The Conviction Play
What makes a company exceptional?
Most businesses are ordinary. A few have economics so durable they keep compounding for decades — and telling one from the other is a question with an answer, not a matter of taste. Every piece here puts one company through the same five structural tests and says plainly what came back. The ones that clear all five join the Exceptional Company Series. The ones that don’t get published too — a bar that nothing fails is not a bar.
422
measured
46
exceptional record
14
…and no open questions
39
just misses
A company is judged on two separate things: what twelve years of earnings show, and what the latest accounts leave open. The record decides whether it is exceptional; the condition is reported next to that, never folded into it. See what we measure, and where every name sits →
The Exceptional Company Series
What exceptional earnings actually look like: profits that compound durably and for reasons that hold. Eleven of the twenty-one companies we have written up on this basis clear all five tests. An assessment of business quality, not a buy signal — exceptional can also be expensive.
Candidates — clear the same five tests, and fall short of the grade on measure rather than on kind.
What the grade means, and why a company you expected is not on it →
Where must capital flow? A second framework runs alongside the assessments: seven forces that decide which industries have a tailwind at their back, and which are working against one. See the forces →
Recent
The assessments themselves, as they were written, most recent first — the companies above and the ones that did not clear the bar alike.
2026-08-28ResMed (RMD): The Mask, Not the MachineSells the pump once, then the mask that wears out every few months for as long as therapy lasts — with the reimbursement pathway keeping score.
San Diego, California, United States
2026-08-27IDEXX Laboratories (IDXX): The Vet's SubscriptionSells the analysers that sit in veterinary practices, then earns again on every test they run. The piece asks what makes that annuity durable.
Westbrook, Maine, USA
2026-08-24Intuitive Surgical (ISRG): The Surgeon's SubscriptionSells the da Vinci surgical robot, then earns again on every operation it performs. The piece asks what makes that annuity durable.
Sunnyvale, California, USA
2026-08-17Moody's (MCO): Permission to BorrowCharges issuers for the credit rating that lets them borrow. The article asks what that permission is worth.
New York, USA
2026-07-07Oracle: Underwriting the AI Investment CycleSelling compute into the AI build-out. The article asks whether the spending is demand or hope.
Austin, USA
2026-07-05Netflix: Priced Like the Story's OverThe largest streamer, re-rated down hard. The article asks whether the economics followed.
Los Gatos, USAWhat it is
Written assessments of individual public companies — whether the business is durable enough to keep compounding. One company at a time, against the same five tests.
What it is not
Never a view on the price. No price targets, no portfolio, no alerts, nothing to buy. A company can be exceptional and still be far too expensive — those are two different judgements. Free, with no paid tier.
Who writes it
One person, investing his own money on the same conclusions, disclosing his position in every piece.
What we measure · How a company is judged · Our team, and what we get out of it
Each piece is an assessment of what is known on the day it is written. Company reports bring genuinely new information, and a view that holds this quarter may not hold the next. Nothing here is a prediction, and nothing here updates itself.
The evidence
What would prove it wrong.
Every piece ends by naming the conditions that would prove it wrong, and the scheduled event that will test them. This is what happened next — kept up to date, including when the answer is that we got it wrong.
Nothing on this page is ever deleted or quietly revised. A test can only gain an outcome, and changing our mind is recorded as a dated event with a reason, not as an edit. That is deliberate: a record with nothing broken in it would not be a record.
Where the record starts, and why
The tracker begins with Oracle, in July 2026, when every piece started naming in advance, with a date attached, the evidence that would break it. Earlier work stays published and linked from the record, but is not tracked: a falsifier written after the outcome is known is not a falsifier.
How long a thesis stays open
A record nobody maintains stops being a record, so the rules are fixed and deliberately small.
- A thesis is live until its last dated test resolves, then it closes with a stated outcome. It ends by construction rather than by anyone deciding to stop looking.
- Two years is the maximum life. If the tests have not resolved within twenty-four months of publication the thesis expires and says so — a claim that cannot be tested in two years was not a testable claim.
- One sweep a year, every January, checking every live row against what actually happened. Rows retire as fast as they are added, so the work does not grow with the archive.
- Closed theses stay on this page, broken ones included.
- Nothing is ever added to a thesis after publication except an outcome. Falsifiers are set when the piece runs, or not at all.
Reading a test
Pending — the event hasn't happened yet.
Confirmed — the event happened and the thesis held.
Broken — the condition came true. Evidence against us.
Each line is written as the thing that would damage the argument, so a test passing means the damage did not occur. Dates are when the evidence is scheduled to arrive, not when we expect a share price to do anything.
Nothing tracked yet.
The companies
Every company, on its own page.
Every company we have written up, with what we assess it to be stated on its own page. Sector, employees, market cap and a five-year price chart — raw market data, no valuation view.
Some of them clear all five structural tests and stay clear — the Exceptional Company Series, with a handful more named as candidates. The rest are here for the same reason: a bar that nothing fails is not a bar, and several of them are judged on a different basis entirely.
27 companies, eight places. Every company placed where it operates from. The shape is the point: a row of large American franchises, and then three companies in one small city in the Caucasus.
Hover a marker, or pick a company —
The Forces
Where must capital flow?
Not where it might, or where a story says it should — where it has no real choice.
Seven destinations, and the twenty-two themes beneath them. Each is a shift large enough and slow enough to still be running in ten years — which is the only kind worth underwriting against.
What is a force?
A destination, not a driver
Somewhere capital has to go, whatever is pushing it. Decarbonisation, ageing and coercion risk are the pressures; these seven are the places the money lands. Hover a node to see one.
Hover or focus a node for a preview; click for the full section.
The standing question
Which companies here, if any, are exceptional?
The map decides what gets looked at — not what the news happens to be talking about that week — and every candidate gets the same five structural tests. It is why companies turn up here that almost nobody is writing about.
Some forces turn up nothing, and that is an answer rather than a gap: capital has to go there and nothing durable enough has been built yet.
Five of the seven have companies written up so far. Two have none yet — which the coverage line under each force states plainly rather than leaving you to count.
The seven forces, and the twenty-two themes beneath them
Revised as the evidence changes — last 26 August 2026. Forces are added, renamed and merged; nothing here is fixed.
The whole map, spoken
4:08 · 4.0 MB · transcript · narrated by a synthetic voice, not by the author
Force 1 of 7
Power and energy systems
Electricity has to be generated, moved, stored and controlled in far greater volume and with far less tolerance for failure than the system was built for. Capital lands here whether the driver is decarbonisation, industrial load growth or simple replacement of assets past their life.
Listen to this force
3:11 · 3.1 MB · transcript · narrated by a synthetic voice, not by the author
Grid Expansion, Reinforcement & Modernisation
three expressions
Massive capital deployment into transmission, distribution, and grid resilience to support higher loads and bidirectional power flows.
End-Use Electrification of Transport, Industry & Buildings
four expressions
Sustained replacement of combustion-based systems with electric alternatives across mobility, industry, and the built environment.
Power Management, Efficiency & Control Systems
three expressions
Structural demand for equipment and systems that optimize conversion, efficiency, stability, and control across an electrified stack.
Force 2 of 7
The built environment and its resilience
Physical infrastructure gets hardened, duplicated and redesigned to survive things it was not specified for — extreme weather, deliberate attack, and resource scarcity. The threat differs; the concrete, the water system and the redundancy are the same spending.
Listen to this force
3:13 · 3.1 MB · transcript · narrated by a synthetic voice, not by the author
Climate-Resilient Infrastructure & Adaptation
four expressions
Rebuilding infrastructure to survive physical stress it was not specified for: storm-resistant materials, stormwater and drought systems, grid and telecom hardening, and the analytics that price the exposure. The threat is physical and statistical, not hostile.
Strategic Infrastructure & Critical Systems Protection
three expressions
Protecting infrastructure against a deliberate adversary and against loss of national continuity: operational-technology cybersecurity, hardened and secure construction, and independent power for sites that cannot go dark. The threat is intentional. Weather-driven hardening is climate-resilient infrastructure.
Water, Waste & Resource Efficiency Systems
five expressions
Sustained capital flows into systems that reduce resource intensity, recycle inputs, and secure reliable access to water and critical materials under constraint.
Force 3 of 7
Industrial capacity and supply chains
Where things are made, and what they are made from, is being rebuilt for security of supply rather than lowest landed cost. Reshoring, redundancy and material substitution are one destination reached by two roads — coercion risk and physical scarcity.
Listen to this force
2:44 · 2.6 MB · transcript · narrated by a synthetic voice, not by the author
Supply Chain Resilience, Localisation & Redundancy
four expressions
WHERE production sits and how goods move: reshoring and the industrial capex it pulls, freight and intermodal capacity, distribution and buffer inventory, and the software that makes a chain legible. A logistics and capacity theme, not a materials one.
Covered on this site: Descartes Systems Group
Input Substitution & Material Innovation
three expressions
What products are made OF: critical minerals secured outside a single jurisdiction, circular and recycled feedstocks, and engineered or bio-based substitutes for scarce inputs. A materials theme. Where things are made and how they move is supply chain resilience.
Force 4 of 7
Automation and machine-performed work
Work moves from people to software and machines. Whether the reason is a shortage of workers or the cost of employing them, the capital goes to the same place: systems that perform the work, and the decision layers that direct them.
Listen to this force
3:20 · 3.2 MB · transcript · narrated by a synthetic voice, not by the author
Software-Led Process Automation
five expressions
Software that performs or removes work inside an organisation: horizontal workflow platforms, function-specific automation in finance, HR and service, vertical SaaS with automation embedded, and the systems that manage the workforce itself. Not machines doing physical work, and not the decision layers that direct them.
Covered on this site: Accenture · Microsoft · Oracle · ExlService
Intelligent Physical Automation
six expressions
Machines that perform physical work, sold as horizontal capability rather than to one industry: vision and inspection, mobile and collaborative robots, field and off-highway autonomy, and the automation of factories and warehouses. Not software that reorganises how people work — that is software-led process automation — and not labour substitution scoped to a single sector.
Covered on this site: AeroVironment
Algorithmic Decision Infrastructure
four expressions
Firms institutionalise algorithmic decision-making to improve speed, consistency, and outcomes across operations and planning.
Covered on this site: Arista Networks, Inc. · Moody's Corporation · MSCI · Microsoft · Oracle
Automation of Routine Human Labour
three expressions
Labour substitution scoped to one sector at a time, where the constraint is that the people are not there to hire: self-service in customer-facing retail, autonomy in agriculture, administrative substitution in healthcare. The distinction from the horizontal robotics and software themes is deliberate — here the theme is a labour market, not a technology.
Force 5 of 7
Defence, security and strategic autonomy
States rebuild the capacity to defend themselves and to operate without depending on anyone else — across land, cyber and orbit. The buyer is sovereign, the budget is set by threat rather than by return, and it does not turn off in a downturn.
Listen to this force
2:38 · 2.5 MB · transcript · narrated by a synthetic voice, not by the author
Defence, Deterrence & Security Capacity Expansion
one expression
Governments structurally increase spending to rebuild and sustain military, cyber, and homeland security capabilities.
Covered on this site: AeroVironment · Fortinet, Inc. · HEICO · Curtiss-Wright
Orbital Access & Space Infrastructure
two expressions
Falling launch cost turns orbit from a scarce, state-controlled destination into usable infrastructure — communications, observation and positioning that both commercial and defence customers come to depend on.
Covered on this site: SpaceX
Force 6 of 7
Health and wellbeing
Populations live longer with more years of managed illness, and the care has to be delivered by a workforce that is not growing. Spending lands on diagnosis, intervention, chronic therapy and the systems that deliver them with fewer hands — and, increasingly, on keeping people out of that system at all: nutrition, prevention and the things people buy to stay well. The earlier the money is spent, the cheaper the outcome, which is why the same destination now attracts capital at both ends.
Listen to this force
3:48 · 3.7 MB · transcript · narrated by a synthetic voice, not by the author
Precision Diagnostics & Testing Infrastructure
two expressions
Diagnosis moves earlier and gets more specific, which means more tests per patient and more instruments in more places. The economics sit in the recurring consumable, not the analyser.
Covered on this site: IDEXX Laboratories, Inc. · Thermo Fisher Scientific
Interventional & Platform Medicine
two expressions
Procedures migrate onto platforms — robotic, catheter-based, image-guided — that a clinician trains on for years. The platform earns on every procedure it performs, and the training is the moat.
Covered on this site: Intuitive Surgical, Inc.
Metabolic & Chronic Disease Therapeutics
one expression
Chronic metabolic disease is treated for decades rather than cured, and a class that works at scale reshapes demand across every specialty downstream of it.
Preventative & Self-Directed Health
two expressions
Spending to stay well rather than to be treated: screening bought out of pocket, nutrition, sleep, recovery and mental health. Elective, largely unreimbursed, and rising with age and income together.
Care, Health & Dependency Infrastructure Expansion
four expressions
Delivering care with a workforce that is not growing: home and residential care capacity, the staffing and scheduling systems behind it, and the infrastructure that lets fewer people look after more. About who does the work, which is what separates it from the medicine themes beside it.
Covered on this site: ResMed Inc.
Force 7 of 7
Experience and discretionary time
As incomes rise and populations age into more leisure hours, a growing share of spending goes to how time is filled rather than to goods. The winners own a habit and the distribution to serve it.
Listen to this force
2:58 · 2.9 MB · transcript · narrated by a synthetic voice, not by the author
At-Home Entertainment Platforms
one expression
Scaled subscription platforms that amortise content across a global base, where each additional market is served at near-zero marginal cost and the catalogue is the barrier.
Covered on this site: Netflix
Live, Travel & Experience Economy
two expressions
Spending shifts from owning things to being somewhere — travel, live events, hospitality — and the durable economics sit in the booking layer and the venue, not in the operator carrying the asset.
Interactive & Social Entertainment
two expressions
Entertainment people participate in rather than watch — gaming, real-money interactive, social platforms — where engagement compounds through the network and monetisation is per-session, not per-title.
Covered on this site: Evolution AB · Grindr
At the centre
Our world
Every one of these forces acts on the same planet, and the physical state of it — climate, heat, water, air, materials — is a condition on all of them rather than an eighth force beside them.
That is why it has no spoke of its own. These forces sort capital by where it lands, and the environment is a reason money moves rather than a place it arrives. A spoke named “climate” would cross every other spoke and locate nothing.
Listen to this
2:57 · 2.9 MB · transcript · narrated by a synthetic voice, not by the author
Where the physical fact does most of the work
Take the environment out of these three and the force itself changes shape.
Heat raises the peaks, decarbonisation raises the volume, and the grid has to carry both at once.
Storm, flood, heat and drought are the load cases everything now gets rebuilt against.
The physical world decides what is scarce, and where it can still be dug, grown or made.
Where it reaches, but does less
Real in all four, and not the thing driving them.
Compute is a physical plant. Power draw and cooling water are siting decisions before they are engineering ones.
Water, food and habitability move people; resources are half of what strategic autonomy means.
Heat, air quality and the range of disease are health inputs, not background conditions.
Where people can comfortably be, and in which months, is being redrawn season by season.
Every force is reached; three of seven are shaped by it. The line under each says how.
Take it with you
The whole map on five sides of A4 — every force and every theme, for reading away from a screen.
A force here is a destination, not a driver. Supply chains, infrastructure and power are not pressures — they are places money has to end up. The pressures are decarbonisation, ageing, labour scarcity and coercion risk, and several of them push toward the same destination at once. That is exactly why the map is drawn this way round: sorting by pressure would put one company in four categories at the same time, and sorting by where the money lands puts it in one.
The record
Everything, in order.
Nineteen pieces published, filterable and sortable — fewer than the companies we cover, because a company gets a profile as soon as it is assessed and an article only when there is something worth saying. This is the work; how it was judged is in the method, where it belongs.
Every piece has a key-points card. Company, what the business is, the figures the piece turned on, the call where one was made, and what would prove it wrong — about twenty seconds each, and you can cycle straight through them without opening anything.
The archive
Nothing matches that filter.
The method
Seven questions, every time.
Every piece answers the same seven questions in the same order. That is what makes two articles about completely different businesses comparable — and what makes it obvious when one of them is dodging something.
- What does the market believe?
Stated as the market would state it, not as a strawman. - Why does it believe that?
The bear narrative's own logic, taken on its merits. - What evidence supports it?
Figures traced to public filings, not to a summary of them. - What evidence contradicts it?
The gate. If this answer is weaker than the one above, the piece goes back. - What would invalidate the thesis?
Named conditions, each with the date the evidence arrives. - What remains genuinely uncertain?
Written as open questions, never as instructions. - What follows from all of it?
A personal stance, with the position disclosed.
How a piece gets made
Research runs through Legend and produces an evidence pack: the market narrative, the bull and bear cases, the verified figures, the dated falsifiers, the unknowns. Only then does anything get written.
The draft is written against a fixed structure, then put through a second, adversarial pass that compresses it, strips decorative metrics, calibrates every claim, and re-audits it line by line against the rules below. Two automated checks run before publication. Neither of them can approve a piece — every discretionary call is signed off by a person.
How a piece is marked
Ten criteria, one to five each, applied to every finished draft. They are fixed in advance and they do not change between companies, which is the only thing that makes two scores comparable.
The ten criteria, and what each one asks
- Central question
One underwriting question, stated early. A reader can state the debate in a sentence after three paragraphs. - Opening and reframe
Opens on a principle, with the company as the evidence. Pinpoints where the mispricing sits. - Concept teaching
Teaches a transferable idea, not just a stock. How to think, not what to hold. - Evidence quality
Specific, checkable facts do the analytical work. Not adjectives. - Data discipline
Every figure traces to a source and reconciles. No invented numbers. - Bear-case honesty
As strong as the bull case, led by the biggest real risk, no strawman. Below 4 here and the piece cannot publish — the only criterion with a hard gate. - Falsification and catalysts
Names in advance what would change the author's mind, and the dated event that tests it. - Structure and economy
Each section earns its place. Each load-bearing idea said once. - Voice
Calm, plain English, honest. At least one memorable line. Not promotional. - Discretion
No trigger prices, no ratings, disclosure correct, the system invisible.
Criterion six is the one that does the work. It was added after a review of the first twelve pieces found the bear case got weakest exactly where the author was most bullish — the signature of a pitch rather than an assessment. It is now the only criterion that can stop a piece being published.
Editorial scoring
| Date | Company | Score | Biggest weakness found |
|---|---|---|---|
| 2026-07-07 | ORCL | 4.7 | Swing factor restated across three sections |
| 2026-06-19 | MTCHnot published | 4.7 | No dated catalyst tying the test to a schedule |
| 2026-06-18 | MSFT | 4.7 | Central idea restated four times; mid-sections loose |
| 2026-06-17 | TMO | 4.7 | Durability drivers repeated across closing sections |
| 2026-06-30 | UZNF | 4.5 | "Can I actually buy this" section read brochure-ish |
| 2026-06-18 | SPCX | 4.5 | A "do nothing" conclusion risks feeling anticlimactic |
| 2026-06-19 | ACN | 4.3 | Thin on checkable facts; ran on reasoning not filings |
| 2026-06-16 | MSCI | 4.3 | No dated falsifier; "wait for a panic" is untestable |
| 2026-07-05 | NFLX | 3.8 | Soft rounding; bear case pre-rebutted before it was made |
| 2026-03-06 | GRND | 3.5 | No real bear case; earnings section read as a pitch |
| 2026-04-07 | AVAV | 2.8 | No bear case; leaned on statistics instead of argument |
| 2026-01-16 | CBRL | 2.3 | No thesis, no bear case, no falsifiers, thin financials |
The four weakest pieces shared one fault, and it is the fault this journal now exists to avoid: the case against was missing. The editorial principles were rewritten around that finding, and the review pass tests for it on every draft.
What this journal will never print
- A price to buy at, or a price to sell at.
- A rating. No buy, hold or sell — on any company, ever.
- A manufactured reason that today is the day.
- A claim that cannot be traced back to a public filing.
- A position held but not disclosed.
These are editorial lines, not legal ones. They exist so the writing stays analysis rather than a recommendation, and they are checked on every piece before it goes out.
Built on Legend
The research runs on a system.
Legend is a rules-based investment research system, built over several years and used on every company in this journal. It is not a product, it is not for sale, and it does not issue signals to anyone. It is the reason the analysis is consistent rather than a matter of mood.
Different businesses, one method — the straight line never bends to fit the answer.
What it does
It values a business by what kind of business it is.
A bank, a cyclical, a compounder and a holding company do not become worth something for the same reasons, so they are not valued the same way. Legend picks the method that fits the business and applies it identically to every company of that type — which is what stops the method being chosen to suit the answer.
It separates value from quality from timing.
Three different questions that get conflated constantly: is it worth more than it costs, is the business actually good, and is now a sensible moment. Kept apart, they disagree usefully. A cheap price on a deteriorating business looks nothing like a fair price on an improving one. Only the quality answer is published here. Nothing on this site is a view on a share price.
It judges a record and a condition, and keeps them apart.
Whether a business is exceptional is decided by what twelve years of earnings show — how often they rose, how far they fell, how fast they compounded. What is strained in the latest accounts is reported next to that, never folded into it. A company can have an exceptional record and an open question today, and the honest thing is to say both.
It writes down what it thought, when it thought it.
Every assessment is recorded at the time it is made, in a record that is only ever added to. That is what makes an honest review possible later — including of the calls that aged badly.
It is the same system running the author's own money.
Legend was not built to produce a newsletter. It was built to make decisions about a real portfolio, and it still does. The writing is a by-product of that work, not a business it serves.
From filings to insight
The three questions are kept apart on purpose: a cheap price on a deteriorating business looks nothing like a fair price on an improving one. AI helps with the research and the drafting; it never assigns a verdict.
- Consistent
- Unbiased
- Repeatable
- Scalable
The system's internals stay private: the formulas, the thresholds, the levels at which it changes its mind, and whatever it currently concludes about any named company. What appears in an article is the reasoning and the public evidence behind it. The tool shows through the discipline, not through screenshots of a dashboard.
How I make this
I research every company through Legend, my rules-based investment research system. It values businesses using the right method for their type, then tests that value against business quality and timing. AI helps me research, challenge the thesis, and write the piece. The judgement is mine, and so is the money.
This statement appears on every piece published. It names the AI assistance outright rather than describing everything except the drafting — an under-disclosure costs more trust than the disclosure does.