Practitioner framework (ARK × Glassnode, 2023): methodologically clear, not peer-reviewed, no out-of-sample test
ORIGINATED BY David Puell (ARK Invest) & James Check (Glassnode) — ARK × Glassnode whitepaper, 2023. Builds on the liveliness concept introduced by Tamás Blummer (2018).
Most on-chain valuation treats every satoshi as equal. A coin lost in a landfill in 2011, a coin frozen in cold storage since the Genesis block, a coin that changed hands this morning — all counted the same. Cointime Economics refuses that. It reweights the entire ledger by time, asking not merely how many coins exist but how much each one has actually been held. The AVIV Ratio is what falls out: a valuation multiple built to ignore the dead, the dormant, and the never-moved — and to price only the coins doing real economic work.
The unit at the heart of the framework is the coinblock: one bitcoin held for one block. Every block that passes, the network creates coinblocks proportional to the coins outstanding; whenever a coin moves, it destroys the coinblocks it accumulated while sitting still. This is simply a way of accounting for supply weighted by holding time, rather than treating a coin's existence as a single instantaneous fact.
From coinblocks come two mirror quantities. Liveliness is coinblocks destroyed divided by coinblocks created — the fraction of the network's stored time that has been 'spent', a proxy for how much of the supply is genuinely in circulation. Vaultedness is simply 1 − liveliness: the dormant, long-held, 'vaulted' portion of supply that includes lost coins, deep cold storage, and never-moved coinbase rewards. The credit for liveliness goes to Tamás Blummer (2018); Cointime Economics is the framework built on top of it.
The intuition behind the headline metric is this: standard MVRV compares market cap to realized cap (the price at which each coin last moved), but realized cap still carries the weight of coins that are effectively out of the market — lost forever, or held by owners who will never react to price. If those coins do not participate, why should they sit in the denominator of a valuation ratio? Cointime Economics strips them out on both sides.
On the numerator, Active Cap = Market Cap × Liveliness — market value scaled down to only the economically active supply. On the denominator, Investor Cap = Realized Cap − Thermocap, which subtracts the cumulative value paid to miners (their revenue, or 'thermocap') to isolate the value invested by ordinary holders rather than injected by issuance. The AVIV Ratio — Active Value to Investor Value — is the quotient of the two.
Read it like a refined MVRV. A value near 1 means active market value roughly equals investor cost basis — a rough equilibrium. The authors identify an overbought band above 2.5 (250%) and an oversold band below 0.55 (55%), proposing these as historically meaningful tops-and-bottoms thresholds. The pitch is that by excluding lost and dormant coins, AVIV cuts through the noise that distorts vanilla MVRV, giving a cleaner read on where price sits relative to the coins that actually trade.
THE MATHS
coinblock = 1 BTC held for 1 block Liveliness = coinblocks destroyed / coinblocks created; Vaultedness = 1 − Liveliness Active Cap = Market Cap × Liveliness Investor Cap = Realized Cap − Thermocap AVIV = Active Cap / Investor Cap (overbought > 2.5, mean ≈ 1, oversold < 0.55)
LIVE READ
The indicator, as it reads right now.
THE HONEST READ · LIMITATIONS
Where this indicator lies to you.
The framework is methodologically clear but unvalidated. Cointime Economics is a well-constructed practitioner whitepaper from ARK Invest and Glassnode — not a peer-reviewed result. The bands (2.5 overbought, 0.55 oversold) are calibrated on Bitcoin's entire single price history, which is a sample of a handful of cycles at most. There is no out-of-sample test, no walk-forward validation, and no analysis of overfitting. When you fit thresholds to the only history that exists, they will always look prescient in-sample; that tells you almost nothing about the next cycle.
The central marketing claim — that AVIV is more accurate than MVRV — is carefully hedged, not demonstrated. The paper reaches for language like 'seems to', 'perhaps because', and 'according to our research' rather than a formal statistical comparison. There is no reported error metric, no head-to-head backtest with confidence intervals, no null-hypothesis test that would let a reader judge whether the improvement is real or wishful. Tellingly, the standard 'past performance is not indicative of future results' disclaimer appears seventeen times in the document — the epistemic load is carried by a legal disclaimer, not by evidence.
There are also structural fragilities. Liveliness is a slow, near-monotonic aggregate that drifts over years, so the ratio inherits a long, cycle-dependent trend — comparisons across eras are not clean. 'Vaulted' supply lumps genuinely lost coins together with patient long-term holders who could sell tomorrow, so the active/dormant split is an assumption, not an observation. And because ARK held a commercial and directional interest in Bitcoin when publishing, the reader should treat the confidence of the framing with appropriate skepticism. Use AVIV as one lens among several — not as a settled, tested edge.
THE CITATIONS
Read the sources. Check our work.
FIELD TEST · SIGN-IN REQUIRED
Take the quiz. Track the curriculum.
The module text is open to everyone — no account required to read. The 4-question field test and the curriculum-wide graduation stamp need a free account so we can record your progress.
SIGN IN TO TAKE THE QUIZFREE · NEW ACCOUNT TAKES 30 SECONDS