Block II

This is not a claim about price, and it does not become false if the price falls. The performance claim lives in Block I position 4, with its own horizon and its own benchmark. Separating them is the point: Block Ø collapsed a value and a prediction into one superlative and had to be taken apart afterwards. This states the property. That one states the bet.

The property is narrow and it is the whole of it: a supply schedule nobody can change, and settlement nobody can reverse, with no entity holding the power to do either. Not scarcity — gold is scarce and its supply still responds to price. Not decentralisation as a slogan, which describes a diagram rather than a capability. The test is adversarial and it has one question: who could alter issuance or reverse settlement if sufficiently motivated, and what would stop them? An answer naming any person, foundation, company or validator set is an answer that the property is absent.

Almost everything sold in the same vocabulary fails that question on inspection, and usually fails it in the documentation rather than in some hidden place. A foundation that can ship a change. A validator set small enough to be subpoenaed. A supply schedule revised by governance vote. None of that is fraud and none of it has to be — it is a different product, with different properties, using borrowed words. The borrowing is the part worth noticing, because the words are doing the work the properties are not.

Which is also the limit of the claim, stated plainly because it is the half most often dropped. The property being unique does not make the asset a good investment, and nothing here says it does. It makes it a different category of thing. What follows from holding it is a separate argument with a separate horizon, and it is filed separately.

Both sides of that ratio get priced, and most arguments about an intervention price exactly one of them.

An unknown risk files as high. Never as low, and never as absent. “No reported side effects” is a statement about a reporting system, not about a compound — it describes how many people looked, for how long, with what incentive to write it down. A surveillance gap reads identically to safety on the page and is its opposite in the body. Treating the two as one thing is the most common error on the risk side, and it runs in both directions: the same sentence waves a compound through and waves a product through.

Then the reward side, where the ratio does its real work and where it is almost never run. A young healthy person has nothing to gain and a full unknown to lose. No deficit to correct, no decline to arrest, no function to restore — so the numerator is close to zero and any unknown in the denominator dominates it. Their upside is still in light, food, sleep, movement and time off the screen: interventions with centuries of record, a known mechanism, and no unknown on the risk side at all. The ratio does not tell them to be adventurous. It tells them they have not earned the trade.

The same arithmetic inverts the moment the known cost is large. Set a standard hernia repair on one side — three new holes in the abdominal wall, a permanent synthetic mesh, an irreversible procedure whose failure modes are documented and whose revision surgery is harder than the first. Against that, a reversible course with a named mechanism and a long safety record is the conservative option, not the reckless one. It is reversible; the mesh is not. And doing nothing is never free — it is an intervention with its own side of the ratio, usually priced at zero because inaction produces no event to attribute. The cost of waiting arrives anyway. It arrives without a name on it.

Ivermectin is the worked example, and the finding is not the one the argument is usually had about. The compound has a safety record across billions of administered doses, sits on the WHO Essential Medicines List, and its discovery took the 2015 Nobel Prize in Physiology or Medicine. Set against it a novel product with no comparable record and no accumulated surveillance. One was removed from consideration for want of trials; the other was mandated while its own trials were still running. That asymmetry is the finding — not whether the compound worked. Doctors who said in public that it did paid licences, hospital privileges and standing for saying it, which under this publication’s own method is evidence about the claim rather than about them.

Start from the standard Sweden’s state epidemiologist stated and was attacked across the international press for holding: an extreme action requires extreme evidence, and the burden sits on whoever takes the action. Anders Tegnell declined the general lockdown and kept the schools open under exactly that rule. Hold every party to it and the record sorts itself, because the actions taken were extreme and the evidence moved the wrong way three separate times.

The claim moved. It opened as protection from infection. It became protection from transmission. It settled on protection against severe illness and hospitalisation. Each retreat was published as though it had always been the claim, and not one was announced as a change. The policy the withdrawn claim had justified did not move at any step — a mandate argued for on the grounds that it stopped a person infecting others outlived the transmission claim entirely, unamended, and remained a condition of employment after the justification for it had been abandoned in the literature.

The definition moved. “Vaccine” was redrawn mid-campaign: the public definition shifted from producing immunity to producing protection. When the product stopped meeting the definition, the definition was edited rather than the product reclassified. This is not an allegation, which is what makes it the cleanest of the three — it is a revision history, timestamped, on the institution’s own page, and still retrievable.

The liability moved, and it was priced by the sellers themselves. Indemnity was a precondition of supply in every jurisdiction that bought. When India demanded the manufacturer carry liability for adverse events on the ordinary commercial terms every other product carries, the manufacturer declined the market rather than the terms. The risk was moved onto the individual receiving the product, and the route to compensation for an injury was moved out of the courts and into a no-fault scheme with a payout rate close to zero.

Three movements, one direction, and the policy constant through all of them. The confidence of a seller is measured by what they will indemnify, never by what they will assert — and on that measure the public confidence and the contracted confidence were never the same number at any point.

A reference range is a census, not a standard. It is the central 95% of a reference population — the 2.5th to the 97.5th percentile — established per laboratory and per assay, and it answers one question: how does this result compare to the people in that sample. The question actually being asked is a different one: what value does this system work best at. Those two questions have different answers, and only the first is printed on the report.

Take the number the whole field cites. The standard adult-male total testosterone range runs 264 to 916 ng/dL. The functional band — where body composition, recovery and drive sit — runs 600 to 900. Hold the top and the bottom of the official range next to each other: it spans a factor of three and a half. A man at 290 and a man at 900 are both reported normal by the same line of text. No physiological band is 3.5x wide. A result inside that interval carries almost no information about function, and a clinician saying “normal” is reporting, accurately, that the result was unremarkable in the sample — which is all the interval was built to say.

The population doing the describing is also not a well one. Most ranges are built from whoever presented for testing, and that cohort trends with the metabolic health of the country it is drawn from. As that declines, the interval widens to keep containing it. The band tracks the decline rather than resisting it, which makes “your labs are normal” resolve to “you are unremarkable among the unwell” — a true statement, a useless one, and an instance of a true statement built to mislead.

The distinction is not theoretical and it is not fringe: it is already schematised. A production marker catalogue carries 57 markers with the laboratory range and the optimal band as two separate fields, because they are two different objects and cannot be stored in one column. Everything required to read a result against function rather than against the queue already exists. It is simply not what gets handed across the desk.