Skip to main content

Dominance

Dominance is an asset's share of the total crypto market capitalisation. The page splits the whole market four ways — Bitcoin, Ethereum, stablecoins and everything else — and shows how that split has moved.

The four shares​

Three of the four are measured:

  • Bitcoin and Ethereum dominance come from the market-wide totals feed.
  • Stablecoin dominance is the combined market cap of the assets flagged as stablecoins, over the same total.

The fourth, Other assets, is not measured at all. It is the residual: what is left of the market once the other three are counted. That is deliberate, and it is why every column of the chart adds to exactly 100%. If all four were rounded independently, a column would add to 100.01% on some days and 99.99% on others — and a stack that does not add up is a chart you cannot read the way a stack is meant to be read: a gain in one share is a loss in another.

Each share is carried to two decimal places, and the residual is worked out at that same precision, on the server. Nothing is rounded twice.

A share we could not read is shown as "—" rather than as zero, and the residual is withheld with it: what is left of a total you do not know is not a number.

The 30-day change​

Each card carries the change in its share over the last thirty days, in percentage points. Points, not percent: if Bitcoin dominance moves from 59.0% to 59.4%, it gained 0.4 points. Describing that as a 0.68% rise is arithmetically true and reads as a much smaller move than it was.

The comparison is made against the day nearest thirty days back that we hold a recorded reading for. Where the recorded history does not reach back that far — which it does not for the first month after we start recording — the oldest day we hold is used instead, and the card names it. The label always describes the window the figure was actually measured over: "30d" when it is thirty days, and a date when it is not, in either direction. A comparison is never made against the same day's own reading.

The change is the same figure whichever chart window you are looking at — changing the chart changes what you are looking at, not what "the last thirty days" means.

The four changes always cancel out, because a share can only be gained at another's expense.

Where the history comes from​

We record it. It is not fetched. No free source publishes historical global dominance, so the chart is built from our own daily readings, one per UTC day, starting the day we began recording. Nothing is back-filled, interpolated or estimated:

  • The chart states the window it covers and how many days it actually holds.
  • A day we never recorded is a gap in the line, not a straight segment drawn across it.
  • A window we do not yet have enough history for is offered with the date it becomes available, rather than being drawn over a handful of points.

This is why the 1M window can show fewer than thirty days, and why 3M and 6M unlock as the history accumulates. A chart that started on the day we switched the recorder on but implied six months of history would be the kind of thing that costs a data product its credibility.

Capital rotation​

The rotation panel is the same four 30-day changes, drawn as bars around a zero axis so the direction and the relative size are visible at a glance. It is the same computation as the cards above it — there is no second measurement.

Freshness​

The current split is re-read on the market feed's own schedule. If that stops, the page keeps showing the last reading, says when it was taken, and carries on; a clearly dated reading is a better answer than a blank page. The recorded chart is unaffected — it is a different measurement, taken once a day.

Nothing on this page is a forecast, a target or a recommendation. Every figure describes a measured state.