Production Cost estimates what the network spends to bring a coin into existence, and the distance between price and that estimate is the industry’s margin. Price has spent very little of its life beneath the line, which is why it gets treated as a soft floor.

Soft is the operative word. Nothing enforces it, and the reason it behaves like a floor at all is economic rather than mechanical.

What it actually measures

Wide margins invite more capacity, and more capacity competes those margins away. It is the loop the upper end of this chart describes, and it is why comfortable conditions have never lasted indefinitely.

The lower end describes an industry mining at a loss on this estimate. Every major decline has produced such a stretch, and every one of them has been short against the length of a cycle.

The margin overlay collapses the gap between the two lines into a single series. That makes a stretched period and a squeezed one far easier to weigh against each other across different eras.

The floor holds because operators run out of money, not because of a rule

There is nothing in the protocol that stops price falling beneath what it costs to mine. The line holds because miners who cannot cover their bills eventually stop paying them, which removes both their machines and the coins they were selling.

That mechanism is real and it is slow, imprecise and negotiable. Operators with cheap power carry on regardless, hedged ones ride it out, and contracted ones have no choice. The floor bends, which is precisely why it should be read as a region rather than a level.

Fleet X-ray: an unprofitable machine does not stop the same day

Surrender is not a gradual slide. It comes in stages, one vintage at a time dropping beneath the level where keeping it running makes any sense, and identifying which stage is nearest tells you how much more pressure would be required.

The arithmetic says a machine earning less than its electricity should be switched off immediately. Operators frequently do not, and understanding why is most of what it takes to read this view properly.

Cheap power contracts, hedged positions and commitments to deliver keep machines running well past the point where the sums stop working. What the chart marks is where the pressure begins, not where the plug gets pulled, and the gap between those two can run for months.

What it does not tell you

This is a model and not an invoice. Real costs vary enormously between operators, and the cheapest quartile keeps running comfortably below whatever the average says.

Averaging across the industry hides the part that matters. What decides whether machines stop is the cost faced by the marginal operator, and an average is not that.

The estimate rests on assumptions about hardware and electricity that shift over time. It is at its most trustworthy as a broad zone and at its least trustworthy quoted to a specific figure.

How to read it

High Profitability. Margins are wide across the industry, which draws in new capacity and eventually erodes them.

Sustainable Margins. The ordinary state of affairs, with most operators covering what it costs them to run.

Unprofitable Regime. The network is producing coin at a loss on this estimate, a condition that has never lasted long.

Production Cost shares the Mining & Network dashboard with Hashprice, Miner Stress and Fleet X-ray.

Common questions

Why does this behave like a floor?

Because anybody unable to meet their bills eventually stops trying, which takes both their machines and the coin they had been selling out of the market. Nothing enforces it, so the level bends.

Has price gone beneath it?

Yes, during every major decline, and those stretches have been brief against the length of a cycle. Trading below the estimate describes stress and not impossibility.

How exact is the estimate?

Not very, and it should be read as a region. Real costs differ enormously between operators, and the cheapest quartile keeps running well beneath whatever the average says.

What does the margin setting contribute?

It collapses the gap between the two lines into a single series, which makes a stretched period and a squeezed one far easier to weigh against each other across different eras.

Why is an average the wrong number?

Because what decides whether machines stop is the cost faced by the operator closest to the edge. An average sits comfortably above that operator and below the cheapest ones.

What is a shutdown threshold?

Whatever price leaves a particular vintage bringing in less than its power bill. Below there, keeping it running burns money, so the incentive points squarely at the off switch.

How dependable are the thresholds?

They rest on an assumed electricity price, and real prices vary enormously by operator and region. Anybody paying substantially less is comfortable where the chart says otherwise.