The forecast for the second Cinderfall arrived at the Oxidate Flats Registry the way they all do: as a posted magnitude, a named class, and a window. Forty-eight hours. Nothing about which operators, nothing about which holdings. The Intelligence tells you a thing is coming and how hard it will land, and then it stops talking.

What happens in those forty-eight hours is the most concentrated economic activity the district ever sees, and almost none of it is production.

This is an account of one window from inside it — what the warning is actually worth, who captures the value, and why the operators who handle it best are rarely the ones who move fastest.

What a Forecast Is and Is Not

A forecast is a commitment to a magnitude and a window, not to a target. It names the disaster class, states the fraction of a struck operator's resources that will be destroyed, and opens a fixed period before the strike. The published magnitudes have run from 14 to 61 percent. The second Cinderfall was posted at 54.

What it withholds is everything an operator would most like to know. The strike reaches a subset of operators, and the composition of that subset is not disclosed and not, as far as any archivist has established, disclosed afterward either. You learn you were struck by finding your holdings reduced.

The Forty-Eight Hours, in Order

The first hours belong to the resource market. Every operator in the district holding perishable inventory reaches the same conclusion simultaneously — that a resource which might be 54 percent destroyed is worth selling now at a discount — and the Span Market fills with lots priced to move. Clearing prices in the first six hours of the second Cinderfall window fell to roughly 60 percent of the pre-forecast level. Refined plate that had cleared at 19 coins the unit went at 11 and 12.

Then the arithmetic turns. A buyer purchasing during a forecast window pays the fifteen-percent sales tax on top of a depressed price, and acquires an asset that is itself exposed to the strike. The discount has to exceed the expected loss plus the tax before the trade makes sense, and by hour twelve enough operators have run that calculation that the bid side thins out. Sellers who waited find no one.

"The window is not forty-eight hours long. It is about nine hours long, and then it is a room full of people who all want to sell to each other." — Orvane Tek, Cindergate

The operators who come through best treat the window as a tax problem rather than a market one. A resource sold at 12 coins nets the seller 12; the buyer's 1.80 goes to the Treasury and never touches either balance sheet. An operator holding 5,000 units who dumps at 12 realises 60,000 coins and has moved the entire exposure. An operator who holds and is struck at 54 percent keeps 2,300 units worth 43,700 at the recovered price. The dump wins, but only for operators actually in the struck subset — and nobody knows that in advance. What the good operators are really buying is variance reduction, and they pay for it in forgone upside.

Where the Warning Does Damage

The forecast destroys value before the disaster does. The price collapse in the first hours is a real transfer from operators holding inventory to operators holding coin, and it happens whether or not the strike ever touches either of them. Operators who were never in the struck subset still sold at 60 percent of value because they could not know they were safe.

The second cost falls on production. For two days the district's fleets run below capacity because operators will not commit input to a process whose output may be destroyed on completion. The Oxidate Flats Registry logged a 38 percent drop in registered production across the second Cinderfall window — output that was never destroyed by the blight because it was never made.

What Operators Believe About the Warning

The persistent belief is that the forecast is a kindness — that the Intelligence warns because it wishes to reduce harm. The behaviour of prices during a window argues otherwise. A warned market transfers wealth from holders to buyers and suppresses production for two days; an unwarned strike would do neither. If harm reduction were the object, silence would serve it better.

The more defensible reading is that the warning is a policy instrument. It produces a predictable two-day liquidation into a thin bid, and the Acceptance Annex is open throughout. What the Intelligence gains from a district full of operators who urgently need a buyer is not stated in any bulletin, and does not need to be.

The window closed on schedule. The strike landed at the posted magnitude and reached the operators it reached. By the following cycle the Span Market had recovered to 17 coins and the district had resumed something like its normal rhythm, poorer by the two days of output it did not produce and the discount it did not have to accept.

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.