privacy · proofs · compute
v3.2 · checksummed
2026-08-16 · v1.8

v1.8 — Duration of the Claim, Duration of the Project

v1.8 — Duration of the Claim, Duration of the Project

Date: 2026-08-16

Version 1.8 makes the thesis claim narrower in one place and stronger in another. Duration-neutrality of the monetary object remains a first principle. What v1.7 had treated as the whole temporal problem — don’t be a bond — is now named as only the monetary half. Plants, grids, and data centers still take decades. Someone must warehouse that interval. AfterFiat is not that someone.

Nothing in the architecture changed. Layers 0 through 6 are not renumbered. The Value Capture Lemma keeps its five conditions. The duration warehouse sits in the diagnosis of the incumbent order, beside the collateral loop, not inside the stack.

The unbundling

Treasuries currently smash two functions together: politically usable collateral, and the conversion of present savings into long-lived capacity. The amendment separates them.

A post-fiat order that cannot tell those functions apart will recreate opaque leverage. A post-fiat order that refuses credit altogether will not build Layer 0. That is now §2.7, and it is the answer to §30.23.

Buyer quality

§2.5 already had Gromen’s leveraged residual buyer. It now has Green’s automatic-versus-marginal structure: a recurring dollar inflow can absorb less DV01 after a selloff because market-value weights shrink. “There is still demand for Treasuries” is not an answer if the remaining demand is a rule rather than a liability. VerifyFlow publishes the distinction as a checklist, not as a scalar.

What this version refuses

It refuses to import a tactical long-bond rally, a mortgage-convexity point forecast, or a unique causal attribution of the latest thirty-year move. Those claims are quarantined in Appendix I. It also refuses the crypto failure mode in which a settlement rail is mistaken for a capital-formation system.