the Amazon warrant; first reported data-centre revenue by early 2027.

~62x EV/LTM revenue with no earnings; the base case models below the price and the bull case needs AWS to scale as the term sheet contemplates and customer concentration to broaden.

Price $212.41→Target $177FY2028

What would prove it wrong

The remaining performance obligation stops growing. +5 more

Watch for

The Q2 2026 prepared remarks (only a UBS restatement of 'revenue up more than 3x next year' is in hand); 36.1m customer-warrant shares vesting; first S-3 eligibility.

The expensive one: ~41x FY27 EPS even if Data Center doubles in a year. The price is the option on the second-source thesis, and the OpenAI/Meta warrants (320m shares at $0.01) dilute holders if it works.

Price $623.77→Target $518FY2027 (Dec 2027)

What would prove it wrong

Data Center revenue fails to more than double in 2027. +5 more

Watch for

Q3 results early Nov; Helios and MI500 ramp; warrant vesting in each 10-Q; the '>$20 EPS' target timeline.

The base case is below the price even on management's 60%/60% data-centre growth guides; the bull ($259) is roughly today's price. SBC and the NVIDIA preferred (21.8m shares at ~$92) dilute.

Price $262.36→Target $203FY2028 (Jan 2028)

What would prove it wrong

Distributor A's share of revenue keeps climbing. +5 more

Watch for

Data-centre growth vs the 60% FY27/FY28 guides; Q3 FY27 non-GAAP EPS $1.05-1.15 guide; custom-silicon program wins.

The only cash-generative name in the ai-capex cluster, deliberately trading that status for AI growth. Q2 2026: revenue +29%, gross margin 59.9% to 55.0%, operating income -18%, diluted shares +25.8%, diluted EPS $0.39 to $0.29. At ~104x guided FY2026 non-GAAP EPS the market pays for the transition arriving; the base case here does not reach the current price, and the CFO has twice declined to give free cash flow including lease payments.

Price $146.00→Target $107FY2028

What would prove it wrong

Gross margin keeps falling. +4 more

Watch for

One quarter with gross margin flat or up sequentially, operating income growing year over year and a flat diluted share count; an all-in free-cash-flow guide (asked for in two consecutive quarters, not given); the adjusted-FCF margin guide not cut a third time; the 2027 revenue guide formalised above 50%; Access Industries (15.86%, Schedule 13D) not filing a sale.

The fastest capital-to-megawatt converter on the list, financed by customers ($5,975.2m of deferred revenue), convertible buyers (carried at $8.5bn, fair value $20.8bn) and NVIDIA ($2.0bn for warrants at $0.0001) rather than by earnings. At ~51x this year's guided adjusted EBITDA the market already pays for the base case; the bear column is what the same model gives if 2027 capacity breaks pricing before the 2026 capex is depreciated.

Price $236.12→Target $160FY2028

What would prove it wrong

The prepayment engine stalls. +4 more

Watch for

Deferred revenue holding or rising quarter-on-quarter; the FY2026 ARR guide ($7-9bn) reaffirmed rather than cut; the first quarter in which 2026 capacity earns revenue at a disclosed price per megawatt; the fixed-asset material weakness (controls over depreciation start dates) closed in the FY2026 20-F; no new convert above 5%.

The only name in this slot whose base case sits well BELOW the traded price. ~$655bn of market value after the August equity sale, on FY2025 revenue of $52.9bn and no net income, is paying for an external foundry anchor customer that management would not name when asked directly — three weeks before selling $23bn of stock. Reported GAAP EPS is currently inversely geared to the share price through the $15.6bn Escrowed Shares derivative, so the headline loss and the price move together for reasons unrelated to the business.

Price $123.86→Target $71.00FY2028 (Dec 2028)

What would prove it wrong

Intel Foundry's external revenue does not leave the hundreds of millions. +6 more

Watch for

Intel Foundry third-party revenue in Note 2 of each 10-Q ($293m in Q2 2026 vs $22m a year earlier — the direct measure of the thesis); 14A PDK 0.9 confirmed complete at the October Q3 call, risk production H2 2027, volume 2028; whether Q3 carries another multi-billion Escrowed Shares mark; gross margin staying 'comfortably into the 40s'; a third capital raise before an anchor customer signs.

~100x+ FY28 earnings; 74% of FY26 revenue growth came from related parties (SoftBank and Arm China) while external revenue grew 7%. Upside needs the AGI CPU and compute-subsystem royalties to inflect.

Price $333.20→Target $169FY2028 (Mar 2028)

What would prove it wrong

Related-party revenue stays above ~30% of the total, or external growth falls back below 10%. +5 more

Watch for

Royalty growth vs the guided 13%; related-party share of revenue; compute-subsystem adoption; SoftBank's 86.4% as an overhang.