Stactray
Menu

Thesis

We look for industries where a standard would settle the argument.

Stactray goes to the parts of the economy that are made of brick, mortar, stock and paper — and looks for the specific kind of problem that is not a shortage of software but the absence of a standard. Then it follows a single unit of that industry — one asset, one record, one transaction — for the whole of its life, and asks what would be different at every step if the standard existed.

This page publishes the method in three moves. It does not publish what the method concluded about any particular platform — those are operating decisions of a private company. What is the reader’s business is whether a method exists at all, and whether it is the kind that can return an unwelcome answer.

One — the screen

The issue has to be an assertion nobody can check.

Every one of these industries already has software. What it does not have is a way for two parties who do not trust each other to agree on a fact without a third one adjudicating it. That is not an application problem.

The sector screen Six brick-and-mortar sectors, each paired with the assertion in it that nobody can independently check. All six converge on one test — whether a single standard would settle that assertion for every party at once — and what passes the test is a candidate sector. SECTORS WE LOOK AT THE ISSUE, AS IT STANDS TODAY THE SCREEN LAND & TITLE who actually holds it CONSTRUCTION what state the project is really in TRADE CREDIT whether the counterparty pays GEMS & JEWELLERY where the stone came from LOGISTICS what is actually in the container AGRI SUPPLY what grade actually shipped A CANDIDATE one standard would settle it
  1. Land and title — who actually holds it.
  2. Construction — what state the project is really in.
  3. Trade credit — whether the counterparty pays.
  4. Gems and jewellery — where the stone came from.
  5. Logistics — what is actually in the container.
  6. Agri supply — what grade actually shipped.

The test at the membrane is a single question: would one standard settle this assertion for every party at once? Where the answer is no, the problem is competition, or price, or execution — real problems, and not ours.

Two — the lifetime

Then we follow one unit for the whole of its life.

Not the market, not the sector, not the category. One asset, one record, one transaction — traced from the moment it comes into existence to the moment it leaves someone’s hands, with every point at which it changes state written down. What the industry asserts at each of those points is above the line. What a standard would write instead is below it.

The lifetime of a single unit One unit traced through five states — origin, custody, transfer, finance and exit. Above the lifeline, in unlit dashed boxes, is what the industry asserts at each state. Below it, on lit nodes, is what a standard writes instead. The lifeline runs dark at origin and bright at exit: the record accumulates. WHAT THE INDUSTRY ASSERTS papers say it exists someone says they hold it a receipt, somewhere a valuation, unsourced the file closes ORIGIN identity, at first sight CUSTODY the holder, with a date TRANSFER the change of hands FINANCE what it stands against EXIT the history, still readable WHAT THE STANDARD WRITES
  1. Origin — asserted: papers say it exists. Written: identity, at first sight.
  2. Custody — asserted: someone says they hold it. Written: the holder, with a date.
  3. Transfer — asserted: a receipt, somewhere. Written: the change of hands.
  4. Finance — asserted: a valuation, unsourced. Written: what it stands against.
  5. Exit — asserted: the file closes. Written: the history, still readable.

The difference between the two rows is the whole business. The top row is record-keeping — software that faithfully stores what somebody asserted, because it was built when checking the assertion was economically impossible. The bottom row is infrastructure, and it is not a module that can be added to a system built on the other assumption.

Three — the durability test

Five axes, and one question asked five ways.

A candidate sector and a traced lifetime say a platform is worth building. Neither says it will still be worth holding. Every platform is scored zero to five on the same five axes, applied identically so that comparison is possible and preference is visible. Each axis asks whether something survives when the analysis layer above it becomes cheap.

The five-axis durability instrument A blank five-axis instrument, scored zero to five on each axis: data gravity, trust position, workflow lock-in, real-world coupling and distribution. The frame is shown unplotted — no platform is plotted on it on this page. 1 2 3 4 5 01 · DATA GRAVITY Data no model can buy 02 · TRUST POSITION A licence someone must hold 03 · WORKFLOW LOCK-IN The record, not a view 04 · REAL-WORLD COUPLING Atoms, custody, legal state 05 · DISTRIBUTION Demand owned, not rented

The instrument is shown unplotted on purpose. How any platform reads against these five axes, the rankings that follow, and the allocation derived from them are operating decisions of a private company and are not published on this page.

Read together the five ask a single question in five ways: what is still here when the clever part is free. Four answers survive that question — proprietary transaction data, a regulatory or standards position, custody of a physical or legal state, and an owned demand channel. Everything else is a wrapper over one of them.

Two conventions, declared rather than buried

The five axes are weighted equally. That is a convention, not a finding, and it is in tension with the instrument’s own use — if distribution is the constraint that binds, an argument exists for weighting it above the four that capability improvements attack. The question is booked for the annual re-run rather than answered mid-cycle, because re-weighting an instrument while its rankings are in view is how an instrument learns to agree with itself.

Every assessment is also read against three futures — capability improving steadily, capability improving very fast, and capability flattening — carried at fixed weights that are likewise a device rather than an estimate. Their only job is to make sure each platform is read against a world it would not choose, with optimism and pessimism carrying equal weight so that neither gets the casting vote. A conclusion that flips under a plausible re-weighting is treated as not established.

How it is used

An instrument that is scored once and filed is an essay. This one runs on a cadence, and the cadence is the part that makes it real.

  • MonthlyOne number and one question: did the platforms judged most durable actually receive the attention the allocation says they should, or did the allocation get captured by whoever had a counterparty who wrote in.
  • QuarterlyA sweep of every assessment’s tripwires — the observable events written in advance that force a change of plan — together with the register of concentrations and single points of failure.
  • AnnuallyA full re-score of every platform, a re-derived allocation, and a re-argument of every decision against what actually happened rather than what was intended.

Platforms are stopped, merged, held as options or funded on the strength of that re-argument. Holding something as an option is an explicit state with a written trigger and near-zero attention against it, because option-shaped work is the work that feels most like progress while producing least.

Three disciplines keep the instrument honest, and each exists because of a specific way this kind of exercise usually fails.

  • The re-score is anchored outThe annual re-run is performed against the underlying assessments with the previous scorecard out of view, and a second scorer is preferred wherever one is available. An instrument re-run in sight of its own last answer reproduces it.
  • Open questions are publishedAn assessment that hides what it does not know is a pitch. Each one ends by naming the questions its conclusions depend on and what each would decide.
  • Waiting is not a stateWhere an outcome depends on a third party, the plan states what happens if they never reply — a unilateral action, our own date, and a defined change of state on silence. Both branches count as closure. A date on somebody else’s decision is an undecided item wearing a deadline.

Boundary

This is a framework for allocating a private company’s own capital and its own attention between businesses it owns. It is not an offer, an invitation or a solicitation of any kind. Stactray does not manage third-party money, does not operate a fund or any pooled vehicle, and does not take outside capital into one. Nothing on this page is advice, and nothing on it is a forecast of any outcome for anyone.

Every forward-looking judgement produced by this instrument is a hypothesis with a confidence band rather than a claim. The assessment is of structure, not of value, and it is not a valuation input. Where a regulatory characterisation appears anywhere on this site it names a regime and nothing more — the register on the Governance page states that distinction in full, and it holds here too.

Questions

Written answers

The questions people actually ask, answered in advance. Nothing on this panel is generated.

Not answered here? enterprise@stactray.tech reaches a named officer, not a queue.