Strong dislocation, clears the index hurdle. 11.5× forward P/E; 8.5× screening P/FCF.
Valuation edition · September 2026
Research byMichaël HillaertSubstack ↗
Research byMichaël HillaertSubstack ↗TWO SEPARATE QUESTIONS
View rankingsValuation × 30-year durability
Select a company to read the case.
| 30-year durability | Strong | Probable | Conditional | No current dislocation |
|---|---|---|---|---|
| A+ | — | — | — | |
| A | — | — | ||
| B+ | ||||
| B | — |
Swipe across to compare all four valuation signals.
- A+
- Unusually high confidence: the compounding engine is particularly difficult to see breaking down.
- A
- Very high confidence, with somewhat more long-term uncertainty than A+.
- B+
- A strong chance of remaining a good compounder, with meaningful structural risks.
- B
- A credible 30-year case, with considerably more uncertainty around technology, competition, cyclicality, concentration or reinvestment.
Intuit and RELX currently combine a strong dislocation signal with B+ durability. Microsoft ranks higher on durability, but I do not currently see a valuation dislocation. A higher durability tier alone does not make a stock a better investment at today’s price.
I intentionally exclude C-tier companies from this matrix: I do not have enough confidence in their 30-year durability. Archived, likely-false and untiered names are also outside this view.
Signals and dates
Durability tiers aligned with the article published 12 October 2026. Shadow, Nvidia and Broadcom dislocation ratings updated 8 October 2026. Valuations and historical company assessments retain the 30 September 2026 snapshot; the matrix uses the updated ratings. Company order within a cell is not a ranking.
Start with the ranked viewBegin with the ten strongest valuation dislocations, or explore the 35-company valuation universe.
TOP TEN · VALUATION DISLOCATION
Ranked valuation dislocations
Shadow Ranking means valuation candidates outside the Top 10. Likely false means apparent cheapness may rely on unsustainable assumptions. Archived means retained for reference but not currently assessed.
Coverage groups are selected above. Signal and hurdle filters apply only to the Top 10 and never substitute for Shadow Ranking, Likely false or Archived. Signal: Conditional dislocation ≠ Hurdle: Conditional.
CompanyCurrent status Signal measures apparent mispricing. Hurdle asks whether the stock currently looks preferable to the S&P 500.Valuation Metrics differ by company and are screening references, not directly comparable scores.Monitoring / reasonValuation absorption / date How quickly and reliably conservative earnings or free cash flow make today’s valuation attractive. 1 = slow or uncertain; 5 = fast and reliable. This is not a business-quality score.
Ranked valuation dislocationsThe Top 10 receives full rank, signal, hurdle, valuation and monitoring detail.
Strong dislocationHurdle · Clears11.5× forward P/E; 8.5× screening P/FCF5/5
Strong dislocationHurdle · Clears16.7× forward P/E; 15.3× screening P/FCF4/5
Probable dislocationHurdle · Clears8.9× forward P/E; 8.8× screening P/FCF4/5
Probable dislocationHurdle · Clears16.1× forward P/E; 15.4× screening P/FCF4/5
Probable dislocationHurdle · Clears16.6× forward P/E; 13.5× screening P/FCF4/5
Conditional dislocationHurdle · Conditional15.7× forward P/E; 15.6× screening P/FCF4/5
Conditional dislocationHurdle · Clears19.0× forward P/E; 43.4× screening P/FCF4/5
Conditional dislocationHurdle · Conditional20.3× forward P/E; 42.6× screening P/FCF3/5
Probable dislocationHurdle · Clears22.6× forward P/E; 45.1× screening P/FCF3/5
Strong dislocationHurdle · Conditional12.3× forward P/E; 20.0× screening P/FCF4/5
September changes — 20 companies updatedBriefing, Top 10 comparison and company-by-company researchShow detailsHide details
Intuit remains #1. Autodesk rises to #4. Broadcom enters at #8.
Autodesk rises 4 places from 31 August 2026. Probable dislocation; Clears hurdle.
Apparent cheapness is excluded where peak-cycle earnings, poor cash conversion, financing or balance-sheet risk dominate the headline multiple.
September 2026 edition. Operating evidence and prices through 30 September 2026. Provider valuation denominators observed on the first post-cutoff retrieval.
Top 10 rank comparisonPrevious edition: 31 August 2026 · Current edition: 30 September 2026. Shadow Ranking is non-ordinal.
CompanyPrevious editionCurrentMove
Intuit#1#1Unchanged
RELX#2#2Unchanged
Adobe#3#3Unchanged
Autodesk#8#4Up 4
Experian#6#5Up 1
Constellation Software#5#6Down 1
NVIDIA#7#7Unchanged
BroadcomOutside Top 10#8Entered Top 10
Meta Platforms#4#9Down 5
Tencent#9#10Down 1
Microsoft#10Shadow RankingExited Top 10
Published field changesChanges in evidence and assessment since the previous monthly edition.
IntuitLatest evidence / Assessment / Monitoring triggers
Latest evidence: Investor day: midmarket revenue +39% and customers +28%; IES reached nearly $150m annualised revenue, about 4× year on year. Assessment: Enterprise Suite upgrades, payroll and payments adoption could deepen Intuit’s role in customers’ finances. The Credit Karma tax pilot matters if it brings incremental paying customers rather than shifting existing users. Customer growth and recurring cash conversion remain the tests. Monitoring triggers: Upgrade if Enterprise Suite, payroll, payments and the Credit Karma tax pilot bring incremental revenue and stronger retention.30 September 2026. Compared with the previous monthly edition.
RELXLatest evidence
Latest evidence: September's investor presentation reiterates H1 underlying revenue growth of 7%, adjusted operating profit growth of 9% and adjusted EPS growth of 11% at constant currency.30 September 2026. Compared with the previous monthly edition.
AdobeLatest evidence
Latest evidence: Q3 revenue rose 13% to $6.76bn; ARR reached $27.50bn and AI-first ARR grew over 150% year on year.30 September 2026. Compared with the previous monthly edition.
ExperianLatest evidence
Latest evidence: Q1 organic revenue grew 7%; the September screen is about 16.6× forward earnings and 13.5× screening FCF.30 September 2026. Compared with the previous monthly edition.
NVIDIALatest evidence / Assessment / Monitoring triggers
Latest evidence: Q2 revenue grew 106% to $96.2bn; Q3 guidance is $108bn. Assessment: Strong growth supports the earnings valuation. Buybacks matter when they reduce diluted shares and are funded by recurring cash. GPU-backed financing links buyers and data centres to lenders and insurers: collateral values, refinancing and credit losses could feed back into demand. Review the hardware, software and networking moat in October–November. Monitoring triggers: Upgrade if growth becomes recurring cash and customer demand needs less financing support.30 September 2026. Compared with the previous monthly edition.
BroadcomCoverage / evidence
Q3 revenue +86%, AI semiconductor revenue +221%, and FCF $13.7bn. Strong growth supports the case, but cash valuation, customer concentration and contingent financing limit conviction. Underlying AI concentration may be higher than segment labels suggest, as growth, engineering effort and capital become tied to custom AI infrastructure. Review durability in October–November.30 September 2026. Newly covered company.
Meta PlatformsAssessment / Monitoring triggers
Assessment: AI supports ads and engagement. WhatsApp, Instagram and AI agents could extend Meta into commerce. Cheaper inference may shift value toward distribution, proprietary data, customer relationships and workflow integration. Higher valuation and capex still limit the case. Monitoring triggers: Upgrade if ads stay strong as capex growth slows, or WhatsApp/Instagram/AI-agent commerce produces measurable revenue and cash returns.30 September 2026. Compared with the previous monthly edition.
Novo NordiskLatest evidence / Assessment
Latest evidence: H1 adjusted sales and operating profit both grew 2% at constant exchange rates. Assessment: The lower valuation and improved H1 outlook make the recovery case more interesting.30 September 2026. Compared with the previous monthly edition.
ServiceNowLatest evidence
Latest evidence: Q2 subscription revenue rose 24.5%.30 September 2026. Compared with the previous monthly edition.
SynopsysLatest evidence
Latest evidence: Q3 revenue was $2.477bn; GAAP EPS $2.84 versus $3.91 adjusted.30 September 2026. Compared with the previous monthly edition.
Topicus.comLatest evidence
Latest evidence: Q2 revenue grew 18% to €437.3m, including 4% organic growth.30 September 2026. Compared with the previous monthly edition.
TransDigmLatest evidence / Assessment
Latest evidence: Q3 sales rose 23%, including 13% organic growth; adjusted EPS was $10.87. Assessment: Aftermarket economics remain attractive, and the current price is about 26.4× the new FY2026 guided adjusted-EPS midpoint.30 September 2026. Compared with the previous monthly edition.
TSMCLatest evidence
Latest evidence: August consolidated revenue was NT$514.8bn, up 53.3% year on year; January–August revenue rose 39.3%.30 September 2026. Compared with the previous monthly edition.
Micron TechnologyLatest evidence / Assessment
Latest evidence: Q4 FY2026 revenue reached $54.23bn; adjusted EPS was $33.42. Assessment: Cash generation is substantially stronger than the August evidence.30 September 2026. Compared with the previous monthly edition.
OracleLatest evidence / Assessment
Latest evidence: Q1 FY2027 revenue rose 30% to $19.35bn; OCI grew 121% and RPO reached $664bn. Assessment: Demand and customer funding improved, but headline operating cash flow includes substantial financing-related prepayments.30 September 2026. Compared with the previous monthly edition.
SK hynixLatest evidence
Latest evidence: Q2 revenue reached KRW79.3tn (+257%), operating profit KRW60.5tn (+557%) and margin 76%.30 September 2026. Compared with the previous monthly edition.
Super Micro ComputerLatest evidence
Latest evidence: Q4 FY2026 sales were $11.1bn; GAAP gross margin recovered to 17.5%, versus 9.9% in Q3.30 September 2026. Compared with the previous monthly edition.
Cadence Design SystemsAssessment
Assessment: Cadence is benefiting from structurally rising chip- and system-design complexity.30 September 2026. Compared with the previous monthly edition.
Moody'sLatest evidence
Latest evidence: Q2 revenue rose 15%; the 25.6× forward screen still requires durable issuance and Analytics growth.30 September 2026. Compared with the previous monthly edition.
MSCILatest evidence
Latest evidence: Q2 revenue rose 12.2%; the 25.5× forward screen still assumes strong index flows and retention.30 September 2026. Compared with the previous monthly edition.
Tier and thesis treatment. Top 10 durability tiers match the finalized September workbook and remain separate from the valuation ranking. Changes above compare the two monthly editions; unchanged fields are omitted.
BEHIND THE DASHBOARD
All writing on SubstackLatest research
A 30-Year Durability Ranking of 35 Companies
Published 12 October 2026. The full durability universe, confidence tiers and the hardest borderline calls, with current valuation excluded.
Why Semiconductor Bottlenecks Move
How constraints shift from compute to memory, packaging, power and the capital needed to fund it all.
Valuation Dislocation — September 2026
The monthly research behind this edition’s valuation assessments and ranking changes.
Administrative Friction
A broader look at bottlenecks: what moving within Europe reveals about the institutions behind free movement.
OUTSIDE TOP 10
Research pipeline
MANI · Hamamatsu Photonics · Nakanishi · Salesforce · Hermès · Temairazu · Veeva · Rightmove · NICE Information Service · Moncler · Fairfax · Wolters Kluwer · LSEG
Companies I am considering for future valuation coverage, outside the valuation universe. Some already have a 30-year durability tier. A new Top 10 entry requires new earnings or a material change in the investment case. A lower price alone is not enough.
Strategic technology exposureMulti-horizon · separate from valuation rank and long-term quality.
CHINA TECHNOLOGY CATCH-UP · 27 JULY 2026
Material exposure across time horizons
High · 7Medium · 7
Research snapshot · 27 July 2026. This analysis has not been refreshed for the September valuation edition.
Medium and High describe relevance to the company thesis—not timing, probability, domain confidence or direct China revenue. Companies without a tag have no material exposure identified in the current layer.
CompanyImpactPrimary exposureThesis consequenceClassification effect
HighLeading-edge fabrication · Memory / HBM · Advanced packaging · +3 more
Domestic AI chips and a future closing of the HBM gap threaten Nvidia’s China profit pool, although low-cost models also expand demand for compute and applications.
No reclassificationCurrent signal and 30-year view unchangedHighMemory / HBM · AI chips · Foundation models · +2 more
Domestic AI chips, cloud and applications can strengthen Tencent’s platform, while weak HBM and rising AI investment remain constraints.
No reclassificationCurrent signal and 30-year view unchangedHighLeading-edge fabrication · Mature / specialty chips · Lithography · +2 more
China still lacks production-grade EUV, preserving ASML’s frontier monopoly; near-term risk is export exclusion and gradual substitution in lower-end DUV and fab tools.
No reclassificationCurrent signal and 30-year view unchangedHighLeading-edge fabrication · Mature / specialty chips · Memory / HBM · +5 more
Chinese fabrication and packaging progress can reduce future China-linked demand, while frontier yield, efficiency and advanced-packaging gaps remain meaningful near-term moats.
No reclassificationCurrent signal and 30-year view unchangedHighMemory / HBM
China’s conventional-memory progress already creates pricing pressure; closing the HBM gap would weaken foreign suppliers’ leverage in the AI hardware chain.
No reclassificationCurrent signal and 30-year view unchangedHighMemory / HBM
China’s conventional-memory progress already creates pricing pressure; closing the HBM gap would weaken foreign suppliers’ leverage in the AI hardware chain.
No reclassificationCurrent signal and 30-year view unchangedHighMemory / HBM
China’s conventional-memory progress already creates pricing pressure; closing the HBM gap would weaken foreign suppliers’ leverage in the AI hardware chain.
No reclassificationCurrent signal and 30-year view unchangedMediumAI applications / assistants
Chinese AI applications can compress pricing and export through devices and apps, increasing pressure on Western application-layer products.
No reclassificationCurrent signal and 30-year view unchangedMediumFoundation models · AI applications
Low-cost Chinese models pressure model economics, but near-term pressure is judged smaller on Meta’s core advertising profit pool.
No reclassificationCurrent signal and 30-year view unchangedMediumFoundation models · AI cloud · AI applications
Low-cost Chinese models pressure AI pricing and assistants; lower-cost Chinese clouds can compete in emerging markets, while near-term pressure on core advertising remains smaller.
No reclassificationCurrent signal and 30-year view unchangedMediumFoundation models · AI cloud · AI applications
Low-cost Chinese models compress model pricing, while full-stack Chinese clouds may compete in emerging markets despite limited direct China exposure for Western hyperscalers.
No reclassificationCurrent signal and 30-year view unchangedMediumAdvanced packaging · EDA / IP
China lacks a complete advanced-node EDA flow, but forced localization can reduce China licences and pricing power as domestic design tools improve.
No reclassificationCurrent signal and 30-year view unchangedMediumFoundation models · AI cloud · AI applications
Low-cost models can compress AI pricing, while Chinese full-stack clouds may compete in emerging markets; direct China exposure is limited.
No reclassificationCurrent signal and 30-year view unchangedMediumLeading-edge fabrication · Advanced packaging · EDA / IP · +1 more
China lacks a complete advanced-node EDA flow, but forced localization can reduce China licences and pricing power as domestic design tools improve.
No reclassificationCurrent signal and 30-year view unchangedLong-term quality dashboardThis view does not inherit the valuation rank, signal or hurdle.
30-YEAR DURABILITY VIEW
Business durability, separated from valuation
A 30-year business-durability view. The tier does not mechanically affect the 3–5-year ranking. Current valuation is excluded from the tier.35-company durability universe · Published 12 October 2026. Companies are alphabetical within each tier, without individual ranks. This is a researched selection, not the 35 best companies worldwide, and differs from the valuation universe.
Which businesses have the best chance of compounding value per share at a high rate over 30 years? Margins, reinvestment, capital allocation, dilution and buybacks all matter. The S&P 500 is the long-term hurdle; current valuation is excluded.
Read the tier framework and borderline cases ↗
| Tier | Companies |
|---|---|
| A+ | Microsoft Unusually high confidence: the compounding engine is particularly difficult to see breaking down. |
| A | Alphabet, S&P Global Very high confidence, with somewhat more long-term uncertainty than A+. |
| B+ | Amazon, Experian, Hermès, Intuit, Intuitive Surgical, LSEG, Mastercard, Meta, Moody’s, MSCI, RELX, Visa, Veeva, Wolters Kluwer A strong chance of remaining a good compounder, with meaningful structural risks. |
| B | ASML, Autodesk, Broadcom, Cadence, Constellation Software, Copart, MANI, NVIDIA, Schneider Electric, Synopsys, TransDigm, TSMC A credible 30-year case, with considerably more uncertainty around technology, competition, cyclicality, concentration or reinvestment. |
| C | Adobe, Netflix, Novo Nordisk, ServiceNow, Tencent, Topicus Insufficient confidence to underwrite the business model for a full 30 years; this does not mean a bad company. |
Historical July ranking · 10 companies
July’s ranks for the companies then in the valuation Top 10 are retained below. The tiers shown are current, not a reconstruction of July’s tiers, so they may differ from the historical order.
July 2026 rankCompanyCurrent tierCompounding engineRationale
#1MicrosoftA+
Office, Azure and developer tools sit inside core business workflows and reinforce each other.
Multiple embedded platforms, strong cash generation and unusually broad reinvestment optionality, balanced by higher AI capital intensity.
#2IntuitB+
Tax, accounting, payroll and credit data turn recurring compliance work into a trusted ecosystem.
Deep trust and recurring compliance workflows support a long, focused reinvestment runway.
#2Meta PlatformsB+
Identity, discovery, social proof, messaging and business contact form a combined network.
Powerful network effects and cash generation, balanced by attention shifts, regulation and high capital spending.
#2RELXB+
Proprietary legal, scientific and risk data are embedded in professional decisions.
Embedded professional workflows, proprietary data and reliable cash conversion support durable compounding.
#5AutodeskB
Design files and workflows are deeply embedded across architecture, engineering and construction.
An embedded design-workflow franchise with meaningful long-term AI and platform risk.
#5Constellation SoftwareB
Decentralized vertical-software acquisitions turn recurring cash flow into a long reinvestment runway.
An exceptional acquisition culture whose next phase depends on maintaining returns at greater scale under post-founder leadership.
#5ExperianB+
Credit data, scoring, identity and decision tools are embedded in lending and consumer workflows.
A durable data-and-workflow franchise with meaningful cyclical and regulatory exposure.
#8AdobeC
Creative tools, professional workflows and the PDF standard still support a broad ecosystem.
A broad creative and document ecosystem whose long-term pricing power is being retested.
#8NetflixC
Global distribution, recommendation and engagement support pricing and content efficiency.
A strong global platform with a less visible long-term reinvestment runway beyond entertainment.
#8TencentC
WeChat, gaming, payments and advertising form a deeply connected consumer ecosystem.
A high-quality ecosystem whose outcome is partly capped by regulation and geopolitics.
Current valuations for other companiesValuation only · 30 September 2026
Alphabet25.6× forward P/E; 79.0× screening P/FCF30 September 2026
Amazon26.6× forward P/E; P/FCF N/M30 September 2026
ASML41.5× forward P/E; 61.2× screening P/FCF30 September 2026
Intuitive Surgical36.0× forward P/E; 44.6× screening P/FCF30 September 2026
Mastercard25.9× forward P/E; 28.9× screening P/FCF30 September 2026
Microsoft26.0× forward P/E; 56.9× screening P/FCF30 September 2026
Netflix20.1× forward P/E; 26.0× screening P/FCF30 September 2026
Novo Nordisk12.0× forward P/E; 14.9× screening P/FCF30 September 2026
S&P Global20.9× forward P/E; 20.9× screening P/FCF30 September 2026
Schneider Electric26.0× forward P/E; 26.6× screening P/FCF30 September 2026
ServiceNow29.6× forward P/E; 30.3× screening P/FCF30 September 2026
Synopsys24.8× forward P/E; 30.3× screening P/FCF30 September 2026
Topicus.com18.9× forward P/E; 10.7× screening P/FCF30 September 2026
TransDigm26.4× FY2026 adjusted P/E; 37.7× annualised H1 P/FCF30 September 2026
TSMC20.1× forward P/E; 57.1× screening P/FCF30 September 2026
Cadence Design Systems38.5× forward P/E; 54.2× screening P/FCF30 September 2026
Moody's25.6× forward P/E; 26.5× screening P/FCF30 September 2026
MSCI25.5× forward P/E; 24.5× screening P/FCF30 September 2026
Visa24.9× forward P/E; 31.4× screening P/FCF30 September 2026
ABOUT THE AUTHOR
Michaël Hillaert
I’m a Belgian independent investor and writer. I study valuation dislocations, durable businesses and strategic technology shifts. This dashboard tracks my assessments; on Substack, I explain the thinking behind them.
For informational purposes only. Not investment advice. Data and valuations are shown as of their stated dates.