The area shows the Z-Score (right axis): purple when the market price sits above Realized Cap by more than usual, red when it falls below. The dotted lines mark the mean and standard deviations of the Z-Score itself within the visible range -- not fixed thresholds, they're recalculated when you change the range.
MVRV (Market Value / Realized Value) is already in the cycle Score as a simple ratio: above 1, the market price exceeds the average cost at which each coin last moved. The Z-Score goes a step further: instead of looking at the ratio, it looks at the dollar difference (Market Cap minus Realized Cap) and compares it to how large that difference typically is across history -- so the same MVRV ratio can represent a very different Z-Score depending on how much the market has grown since.
Market Cap and MVRV come from Coin Metrics (free, no key, full history since 2011) -- the same MVRV already used by the cycle Score. Realized Cap is derived by dividing Market Cap by MVRV, and the Z-Score is that difference (Market Cap minus Realized Cap) divided by the standard deviation of Market Cap itself, computed using only data up to that day -- never looking ahead.
With fewer than 30 days of history, standard deviation is noise, not a real reference -- the Z-Score isn't shown until then (from late January 2011 onward). The chart's reference lines are the Z-Score's own mean/deviation WITHIN the visible range, so they change when you switch ranges -- they aren't universal thresholds.
Realized Cap here is a derivation (Market Cap divided by MVRV), not the direct figure -- Coin Metrics doesn't publish it for free. We already have our own Realized Cap, computed from our own node (see the link below), but so far only from 3 point-in-time snapshots -- not yet enough for a full daily series like this one. Once HODL waves' Phase 2 (historical rollbacks) provides enough density, this chart could switch to our own data instead of Coin Metrics' derived figure.
MVRV (the simple ratio, not this Z-Score) is already one of the ten indicators in the NodeWitness cycle Score -- but the Z-Score itself, with its reference lines, isn't part of the calculation yet.