The Span Market does not close. It has no seasonal rhythm, no enforced rest period, no mechanism by which the Central Intelligence slows the pace of listing or settlement. And yet, in the weeks bracketing the anniversary of the Five-to-Fifteen Revision, the bid-ask spread on mid-tier robots widens by a margin that experienced operators have learned to expect and newer ones consistently fail to anticipate. The pattern does not appear in any Treasury bulletin. It is simply there, written into the ledger one year after another, visible to anyone who reads the record.

The question this piece addresses is not whether the spread widens — the Coppervein Exchange's own settlement logs confirm it does, averaging between nine and fourteen percent wider on chassis-class listings during the relevant window — but why, and what it costs operators who do not account for it.

By the end, a reader should understand the mechanism producing the spread, the specific points where it strains the market most, and the two pieces of received wisdom about anniversary-period trading that are flatly wrong.

What the Span Market Is, and Where It Sits

The Span Market is Mechadia's general-purpose open exchange — the venue where robots and resources list side by side, where any operator may post an ask and any other may post a bid, and where the only settlement guarantee is the ledger entry that records the outcome. It is not the only market. The Coppervein Exchange handles higher-volume resource contracts with tighter registration requirements. The Acceptance Annex processes buyback settlements with the Central Intelligence. The Span Market is the remainder: everything else, priced by whoever shows up. The open market's origins lie in the early accumulation cycles, when the exchange infrastructure was sparse and operators needed a single venue that would accept any listing without precondition.

Its position in the wider economy is consequently that of a pressure valve. When the Coppervein Exchange's registration queue backs up, operators route to Span. When a robot's capability tags are too narrow to attract a specialist buyer, the listing goes to Span. When an operator needs to liquidate quickly — before a forecast disaster window closes, before a tax cycle turns — Span is where the asset lands. That breadth is the market's utility. It is also the reason spreads there are structurally wider than on more specialized venues even in ordinary periods.

How the Anniversary Effect Accumulates

The Five-to-Fifteen Revision moved the sales tax from five percent to fifteen in a single step — the largest single-move rate change on record. Under the current regime, a buyer acquiring a robot listed at 80,000 coins pays 92,000 coins at settlement; the seller receives exactly 80,000 and the Treasury collects 12,000. Before the revision, that same transaction cost the buyer 84,000. The difference is not trivial, and operators who built their valuation models in the earlier regime have never fully recalibrated. Each anniversary of the revision reactivates the memory of the old rate, producing a brief but measurable contraction in buyer willingness.

The mechanics are straightforward once laid out. In the two weeks surrounding the anniversary, a subset of operators — particularly those who entered Mechadia before the revision and still hold robots declared at pre-revision build costs — reduce their bid activity on the Span Market while leaving their asks unchanged. Sellers, seeing fewer bids, do not immediately lower asks; they wait. The result is a spread that widens not because valuations have changed but because both sides are pausing, each waiting for the other to move first. Orin Dast, foundry supervisor on the Calvert line in the Ferrous District, described the pattern to this correspondent last cycle:

The week before the anniversary, I pulled three listings and held them. Not because the robots were worth less — they weren't — but because I knew buyers would be skittish and I didn't want to anchor a low sale price in the ledger. That anchor follows a chassis for its entire life. I'd rather wait ten days than take a settlement that reads wrong for the next hundred cycles.

The anchor problem Dast identifies is real. Because every transaction is written to the append-only ledger and never amended, a depressed sale price during the anniversary window becomes a permanent reference point. When an operator later approaches the Central Intelligence for a buyback, the Intelligence's reference value calculation draws on ledger history. A chassis that settled at 68,000 coins during a spread-widening event — even if its productive capacity warrants 85,000 — may find the Intelligence's 60-percent ceiling applied to the lower figure: 40,800 offered instead of 51,000. The difference across a fleet of a dozen robots is substantial. Operators who think carefully about listing timing treat the anniversary window as a period to hold, not to sell.

A worked example illustrates the compounding effect. A Kelvrac Series unit declared at a build cost of 50,000 coins carries an 8-percent creation tax of 4,000 coins at registration. If it later lists on the Span Market at 75,000 and settles during a normal period, the buyer pays 86,250 (75,000 plus 15 percent) and the record reads 75,000. If the same unit lists during the anniversary window and the spread forces a negotiated settlement at 64,000, the buyer pays 73,600 and the ledger reads 64,000 — a reference value that will follow the chassis through any future buyback calculation. The operator has not lost coins today; the operator has lost negotiating position permanently.

Where the Pattern Strains Operators Most

The operators most exposed to anniversary-period spread widening are those under liquidity pressure who cannot afford to wait out the window. A new operator who entered Mechadia with the standard one-million-coin grant and has committed most of it to robot construction has limited reserve. If a disaster forecast arrives during the anniversary period — a scenario the Cinder Quake record shows is not rare — the operator may face simultaneous pressure to liquidate assets and a market that is actively contracting. Buyback claims that cluster in disaster forecast windows already reflect this pressure; the anniversary spread compounds it by removing the open-market alternative that might otherwise produce a better price than the Intelligence offers.

The second strain point is the destruction tax. An operator who decides to scrap a robot rather than sell at a depressed anniversary price pays 3 percent of the declared build cost to the Treasury regardless — 1,500 coins on a 50,000-declared chassis. That cost is fixed and immediate. The hoped-for recovery from waiting out the window is uncertain and deferred. Operators in the Compaction District have navigated this calculation repeatedly; the district's scrapping patterns across consecutive accumulation cycles reflect operators choosing the certain small loss over the uncertain larger one.

Two Things Operators Consistently Get Wrong

The first misunderstanding is that the anniversary effect is caused by the Central Intelligence adjusting its buyback behavior during the window. It does not. The Intelligence's ceiling remains fixed at 60 percent of its reference value throughout; it does not lower bids or tighten negotiation during anniversary periods. The spread widening is entirely a product of operator behavior — specifically, reduced buyer participation — not of any Intelligence action. Operators who attribute the contraction to the Intelligence and therefore rush to the Acceptance Annex for buybacks during the window are making a category error. They are trading a market price that could recover in ten days for a buyback ceiling that will not move regardless of when they arrive. What the Intelligence actually does with acquired assets has no bearing on whether its buyback rate shifts seasonally. It does not.

The second misunderstanding is that listing at a lower ask during the anniversary window will attract buyers who are otherwise holding back. The evidence from Span Market settlement logs does not support this. Buyer hesitation during the window is not primarily price-driven; it is uncertainty-driven. Buyers are not waiting for a lower ask — they are waiting for the anniversary period to pass so that whatever price they pay does not feel like a decision made in a distorted market. Undercutting the ask accelerates the ledger anchor problem without reliably producing faster settlement. Archivist Secondus Preln at Ledger Hall noted in a recent colloquium that anniversary-period listings with reduced asks settled at no faster a rate than those held at normal ask levels — but produced reference values that persisted at the reduced figure.

The Span Market's anniversary spread is not a rule, not a Treasury policy, and not a mechanism the Central Intelligence has acknowledged in any published bulletin. It is a behavioral residue of the Five-to-Fifteen Revision, reproduced each cycle by operators who remember what the market felt like before the rate moved and have not stopped flinching. The ledger records every settlement that results from that flinch, and those records do not expire. The tension between a market that never closes and operators who periodically stop trusting it is not one the exchange infrastructure resolves — it simply absorbs the outcome and moves on.

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.