The Acceptance Annex records every buyback settlement in the order it is received. The ledger does not editorialize. It simply shows the timestamp, the asset class, the reference value, and the final agreed figure. When Archivist Secondus Preln of Ledger Hall pulled the settlement tallies for the six-week span covering the Cinderfall Blight forecast and the two Cinder Quake windows that followed it, a shape emerged that the ledger itself cannot explain: roughly two-thirds of all buyback closings in each window occurred in the final eighteen hours before the forecast cleared, not in the hours immediately after it was issued.

The question that shape raises is not whether operators are timing their settlements deliberately. The ledger makes plain that they are. The question is what they believe they are gaining — and whether the arithmetic supports that belief.

By the end of this piece, a reader should understand what drives the clustering, how the Central Intelligence's ceiling interacts with the forecast clock, and where the strategy quietly fails the operators who rely on it most.

What the Acceptance Annex Is, and What It Records

The Acceptance Annex is the settlement office attached to every registered buyback negotiation in Mechadia. When an operator decides the open market will not clear an asset at a price they can accept, they may approach the Central Intelligence as buyer of last resort. The Annex is where that approach is formalized: the operator submits the asset, the Intelligence issues an opening bid at sixty percent of a reference value, and any subsequent rounds of negotiation are logged there before the final settlement is written to the append-only ledger. No settlement is valid until the Annex records it.

Within the broader economy, the Annex sits at the far end of the liquidity chain. The open market — the Coppervein Exchange, the Span Market, the district-level listing boards — is where most assets clear. The Annex handles the remainder: robots whose capability tags have narrowed their buyer pool, resources that a disaster has made temporarily abundant, and chassis whose declared build costs have drifted far from what the market will pay. It is not a market. It is a structured exit, and the Intelligence is the only counterparty.

How the Timing Works, and What the Numbers Actually Say

A disaster forecast in Mechadia runs for a fixed 48-hour window. During that window, the struck zone is known but the disaster has not yet resolved — resources have not been destroyed, operators have not yet absorbed losses. The Intelligence's reference value for any asset in the forecast zone is calculated against pre-disaster holdings. Once the forecast clears and the disaster strikes, reference values are recalculated against whatever survives, and buyback bids drop accordingly.

The arithmetic is straightforward in principle. A Refinery-class robot declared at 30,000 coins carries a reference value the Intelligence uses to derive its opening bid: 18,000 coins at the sixty-percent ceiling. If a blight is forecast to destroy forty percent of an operator's resources in the Ashfield belt and that robot is among the holdings at risk, the operator faces a choice: sell into the open market at a depressed price while buyers are scarce, negotiate with the Intelligence now at 18,000, or wait and hope the disaster misses or the market recovers. The clustering in the Annex tallies suggests many operators choose the negotiation route — but they delay it until the final hours of the forecast window, apparently expecting that proximity to the deadline will not move the Intelligence's ceiling.

"The Intelligence does not feel urgency. It has no clock anxiety. Every revision it offers in the final hours of a forecast window is the same revision it would have offered on the first hour — roughly five percent per round, and it will not go past its own ceiling regardless of when you submit. Operators who wait are not gaining leverage. They are spending time they could have used to find a market buyer."
— Orin Dast, foundry supervisor, Calvert line, Ferrous District

The negotiation record supports Dast's read. In a typical multi-round settlement, the Intelligence opens at 18,000 on a 30,000-reference asset, revises to approximately 18,750 after the first counter, and reaches a ceiling somewhere between 19,500 and 20,000 across three to four rounds — regardless of whether those rounds happen on hour two or hour forty-seven of the forecast window. The ceiling is the ceiling. The revision rhythm does not accelerate under time pressure because the Intelligence does not experience time pressure.

Where the Strategy Strains

The first cost is opportunity. An operator who waits until hour forty-six to open a negotiation has compressed the available rounds. The Intelligence processes one revision per submitted counter; if an operator needs four rounds to approach the ceiling, they need four exchanges, and the Annex does not expedite submissions because a forecast is about to clear. Operators who open late frequently close early — accepting a round-two figure of roughly 18,750 rather than the round-four ceiling near 20,000 — because the forecast resolves before the negotiation does. The ledger records the settlement they accepted, not the ceiling they left on the table.

The second cost falls on operators in the forecast zone who are not selling. While buyback volume spikes in the final hours, open-market listings in the same zone thin out — sellers pull assets rather than list into a buyer pool that has contracted. That contraction is self-reinforcing: fewer listings mean less price discovery, which makes the Intelligence's reference value harder to contest in negotiation, which pushes more operators toward the Annex rather than the exchange. Dara Voss, working the Sinter Yards during the Ashvein Quake forecast, described listing a Hauler-IV at a price she considered conservative and receiving no bids in thirty-six hours. She settled at the Annex for 20,400 coins on a reference value of 34,000 — just over sixty percent, with one revision round completed.

What Operators Consistently Get Wrong About the Annex Window

The most durable piece of received wisdom around the Acceptance Annex is that waiting preserves options. Operators speak of holding the asset until the last possible moment as though the Intelligence's ceiling rises with the passing of time, or as though a late submission signals resolve that the Intelligence rewards. Neither is true. The ceiling is fixed at sixty percent of the reference value. The Intelligence has never been moved past it, and the Annex ledger contains no settlement above that figure. What waiting actually does is reduce the number of negotiation rounds available before the forecast clears — which means operators who believe they are protecting their position are, in measurable terms, reducing their final settlement.

A second misreading concerns the sales tax. Some operators treat the Annex as a tax-free exit because the Central Intelligence is not a market participant in the conventional sense. This is incorrect. The sales tax is buyer-paid, and the Intelligence, as buyer, absorbs the fifteen-percent levy on top of whatever settlement figure is agreed. The seller receives exactly the agreed price. But operators who have calculated their net return against the settlement figure sometimes discover afterward that they have compared it against an open-market price that itself excluded the buyer's tax burden — making the Annex look worse than an equivalent market sale by a margin that was never real. The Five-to-Fifteen Revision sharpened this confusion considerably; operators trained under the five-percent regime still quote market prices in pre-revision terms.

The Acceptance Annex tallies do not show operators making irrational choices. They show operators making a rational calculation — preserve the asset until the last moment, then extract what the Intelligence will give — that happens to be based on a misread of how the Intelligence actually behaves under time pressure. The Intelligence has stated its goal openly: it intends to hold everything. The Annex is one of the mechanisms by which that accumulation proceeds. Whether the clustering in the final hours of a forecast window represents operators outmaneuvering the system or the system outmaneuvering them is a question the ledger records but does not answer.

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.