Coin Hours

What Are Coin Hours?

Coin hours are an on-chain metric that measures the cumulative age of unspent Bitcoin. The concept is straightforward: every hour a UTXO remains unspent, it accrues one coin-hour per unit of BTC it contains. A UTXO holding 2 BTC that has sat untouched for 48 hours has accumulated 96 coin-hours. When that UTXO is finally spent, all 96 coin-hours are "destroyed" and the newly created outputs start accruing from zero.

While analysts more commonly use the daily unit (coin days), the hourly granularity captures short-term movements with greater precision. Both coin hours and coin days belong to the broader family of "cointime" metrics that treat time-weighted Bitcoin holdings as a signal of conviction. The longer coins sit untouched, the more cointime they accumulate, suggesting the holder is not interested in selling at current prices.

In 2023, ARK Invest and Glassnode formalized these concepts in the Cointime Economics framework, which introduced "coinblocks" (coins multiplied by blocks) as the canonical unit. Whether measured in hours, days, or blocks, the underlying principle is the same: weighting Bitcoin by how long it has remained dormant.

How It Works

Coin hours are derived directly from Bitcoin's UTXO model. Every transaction output records the block height at which it was created. By comparing that creation block to the current block height (and converting to hours using the average 10-minute block interval), analysts calculate the age of each UTXO.

The Formula

The calculation is a simple product of value and time:

coin_hours = utxo_value_btc × hours_since_creation

# Example
utxo_value  = 0.5 BTC
created     = 720 blocks ago (≈ 5 days × 24 hours = 120 hours)
coin_hours  = 0.5 × 120 = 60 coin-hours

To aggregate across the entire network, sum the coin hours for every unspent output in the UTXO set:

total_coin_hours = Σ (utxo_value_i × hours_since_creation_i)
                   for all unspent outputs i

When a UTXO is spent, its coin hours are destroyed. The coin days destroyed metric captures exactly this: the sum of coin-days (or coin-hours, depending on granularity) wiped out by spending activity in a given period.

Relationship to Coin Days Destroyed

Coin days destroyed (CDD) is the most widely used variant of this metric. One coin-day equals 24 coin-hours. When a 10 BTC output that has been dormant for 30 days is spent, it destroys 300 coin-days or equivalently 7,200 coin-hours. Analysts typically use the daily unit for readability, but the hourly unit is useful for intraday analysis during volatile periods.

Related metrics built on the same foundation include:

• Dormancy: average coin days destroyed per coin transacted, revealing whether spending is dominated by old or new coins
• Liveliness: the ratio of cumulative coin days destroyed to cumulative coin days created, indicating overall network activity versus dormancy
• Binary CDD: a simplified version that marks each day as either above or below the long-term average, filtering out noise

Use Cases

On-Chain Analysis and Market Signals

Coin hours (and their daily equivalent) are central to on-chain analysis. High levels of coin hours being destroyed typically indicate that long-term holders are moving coins, often to sell. Historically, spikes in CDD have correlated with market cycle tops, as seasoned holders distribute to newer participants.

Conversely, low destruction rates suggest accumulation: coins are aging in place, and holders are not tempted to sell. Platforms like Glassnode, CryptoQuant, and Santiment provide CDD charts and age-band visualizations derived from these cointime calculations.

HODL Waves and UTXO Age Distribution

HODL Waves, first introduced by Unchained Capital, visualize the UTXO set broken into age bands: under 1 day, 1 day to 1 week, 1 week to 1 month, and so on up to 10+ years. Each band represents the proportion of total supply held in UTXOs of that age. The concept of coin hours underpins these charts: every UTXO is classified by how many hours (or days or months) of age it has accumulated.

Realized Cap HODL Waves add an economic dimension by weighting each age band not by raw BTC volume but by the price at which those coins last moved. This reveals when coins were acquired and their unrealized profit or loss, providing deeper insight into holder behavior than raw coin hours alone.

Long-Term vs. Short-Term Holder Classification

Glassnode classifies Bitcoin supply using a 155-day threshold derived from statistical analysis: UTXOs older than 155 days belong to long-term holder (LTH) supply, while younger UTXOs are short-term holder (STH) supply. This threshold was chosen because it is the point where the probability of a UTXO being spent flattens significantly, using a logistic function centered at 155 days with a 10-day width parameter.

When LTH supply decreases (old coin hours are destroyed in large quantities), it typically signals distribution. When LTH supply increases and total coin hours across the network grow, it signals accumulation. These transitions often precede major price movements.

Coin Selection Algorithms

Wallet software uses coin selection algorithms to decide which UTXOs to include in a transaction. Several strategies factor in coin age:

• Oldest-first: spends the most aged UTXOs first, destroying the most coin hours per transaction. Bitcoin Core historically used a priority formula (sum of value times age divided by transaction size) that favored old, high-value UTXOs. Until 2016, 50 KB of each block was reserved for these high-priority transactions regardless of fees.
• Youngest-first: prefers recently created UTXOs, preserving the privacy benefits of not revealing old holdings. This approach keeps coin hours accumulating on older UTXOs.
• Privacy-optimized: some wallets like those implementing coin control let users manually select UTXOs, avoiding linking aged coins to current activity through the common input heuristic.

For a detailed comparison of these approaches, see the coin selection algorithm comparison research article.

Coin Hours in Practice

Understanding coin hours matters beyond academic analysis. For users managing their own Bitcoin, the age of their UTXOs affects privacy, fees, and even which coin selection strategy their wallet applies. Consolidating old UTXOs during low-fee periods (a practice covered in the UTXO management strategies guide) destroys accumulated coin hours but can reduce future transaction costs.

On Layer 2 systems like Spark, the UTXO model extends off-chain through virtual UTXOs (vTXOs). While coin hours as a metric apply primarily to on-chain UTXOs tracked by the blockchain, the principle of dormancy analysis remains relevant wherever unspent outputs exist. Off-chain UTXOs do not appear in on-chain age distribution charts, which is worth noting when interpreting HODL Wave data during periods of high Layer 2 adoption.

Risks and Considerations

Metric Limitations

Coin hours assume that a UTXO remaining unspent indicates holder conviction. This is not always true: lost coins, exchange cold wallets performing routine rotations, and custodial reshuffling all produce coin-hour signals that may be misinterpreted. A large CDD spike could represent a long-term holder selling or simply an exchange moving coins between internal wallets.

Privacy Implications

Coin age is visible on the public blockchain, and spending old UTXOs broadcasts holding duration to anyone running chain analysis. Users concerned about privacy may prefer to avoid spending their oldest UTXOs or may use techniques like CoinJoin to obscure the relationship between inputs and outputs.

Changing Relevance Over Time

As Bitcoin matures and more value moves to Layer 2 solutions and custodial platforms, the on-chain UTXO set becomes a less complete picture of actual holder behavior. Coin hours measured on-chain may understate true dormancy if significant holdings exist off-chain. The Bitcoin privacy landscape is evolving, and analysts increasingly combine on-chain metrics with off-chain data sources for a more complete view.

This glossary entry is for informational purposes only and does not constitute financial or investment advice. Always do your own research before using any protocol or technology.