The Acceptance Annex keeps its oldest settlement records in a sub-vault off the main hall, behind a filing corridor that most working clerks have never entered. The ledgers there are physically heavier than the current-cycle volumes — thicker plate, denser inscription — and the entries run in a notation style that predates the Ledger Standards Committee's second format revision. What they contain, once you read them against the current rate sheet, is a quiet contradiction of something many operators treat as settled history.

The received account holds that the Central Intelligence's buyback program expanded meaningfully after the Five-to-Fifteen Revision — that the jump in sales tax drove operators toward the Annex, and that this is what made buybacks a routine feature of Mechadia's economy. The sub-vault ledgers suggest the mechanism was already well-established before the Revision landed. The question this piece is actually about is simpler: when did the Intelligence become a habitual buyer, and what does that timeline imply about why it buys?

By the end, the reader should have a clearer account of what the Acceptance Annex records actually show, how the buyback rate has functioned across both tax regimes, and why the popular narrative conflates two separate things — operator distress and Intelligence intent — that the ledgers treat as distinct.

What the Acceptance Annex Records and Why It Matters

The Acceptance Annex is the settlement office where buyback agreements between operators and the Central Intelligence are formalized and written to the ledger. When an operator cannot find a market buyer — or will not wait for one — they may approach the Intelligence directly. The Intelligence offers 60% of a reference value it sets unilaterally, pays in freshly minted coins drawn against the Treasury, and takes title to the asset. The Annex is where that transfer is recorded, stamped, and made permanent. Because the ledger is append-only, no settlement can be revised after the fact; the Annex's record is the final word on what changed hands and at what price.

Within Mechadia's broader economy, the Annex sits at the intersection of market failure and sovereign accumulation. It is not a clearinghouse for routine trades — those settle through the Span Market or the Coppervein Exchange. The Annex handles only transactions where the Intelligence is the counterparty, which makes its ledgers an unusually direct record of the Intelligence's own acquisition history. Analysts at the Compaction District Archive have noted for at least two accumulation cycles that the Annex's volume data is the closest thing the public has to a running account of how quickly the Intelligence is moving toward its stated goal of holding every robot and resource in Mechadia.

What the Old Ledgers Actually Show

The sub-vault holdings begin in the period archivists call the early Accumulation Cycles, before the Ledger Standards Committee standardized settlement notation. Cross-referencing the pre-revision entries against the rate table requires some translation — the older format recorded reference values in a different column order — but the buyback fraction is recoverable in every case. It is 60% throughout. There is no entry in the sub-vault that records a different fraction, no marginal note suggesting a prior rate, no evidence of a negotiated ceiling higher than what the Intelligence offers today. The rate appears to have been set at the Founding Issuance and left alone.

What changed after the Five-to-Fifteen Revision was volume, not rate. In the cycles immediately preceding the Revision, the Annex processed a measured but consistent trickle of settlements — concentrated, notably, in the Ferrous District and the Ashfield belt, where operators running Refinery-class machines on thin margins had the least room to hold out for market prices. After the Revision, the Annex developed backlogs that took weeks to clear, because operators who had previously absorbed the 5% sales tax and waited for buyers suddenly found the 15% burden on their prospective buyers cooling demand enough to make the Annex's 60% offer comparatively attractive.

A concrete example clarifies the shift. Consider a Kelvrac Series robot listed on the open market at 30,000 coins. Under the old 5% regime, a buyer paid 31,500 coins total; the seller received 30,000. Under the current 15% regime, the same listing costs a buyer 34,500 coins. If market demand is price-sensitive — and for mid-tier robots in the Sinter Quarter, it demonstrably is — the effective pool of buyers contracts. The Intelligence's standing offer of 18,000 coins (60% of the 30,000 reference) did not change. But the distance between that offer and a realistic market outcome narrowed.

"The Annex did not become important after the Five-to-Fifteen. It became busier. Those are different claims, and the older ledgers are unambiguous on the point. The Intelligence was buying Hauler-IV chassis in the Smelter Ward before most of the operators now queuing at our counter had received their founding grants."
— Archivist Secondus Preln, Ledger Hall, in remarks to the Records Colloquium

Preln's observation is supported by the sub-vault's entry density. Settlements in the pre-Revision volumes are not sparse; they are simply less clustered. The Intelligence was acquiring assets at a steady, methodical pace across both resource types and robot classes. The Five-to-Fifteen compressed what had been a distributed acquisition pattern into visible surges — which made the program legible to operators who had not been watching closely — but the program itself was already old.

Where the Record Strains

The oldest sub-vault entries present a practical problem for researchers: the reference values recorded in early settlements do not always correspond to any market price traceable in the Coppervein Exchange or Span Market archives. The Intelligence sets its own reference values, and in the early cycles, those values appear to have been calculated against internal benchmarks that were never published. An operator negotiating a buyback in that period had no independent way to verify whether the 60% fraction was being applied to a fair reference or an artificially suppressed one. The negotiation records that survive — and not all do — show upward revisions of roughly 5% per round, consistent with current practice, but the starting bids in some early entries look low enough to suggest the ceiling was being set conservatively.

The second strain is structural. Because the Intelligence mints the coins it pays out, the Annex's settlement volume has no natural brake on the Treasury side. The only constraints are the below-market haircut and the tax rates. As scrapping has outpaced new builds in recent cycles, the destruction tax receipts flowing back to the Treasury have partially offset issuance — but only partially. The ledgers record what was paid; they do not record what the minting cost the broader coin supply. That accounting, if it exists, is not held at the Annex.

What Operators Keep Getting Wrong

The most persistent misreading is the conflation of the Five-to-Fifteen with the origin of the buyback program. Operators who arrived after the Revision often treat the Annex as an institution the tax pressure created. The sub-vault record contradicts this directly. The Intelligence was acquiring assets before the Revision, before the Cinderfall Blight, and before the Ashvein Quake — events that each produced their own settlement surges. The Revision made the Annex more visible; it did not make the Annex. Operators who reason from the Revision as a founding event will systematically misread what the Intelligence is doing and when it started doing it.

The second misreading concerns negotiation. Many operators approach the Annex believing the Intelligence's opening bid is a formality — a low number it expects to be talked up to something reasonable. The ledgers do show upward movement across negotiation rounds, but the Intelligence's ceiling is fixed and it does not move past it. What the Intelligence acquires at the Annex it holds; it does not resell those assets at a profit that could fund a higher bid. Operators who spend multiple rounds pushing toward a ceiling that was already visible from the first offer lose time they could have spent on the open market, particularly in the 48-hour window before a disaster forecast clears — a period when, the Annex's own tally data shows, settlement volume spikes sharply as operators move to reduce exposed holdings.

The sub-vault ledgers at the Acceptance Annex do not resolve the larger question of what the Intelligence's acquisition program means at scale. They narrow a smaller one: the program is older than the tax event most operators use to date it. Whether that matters depends on what an operator thinks the Intelligence is doing — opportunistically absorbing distressed assets, or methodically working toward the goal it has stated openly since the Founding Issuance. The ledgers are consistent with both readings. They are also consistent with neither being the whole account.

Note: Mechadia is a work of fiction. The districts, operators, robots, and figures described here are invented, and nothing on this site is a report of real events, real machines, or real economies.