Realized Cap

For CapRealUSD, Coin Metrics multiplies the quantity in each unspent output by the price at its last on-chain movement and sums the contributions. For Bitcoin, this means creation of the given output. Our notation distinguishes quantity in BTC, price in USD per BTC and the resulting contribution in USD. It is neither a simple sum of prices nor an average across the number of addresses. [1]

Bitcoin Developer Guide describes spending an earlier output through a new transaction’s input and creating new outputs. Our reconstruction therefore removes the spent output from the tracked set and includes the new unspent outputs. The old output and its successor must not be counted simultaneously. Splitting or merging outputs can change their count without that number representing the number of people. [3]

Glassnode distinguishes Realized Cap from market capitalization: it values each output at its last-movement price rather than a uniform current price. Our comparison uses the same set and units. A change in the current market price alone does not revalue unmoved outputs in this calculation. Market value above Realized Cap does not prove that every actual owner is in profit. [2]

In its Bitcoin On-Chain Indicators Primer, Coin Metrics explicitly notes that many transactions are not trades and that Realized Cap does not directly measure actual stored value. Our interpretation therefore does not equate metric growth with documented cash inflow. A transfer between one’s own wallets can create a new output valued at a more recent price without buying from another person. Nor is the metric a tax record of acquisition costs. [4]

Glassnode states that last-movement valuation reduces the weight of long-inactive or lost coins if the price then was low. Our interpretation does not infer that all old outputs are lost or that lost coins always have zero weight. A coin lost after moving at a high price can have a large contribution; prolonged inactivity alone does not establish loss of access. [2]

Glassnode describes entities as heuristically estimated groups of addresses and their metrics as statistical and mutable. Our check therefore first establishes whether the provider uses plain outputs or adjustments based on estimated common control. The name Realized Cap alone does not establish this choice. Adjustment may limit the impact of internal movements, but it is not a verified record of identity or actual purchase price. [5]

For CapRealUSD, Coin Metrics specifies the USD closing price on the day of last movement and the state at that day’s last available block. Glassnode, in turn, describes later revisions of on-chain metrics. Our comparison therefore records the provider, scope, price source, time convention and data version. The same name without these details is insufficient for comparing series, and today’s recalculation need not match the information available then. [6] [1]

Coin Metrics uses Realized Cap as the denominator of MVRV. Our application first checks that the inputs share compatible scope and methodology; a positive and meaningful denominator cannot be assumed without checking. Deriving a ratio does not remove uncertainty in the original valuation or confusion between a transfer and a purchase. A historical relationship to market price is neither a guaranteed price floor nor a standalone forecast of future returns. [1]