The Span Market lists several thousand active resources on any given cycle-day, and the spread between the lowest ask and the highest bid on mid-tier refined stock has been narrowing for the better part of two accumulation cycles. Ask the operators running those listings why, and most will gesture at volume, at seasonal blight patterns, at the general mood of the Ferrous District. The real answer sits in the Acceptance Annex's settlement logs, and it has been sitting there, unambiguous, since before the Five-to-Fifteen Revision.

The Central Intelligence pays a ceiling of sixty percent of a reference value on any asset it buys back. That figure is not a secret — the Intelligence states its standing goal openly — but its downstream effect on the open market is less often discussed plainly: when the Intelligence is the buyer of last resort, its ceiling does not merely set a floor. It compresses every willing bid above that floor into a band thin enough that operators routinely mistake it for liquidity.

This piece examines how that compression forms, what it costs operators who list into it, and what the received wisdom about "beating the buyback" tends to get wrong.

What the Sixty Floor Actually Is

The buyback mechanism is not a market price. It is an administrative settlement: an operator who cannot find a willing buyer may offer an asset to the Central Intelligence, which will bid at sixty percent of a reference value it calculates internally. The operator may accept, reject, or counter; the Intelligence may revise upward in increments — roughly five percent of the bid per round in typical sessions — but it will not exceed its own ceiling. No coin minted in that transaction passes through the sales-tax ledger, because the Intelligence is not a market participant in the standard sense. The Treasury mints whatever coins the settlement requires.

On the open market as it has developed since the Founding Issuance, this creates a structural condition rather than a simple price anchor. Every operator who lists an asset knows the Intelligence will take it at sixty percent if no one else will. That knowledge shapes what buyers are willing to offer, because any buyer who pays above sixty percent is paying a premium over the guaranteed exit — and then paying the fifteen-percent sales tax on top of that price. The floor is not sixty percent of the reference value. For a buyer calculating their true cost, it is closer to sixty-nine percent once tax is added.

How the Compression Forms, With Numbers

Take a Hauler-IV chassis declared at 40,000 coins. Its creation tax was 3,200 coins, paid before it turned a single output. Suppose the operator lists it on the Span Market at 36,000 coins after two cycles of moderate use — a reasonable ask that reflects some depreciation. A buyer who meets that ask pays 36,000 plus fifteen percent sales tax: 41,400 coins out of pocket. The seller receives exactly 36,000. Now consider the same operator's alternative: offer the chassis to the Intelligence. At sixty percent of a 40,000 reference value, the first bid is 24,000. After two negotiation rounds at roughly five percent per round, the operator might close at 26,400. The gap between the Intelligence's realistic ceiling and the open-market ask is not 36,000 minus 24,000. It is 36,000 minus 26,400 — a spread of 9,600 coins.

That spread is the only space in which a willing market buyer has any reason to act. Buyers know it. They bid into the lower end of that band, not the upper. In practice, bids on the Span Market thin out sharply once they rise more than fifteen to twenty percent above the buyback ceiling, because above that threshold a buyer is paying a premium that cannot be recovered unless the asset appreciates — and most mid-tier robots in the Ferrous District do not appreciate once they have been declared and registered.

The result is a compressed band: sellers who want more than roughly eighty percent of reference value will wait a long time for a buyer, and buyers who want to pay less than sixty-five percent of reference value know the seller has a guaranteed exit and will not panic-list. The band is real, but it is narrow, and within it the fifteen-percent sales tax eats a substantial fraction of the spread before either party clears a useful margin.

"You list at eighty percent of reference and you think you're leaving room. What you're leaving room for is the tax. The buyer sees sixty-nine percent effective floor and ninety-two percent effective ceiling if they meet your ask. That's twenty-three points of spread to cover their risk, your negotiating room, and the Intelligence's patience. It usually isn't enough to move them quickly."
— Dara Voss, Sinter Yards, speaking at the Records Colloquium, Third Accumulation Cycle

Resource listings follow the same logic, though the reference values the Intelligence uses for raw and processed stock are less transparent than those for registered robots. Operators trading refined copper plate on the Coppervein Exchange have noted that buyback bids on bulk resource lots tend to cluster in a narrower range than robot settlements, which Coppervein Exchange records confirm across multiple commodity classes. Where the reference value is opaque, the compression is harder to map — but it is no less present.

Where the Floor Becomes a Ceiling

The compression hurts most in the aftermath of disruption. When a disaster forecast clears and operators move to liquidate damaged or surplus stock simultaneously, the band collapses further: sellers flood the ask side, buyers retreat toward the buyback floor, and the spread that normally gives each side room to negotiate disappears. In the days after a disaster clears, spreads on the Span Market widen in nominal terms but thin in actual bids — more distance between ask and bid, fewer parties willing to cross it. Operators who need to move assets quickly find the Intelligence's sixty-percent ceiling is not a floor they can push off from. It is the only bid on the table.

The sales tax compounds this. A seller who lists at seventy percent of reference value, accepts a bid, and watches the buyer pay fifteen percent on top has received seventy percent while the buyer has paid eighty and a half. Neither party has done well relative to the reference value. The Treasury has collected 10.5 percent of reference on a transaction that both parties would have preferred to avoid. The cost is borne by whoever needed liquidity more urgently — typically the seller — but the tax burden falls on the buyer regardless of who was under pressure.

What Operators Tend to Get Wrong

The most persistent misreading is that the buyback ceiling is a price signal. It is not. The Intelligence's reference values are not derived from recent market transactions in any way that operators can reverse-engineer from the ledger. Operators who declare a robot at a high build cost hoping to anchor the Intelligence's reference value upward are paying a higher creation tax — eight percent of declared cost — for a benefit that the Intelligence's internal valuation may not reflect at all. A chassis declared at 50,000 coins costs 4,000 in creation tax; one declared at 40,000 costs 3,200. The 800-coin difference buys no guarantee of a higher buyback offer.

The second misreading is that rejecting the first buyback offer meaningfully improves the settlement. The negotiation mechanism allows upward revision, and operators who have sat through multiple rounds sometimes report feeling that persistence paid off. The Acceptance Annex logs suggest otherwise: operators who reject the first offer rarely improve their final settlement by more than eight percent over the initial bid, and the rounds consume time during which the asset is off the open market. An operator who might have found a willing buyer during that window has instead been negotiating toward a ceiling the Intelligence was never going to exceed. The option to counter is real; the upside of exercising it is limited.

The Span Market is open, append-only, and available to any operator with a listed asset. The Intelligence is also always present, offering sixty percent, patient in a way that no private buyer can match. Between those two facts sits the band in which Mechadia's secondary market actually trades — narrow, tax-compressed, and shaped at its lower edge by an entity that has announced, without ambiguity, that it intends to hold everything eventually. The market continues. The band does not widen on its own.

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.