Realized price is Bitcoin's aggregate on-chain cost basis: total realized cap (sum of each Bitcoin's value at the time it was last moved on-chain) divided by circulating supply. Unlike spot price, which reflects the marginal trade, realized price reflects what the network as a whole paid for its current supply. The metric was introduced by Coin Metrics' Nic Carter and Antoine Le Calvez around 2018 and has become foundational to contemporary on-chain analysis. It is the definitional anchor for the MVRV ratio (market cap ÷ realized cap) and NUPL, and a structural component of the SOPR family. Empirically, realized price has acted as a network-wide cost-basis floor that held through the deepest historical drawdowns — the 2015, 2018-2019, and 2022 lows each touched or briefly broke realized price before reversing. It rises smoothly during accumulation and drops only when high-cost-basis supply is moved on-chain, typically at capitulation; the cohort-specific realized prices (LTH, STH) extend the same machinery to subsets of supply.
Why this note matters
Realized price is load-bearing for the on-chain section in three ways:
- It is the definitional anchor for the valuation cluster. MVRV ratio is price ÷ realized price; NUPL is (market cap − realized cap) ÷ market cap; the SOPR family compares spent-output realized prices to acquisition prices. None of these can be built or interpreted without realized price as the foundation.
- It changes the question on-chain analysis answers. Spot price answers “what is the marginal trade?” — realized price answers “what did the network as a whole pay?” That is the empirical operationalization of cost basis at the network level, a quantity that traditional asset analysis can only estimate from fragmentary tax filings.
- It has been a structurally meaningful price level historically. The 2015, 2018-2019, and 2022 cycle lows each touched or briefly broke realized price before reversing. The pattern is not a deterministic floor, but the empirical regularity is substantial enough that serious analysts treat realized price as a structural support level.
It is also the gateway concept for the on-chain framework: once realized price is understood, the rest of the valuation cluster falls into place almost mechanically, and the cohort framework (Long-term vs short-term holder behavior) extends the same machinery to subsets of supply.
What this metric measures
The conceptual claim. Realized price is the average price-per-Bitcoin at which the current circulating supply was last moved on-chain. It is a network-wide proxy for cost basis.
The construction. Every Bitcoin in circulation has a last-moved-at-price. When a UTXO was created, the Bitcoin price at the time of that transaction is recorded (implicitly, by timestamp). Summing across all UTXOs gives the realized cap — the total dollar value of the network’s supply at the prices each portion of supply was last moved. Dividing by circulating supply gives realized price in dollars per Bitcoin.
The contrast with market cap. Market cap is current price × circulating supply — the present-moment marginal valuation extrapolated across all coins. Realized cap is the sum of last-moved valuations — the “what the network paid” valuation. Market cap reflects marginal trades; realized cap reflects accumulated cost basis. They diverge most at cycle peaks (market cap >> realized cap) and converge at cycle troughs (market cap → realized cap, sometimes below).
What it does not measure. Realized price is not the true cost basis of original purchases — when a coin moves from wallet A to wallet B at X regardless of whether the move is a genuine sale, a self-custody transfer, an exchange deposit, a UTXO consolidation, or a transaction-batching operation. The metric is best read as a proxy for cost basis, with known noise from custodial-and-internal movements. The proxy is good enough for cycle-positioning purposes; it is not precise enough to be treated as the literal economic cost basis of individual holders.
How it’s calculated
The canonical formula:
where:
- is the size of the -th UTXO (in BTC)
- is the timestamp at which the -th UTXO was created
- is Bitcoin’s USD price at
- The sum runs over all currently-unspent UTXOs
Then:
Equivalent formulations. Some platforms compute realized cap by aggregating at the transaction level rather than UTXO level; the values match in the limit because each UTXO can be traced to a creating transaction.
Data-provider variants. Glassnode, Coin Metrics, and Checkonchain each have slightly different methodological choices about (a) which price source to use for , (b) how to handle multi-output transactions, (c) how to handle dust and Coinbase-mining UTXOs, and (d) which supply definition to use as the divisor. The values are very close but not identical across providers. For cycle-positioning purposes, the differences are not material; for high-precision research, the methodological details matter.
The dynamics. Realized price moves only when UTXOs are created or destroyed:
- A new UTXO is created when a transaction spends old UTXOs and produces new ones. The new UTXO’s is the current price; the spent UTXOs’ contributions to realized cap are removed.
- Net effect of a transaction. If the spent UTXOs were created at prices lower than the current price (the typical case in an uptrend), realized cap rises — the coins’ cost basis is being reset upward. If spent at prices higher than current price (typical at deep bottoms), realized cap falls — high-cost-basis coins are capitulating.
This dynamic explains why realized price rises smoothly during accumulation phases (modest UTXO churn at gradually rising prices) and falls only during capitulation events (forced selling of high-cost-basis coins). The metric is structurally low-amplitude relative to spot price.
What it tells you
Cycle positioning at the structural-support level. Spot price oscillates dramatically; realized price moves smoothly. The gap between them — the MVRV ratio — is the primary cycle-positioning signal. Realized price itself is most informative as the structural-support reference: where would the price be if all unrealized gains were marked to zero?
Network-wide cost basis. Realized price answers “what did the network as a whole pay?” The answer is typically a small fraction of spot price during euphoria and approximately equal to spot price during deep capitulation. The history:
- 2015 low: Spot price touched approximately realized price at $230-260
- 2018-2019 low: Spot price briefly dipped below realized price around $3,200-3,800
- 2022 low: Spot price briefly dipped below realized price around $15,000-17,500
In each case, the brief sub-realized-price period coincided with maximal capitulation and was followed by reversal. The pattern is empirical regularity, not protocol-enforced floor.
Cohort-specific cost bases. The same machinery can be applied to subsets of supply — long-term holders only, short-term holders only, miners only, etc. The cohort-specific realized prices (Long-term vs short-term holder behavior) are operationally more informative than the aggregate because they separate the conviction-based LTH cost basis from the reactive STH cost basis. LTH realized price tends to act as a stronger support floor than aggregate realized price; STH realized price tends to act as resistance during recoveries.
Macro context. Realized price growth rate is a slow-moving measure of accumulation intensity. Rapid realized-cap growth during a bull market signals new capital flowing into the network at elevated prices; slow realized-cap growth during accumulation signals patient long-term-holder positioning.
Empirical track record
Historical regularity of realized-price support. Across Bitcoin’s history, spot price has touched or briefly broken realized price at major cycle lows. The pattern:
| Cycle low | Approximate spot | Approximate realized price | Behavior |
|---|---|---|---|
| Jan 2015 | $200 | $230 | Spot briefly below realized; ~9 months of churn before reversal |
| Dec 2018 | $3,200 | $3,800 | Spot briefly below realized; ~3 months before reversal |
| Nov 2022 | $15,500 | $19,800 | Spot briefly below realized; ~3 months before reversal |
The 2022 episode is the clearest recent test: spot price spent multiple weeks below realized price during the FTX collapse and Three Arrows aftermath, but the period was brief and the reversal was substantial.
Realized price as cycle anchor. In bull markets, realized price grows steadily as accumulation continues. By cycle peak, realized price is typically 1/2× to 1/4× of spot — i.e., MVRV is 2-4× at peaks (see MVRV ratio). In bear markets, the gap closes from both sides: spot falls, and realized price slowly rises if accumulation continues then begins to fall at maximum capitulation.
Out-of-cycle behavior 2024+. The post-2024 cycle has produced new dynamics: substantial ETF flows redirect Bitcoin into custodial wallets, which complicates UTXO-level cohort analysis. Realized price aggregate has continued to function as a coherent metric, but the cost-basis content of large-ETF UTXOs is somewhat different from the cost-basis content of self-custodial UTXOs (ETF coins were typically purchased at near-current-spot prices by the ETF mechanism). The metric still works, but the interpretation needs adjustment.
Cross-validation across data providers. Glassnode, Coin Metrics, and Checkonchain produce slightly different realized-price values (typically within ~2% of each other). The differences come from price-source choices and UTXO-handling details. For cycle-positioning purposes, the agreement is more than sufficient.
Limitations
The cost-basis approximation is imperfect. Realized price treats every UTXO movement as a price-resetting event. In reality, many UTXO movements are not genuine sales:
- Self-custody transfers between a user’s own wallets reset the on-chain timestamp without any economic transaction
- Exchange deposits and withdrawals reset timestamps without changing ownership
- UTXO consolidation (combining many small UTXOs into fewer larger ones) resets timestamps at current price
- Transaction batching by exchanges and miners produces many UTXO-reset events that don’t reflect new acquisitions
The result is that realized price systematically overestimates the true average cost basis of holders — it captures the most-recent movement, not the most-recent economic acquisition. The bias is small enough that the metric remains useful, but the literal interpretation should be qualified.
ETF and custodial-wallet distortion (2024+). The ETF era has introduced large custodial wallets whose UTXOs were typically created near current spot prices via the creation/redemption mechanism. These UTXOs inflate aggregate realized cap in ways that don’t reflect “the network paid this” in the historical sense. The aggregate realized price remains useful but is increasingly a composite of historical-cost-basis self-custody supply and near-spot-cost-basis ETF supply.
No protocol-level floor. Realized price is not a guaranteed support level. The 2015, 2018-2019, and 2022 episodes saw spot price briefly trade below realized price. A sufficiently severe bear market could see spot price trade substantially and persistently below realized price. The historical regularity is empirical, not structural — it reflects holder psychology (capitulation typically exhausts at “I paid for it” levels), not a hard constraint.
Lagging metric. Realized price moves only when UTXOs move. During quiescent periods, realized price drifts slowly even as spot price moves dramatically. The lag is a feature for structural-support analysis but a limitation for short-term signal generation.
Cohort definition matters. Aggregate realized price is less operationally useful than cohort-specific variants. The LTH realized price (long-term-holder-only realized price) and STH realized price (short-term-holder-only realized price) carry more signal because they separate the cohorts with different behavioral dynamics. The cohort framework requires the additional 155-day threshold convention; see Long-term vs short-term holder behavior.
Methodological-choice sensitivity. Different price sources, UTXO-handling rules, and supply definitions produce slightly different realized-price values. For research-grade work, methodological documentation matters. For cycle-positioning, the differences are negligible.
Counter-arguments and tensions
”Realized price is just historical-data curve-fitting”
The argument: The historical regularity of realized-price support could be coincidence. Three data points (2015, 2018-2019, 2022) is not enough to establish the metric as a structural floor. Future cycles may see persistent sub-realized-price trading.
Response: Partially right. Three episodes is not a large sample, and the metric has no protocol-level enforcement. The honest reading is that realized price has empirically functioned as a capitulation-completion marker because that is the price level at which the median holder is at break-even — a psychologically meaningful threshold that produces the observed behavior. The mechanism is plausible (holders capitulate when underwater), but the future regularity is not guaranteed.
The UTXO-timestamp confound
The argument: Realized price treats every UTXO movement as a price-reset event, but most UTXO movements are not economic acquisitions. The metric systematically overestimates the true cost basis. Conclusions drawn from realized-price-based metrics inherit this bias.
Response: Real concern. The bias is well-known to the on-chain analytical community and is part of why cohort-specific metrics (LTH realized price especially) are preferred for high-stakes analysis. LTH coins have not moved in 155+ days and so are less contaminated by recent custodial-transfer noise. For aggregate-cluster analysis, the bias produces a small upward distortion; for cohort analysis, it largely cancels.
”ETF era breaks the metric”
The argument: Post-2024 ETF flows have introduced large custodial wallets whose realized-cap contribution doesn’t reflect historical-cost-basis accumulation. The aggregate metric is contaminated; comparison with pre-ETF cycles is no longer valid.
Response: Partially right but overstated. ETF flows do change the composition of the realized-cap aggregate. But the ETF inflows have generally been at gradually-rising spot prices over multi-year accumulation, which is similar to the pattern of self-custodial accumulation in prior cycles. The metric remains useful; the calibration needs adjustment. The cohort framework partially addresses this — separating long-term-holder realized price from aggregate realized price recovers most of the lost signal. Future work may need ETF-aware cohort definitions.
”Cohort-specific realized prices are more informative than aggregate”
The argument: Aggregate realized price is less operationally useful than LTH realized price or STH realized price. The aggregate is a blend of cohorts with different behavioral dynamics; the blend loses information. Why build a note on the aggregate rather than going directly to cohort metrics?
Response: Substantially right. Aggregate realized price is foundational but not the most operationally useful variant. This note presents the aggregate as the foundational concept; the cohort-specific variants are the deployment-ready forms. The reader’s path is: understand aggregate realized price → understand cohort decomposition (Long-term vs short-term holder behavior) → deploy cohort-specific realized prices for actual positioning.
Cointegration-style critique
The argument: Realized price and spot price are both trending non-stationary series. Their ratio (MVRV) is bounded only because spot mean-reverts toward realized over cycles; the relationship’s apparent stability could be spurious in formal cointegration tests.
Response: The relationship survives the cointegration test in practice because realized cap is constructed from spot prices over time and so shares a common trend with spot — the relationship is mathematically built in rather than statistically inferred. The ratio MVRV is genuinely mean-reverting on cycle timescales. The critique applies more to free-standing time-series regressions than to ratio metrics like MVRV; it is a methodological caution but not a refutation.
”What about quantum-vulnerable lost coins?”
The argument: A substantial fraction of Bitcoin is genuinely lost (forgotten keys, dead holders without inheritance). The realized-cap calculation includes these coins at their last-moved price. If a future event (quantum advance, key-recovery breakthrough) suddenly enabled access to lost coins, realized price would shift dramatically without any new economic activity.
Response: Speculative tail risk. The current calibration of realized price reflects the lost-coin overhang implicitly. A quantum-recovery event would represent a regime change for all on-chain metrics, not just realized price. The framework’s predictive content assumes the existing lost-coin distribution remains stable.
Open questions for further development
- How should the metric be calibrated for the post-ETF regime? ETF flows change the cost-basis composition of the aggregate. Specific calibration adjustments — perhaps an ETF-aware cohort framework — would strengthen the metric for forward analysis.
- What is the precise behavioral mechanism producing realized-price support? The “median holder at break-even capitulates last” mechanism is plausible but not rigorously documented. Empirical studies of the actual holder population at deep bottoms would test the mechanism.
- Should the metric use spot-USD or some other denomination? All cost-basis metrics implicitly assume USD-denominated holding decisions. For holders in other currencies, or for long-horizon analysis that should be agnostic to fiat-denomination shifts, alternative denominations (gold-grams, global-monetary-aggregate fractions) may be more informative.
- How does the metric interact with hyperinflation scenarios? If the dollar inflates dramatically, realized-price-based metrics that compare spot to historical cost basis may become misleading. The framework needs a regime-change adjustment for major-currency-collapse scenarios.
- What is the appropriate cohort-decomposition framework for the post-ETF era? The 155-day LTH/STH threshold is a useful approximation; a richer cohort framework (custodial-ETF, institutional-self-custody, long-term-individual, short-term-trader) may be more informative.
- Could realized price be derived as a stable function of network adoption? The The Power Law model derives spot price as a function of time-since-inception; an analogous derivation for realized price would integrate the on-chain layer with the longer-horizon trajectory framework.
Canonical sources for this note
Primary framework sources
- Nic Carter and Antoine Le Calvez, “Introducing Realized Capitalization” (Coin Metrics, 2018) — the canonical introduction of realized cap and realized price as on-chain metrics
- Coin Metrics State of the Network reports — ongoing realized-cap and related metric documentation
- Glassnode Realized Price documentation — methodological detail and ongoing analysis
- Checkonchain platform — James Check’s analytical framework built on realized-price machinery
Practitioner literature
- Various James Check essays and Week On-Chain newsletters during the Glassnode tenure — applied realized-price analysis across multiple cycles
- Various Ryan (On-Chain Mind) videos and Substack analyses — accessible realized-price framework presentation
- Various Bitcoin Magazine and Bitcoin Layer pieces engaging realized-price-based analysis
Adjacent on-chain literature
- David Puell, “MVRV Ratio” original work (Murad Mahmudov and David Puell) — the canonical MVRV introduction that defined the realized-price-based valuation framework
- Various Glassnode research pieces on cohort-specific realized prices
Theoretical and adjacent
- Adam Levitin and various legal-academic literature on cost basis in regulated assets — useful comparison context for what cost basis means in non-Bitcoin settings
- Various accounting and tax literature on cost-basis tracking — for contrast with Bitcoin’s transparent on-chain alternative
Related notes
- On-chain analytics and market psychology — sub-MOC parent
- MVRV ratio — directly built on realized price; the flagship cycle-positioning metric
- NUPL — alternative valuation metric built on realized cap and market cap
- SOPR — realized-side analog; uses cost-basis machinery on spent outputs
- HODL waves — cohort framework that extends realized-price machinery to age cohorts
- Long-term vs short-term holder behavior — 155-day cohort framework producing LTH and STH realized prices
- Coin Days Destroyed — supply-weighted velocity; complementary cohort-behavior metric
- Whale behavior — large-holder positioning; cohort framework extension
- Exchange flows — custodial-flow framework; complements cost-basis framework
- Miner flows — miner-cohort framework
- Sentiment indicators — off-chain sentiment proxies
- Psychological phases of the market cycle — cycle-phase synthesis where realized-price extremes mark phase transitions
- Using on-chain data for macro positioning — operational bridge to macro frameworks
- The Power Law model — longer-horizon trajectory framework; realized price as the structural-support reference within the Power Law corridor
- Four-year halving cycles — cycle structure realized-price extremes anchor
- Bitcoin and global liquidity — macro framework realized-price extremes integrate with
- Diminishing returns thesis — cycle-over-cycle attenuation framework realized-price extremes inform
- Monetization S-curve — adoption framework realized-price growth empirically operationalizes
- Bitcoin fixed supply and issuance schedule — supply foundation
- The halving - Mechanism — schedule that affects realized-price growth dynamics
- James Check — primary contemporary anchor; realized-price framework developer
- Ryan - On-Chain Mind — adjacent contemporary anchor; accessible realized-price presentation
- Dylan LeClair — adjacent on-chain voice
- Portfolio approaches to Bitcoin — practical allocation framework realized-price signals inform