v3.4: the rule-of-its-own pass
v3.4 is the canonical full edition (556 pages). It follows v3.3, the delegation-economy pass of 20 September. It ships with a new concise edition, brief v3.4, and essay v1.4.
The rule manufactures the residual
Part I had stated the buyer-quality problem in two halves: a leveraged residual that withdraws for reasons of margin, and an automatic bid that follows an index weight whatever the yield. Federal Reserve staff research shows they are one chain. Funds that track a benchmark’s duration while holding cash portfolios tilted elsewhere buy the missing duration in futures, and leveraged funds take the other side, short the future and long the bond, financed in repo. The new section in §2, The World Forces New Monetary Primitives states it in one line: the index does not hold the duration; it orders it. The rule-follower at the head of the chain cannot withdraw, and the intermediary at the tail can withdraw in an afternoon, so the demand that appears most stable is the demand financed most fragilely. March 2020 is the stressed case: a shock absorber that works until its mounting fails, and the mounting is the repo market.
What changed in the record
- Why the macro model has states, not a threshold. The binding constraint is the incremental interest-rate risk to be absorbed against the capacity of the absorbers, not the debt ratio. The macro policy states gain duration absorption as a cross-cutting capacity, with the observation that a policy move taken on a filtered yield can tighten into the mounting. Central clearing, phasing in at the end of this year and the middle of next, moves the capacity wall outward without removing it.
- Description becomes instruction. A rule that enough capital follows stops describing a market and starts being one. The thesis reads this as the human-institutional form of the concentration the threat model names for agents, which is why the Native Policy Concentration Ratio is a benchmark’s native analogue.
- Premise 9 from the reference asset. If the price of the world’s risk-free asset cannot be read without first identifying who is speaking and under what rule, the price of a monetary candidate certainly cannot.
- Standing reached by obedience. Standing inversion gains a second route on which every key stays with a person and the person never uses it, because a bad outcome after following an agent is charged to the instruction while one after overriding it is charged to the person. It needs no new instrument: a human-revocable share that holds while the policy concentration of the same cohort rises is standing exercised by nobody. A new objection, “Permanent receipts make obedience rational,” is answered by default privacy and the fresh start, with the limit conceded for audited institutions.
- Thresholds have authors. Layer 6 now says that whoever sets a threshold owns its errors as fully as the signers of an override own theirs, and that an override record judges the rule across many decisions, not the signer on one outcome.
- Indicators and quarantine. The plumbing is read in DV01 and by counterparty class, with the scaling and failing configurations named. The claim that mandate-followers are the Treasury market’s dominant voice is quarantined as a conjecture awaiting its own test, and no move in yields, including the long-end selloff after the September policy move, is attributed to the mechanism.
The book, read as one document
A whole-document read found the argument sound and the book uneven. Sentences in older prose had drifted from what the thesis now holds, and each is resolved in the stronger register. The Introduction’s claim now ends: “This thesis does not ask to be believed. It asks to be measured, and it names in advance every reading that would retire it.” The ZK chapter no longer implies that a dishonest prover can forge a sound proof. The Conclusion assigns the Verify families to the stack and leaves the money to its holders, and it now ends once instead of four times. The Contributions name eight contributions, among them the parts that are new. The delegation buyer, parked since v3.3, reaches the Abstract and the plain-language summary. Version history leaves the body for Appendix J, and the voice is made one throughout.
The questions a practitioner asks
The evaluation framework and the operator checklist had tested accrual, the stack, and wrappers, and never asked the three questions the monetary claim rests on. Both now ask them: would a lender accept the asset without investigating it, does anyone reckon in it, and is holding it worth more when substitutes fail. Both also ask whether anyone holds the asset through stress, and whether issuance is bounded by stressed delivered capacity. The standalone operator checklist is re-extracted from this edition.
Brief v3.4 and essay v1.4
The brief now tracks the thesis it condenses, so it is numbered v3.4. It states eleven premises, adds standing inversion by both routes, the rule-and-leverage bid, and the candidate nineteenth line, and drops two passages of version history. The essay moves to v1.4. It no longer calls the unspanned state “the most probable configuration of the next decade” two sentences before saying it predicts nothing; an unsourced lease-commitment figure and a claim that nobody else publishes kill conditions are removed; the First World War convertibility line is corrected; and it gains the machine buyer, standing, and the rule-and-leverage bid. Brief v3.1 and essay v1.3 remain available, unchanged.
Source concentration
Seven essays by one commentator had come to carry 33 citations, most of them repeat pointers to ideas the thesis had already adopted and named. Each is now cited where its construct is first adopted, in the quarantine table, and in Appendix J, and nowhere else. Fifteen remain. The two figures that had been taken from commentary are withdrawn.
Stated limits
The mechanism is established; its dominance is not. The thesis adopts the chain and the question, and it does not adopt the answer. Nothing is timed, no yield is attributed, and nothing in the pass bears on price. The Bitcoin concession stands, and no new layer is added.
Unchanged, deliberately
No red line, threshold, rating, premise, or count moves: eleven premises, eighteen red lines, one candidate.
Also fixed
“2½%” had printed as “2¡%” in the World Forces sidebar, body, and figure caption, including in the published v3.3 PDF. The four sources v3.4 added had not been registered as citation keys and printed placeholders in the first release compile; the keys are registered, and the build check now fails on any unregistered key.
Reproducibility
The checksum files record the LaTeX source commit the PDFs were built from (a432884, tag v3.4; essay tag essay-v1.4).
Boundaries held
No dated outcome predictions. No fund, token, or gate. No red line relaxed, no threshold moved, no rating raised.