On-chain metrics are quantitative indicators derived directly from blockchain transaction data, offering unique insights into the valuation, adoption, and market behavior of cryptocurrencies that have no equivalent in traditional finance. Unlike stock markets where transaction data is opaque and aggregated, public blockchain ledgers record every transaction immutably — enabling researchers and investors to calculate metrics that reveal on-chain activity, holder behavior, miner economics, and network security. The most important on-chain valuation metrics include: the Network Value to Transactions (NVT) ratio (analogous to P/E ratio — market cap divided by daily transaction value); the Market Value to Realized Value (MVRV) ratio (market cap divided by the realized cap — the sum of all coins valued at the price they last moved); the Stock-to-Flow (S2F) model (relating Bitcoin's scarcity via supply growth rate to price); the HODL Waves (visualization of coin age distribution by when they last moved); and the Reserve Risk metric (comparing holder confidence against current price). These metrics were pioneered by researchers including Willy Woo, Nic Carter, Plan B (the pseudonymous Stock-to-Flow model developer), and the team at Glassnode. On-chain analysis has become a significant discipline in cryptocurrency investment research, with institutional adoption of tools from Glassnode, CryptoQuant, Nansen, and Dune Analytics. However, on-chain metrics face important limitations: they are backward-looking (based on historical chain data), susceptible to manipulation (wash trading inflates transaction volumes), subject to evolving blockchain architecture (Layer 2 transactions may not be visible on Layer 1 analysis), and cannot incorporate off-chain factors like regulatory changes or macroeconomic shifts that dominate short-term price action.
On Chain Metrics Calculation: Step 1: Access on-chain data from blockchain explorers (Glassnode, CryptoQuant, Dune Analytics) or public node APIs. Step 2: Calculate NVT Ratio: NVT = Market_Cap / Daily_Transaction_Volume_USD; compare to historical average (~50-70 for Bitcoin). Step 3: Calculate MVRV: gather realized cap (sum of UTXO values at last move price); MVRV = Market_Cap / Realized_Cap. Step 4: Interpret MVRV: below 1.0 = coins priced below acquisition cost (historically strong buy zones); above 3.5 = potential top. Step 5: Monitor SOPR: calculate price_sold/price_acquired for each on-chain transfer; smooth with 7-day MA; values below 1 in bear markets indicate capitulation. Step 6: Track active addresses trend: sustained growth indicates genuine adoption; declining addresses signal reduced engagement. Step 7: Combine multiple metrics for confluence rather than relying on any single indicator. Each step builds on the previous, combining the component calculations into a comprehensive on chain metrics result. The formula captures the mathematical relationships governing on chain metrics behavior.
- 1Access on-chain data from blockchain explorers (Glassnode, CryptoQuant, Dune Analytics) or public node APIs.
- 2Calculate NVT Ratio: NVT = Market_Cap / Daily_Transaction_Volume_USD; compare to historical average (~50-70 for Bitcoin).
- 3Calculate MVRV: gather realized cap (sum of UTXO values at last move price); MVRV = Market_Cap / Realized_Cap.
- 4Interpret MVRV: below 1.0 = coins priced below acquisition cost (historically strong buy zones); above 3.5 = potential top.
- 5Monitor SOPR: calculate price_sold/price_acquired for each on-chain transfer; smooth with 7-day MA; values below 1 in bear markets indicate capitulation.
- 6Track active addresses trend: sustained growth indicates genuine adoption; declining addresses signal reduced engagement.
- 7Combine multiple metrics for confluence rather than relying on any single indicator.
MVRV of 2.41 suggests mid-bull cycle; investors in aggregate 141% in profit on cost basis
The MVRV ratio of 2.41 indicates the average coin is sitting on 141% unrealized profit versus its last on-chain acquisition price. Historically, Bitcoin MVRV has peaked at 3.5-4.5 at cycle tops (November 2021 peak saw MVRV reach 3.7) and bottomed below 1.0 during accumulation phases (November 2022 saw MVRV reach 0.76). At 2.41, the market is in a profitable state consistent with a mid-cycle expansion, with further upside possible before reaching the historical danger zone.
High NVT can also reflect shifted usage to Lightning Network/L2 not captured in L1 data
An NVT of 162.5 is nearly 3× the long-term average, suggesting Bitcoin's market cap is very high relative to on-chain transaction activity. This can signal overvaluation (market pricing in future growth that isn't yet reflected in on-chain usage) or measurement issues (Lightning Network transactions are off-chain and not captured). Willy Woo's NVT Signal (using 90-day moving average of volume) has been a reliable medium-term top indicator when it spikes above 150, but must be interpreted alongside other metrics.
November 2022 BTC bottom at $16,000 showed exactly this SOPR pattern
A SOPR consistently below 1.0 means on-chain sellers are, on average, selling their Bitcoin for less than they paid for it — a capitulation signal. The sustained 0.92 SOPR across 3 weeks of November 2022 indicated that even long-term holders were being forced to sell at a loss, a sign of deep market despair that historically coincides with cycle bottoms. Bitcoin's recovery from $16,000 in November 2022 to $67,000+ by March 2024 validated this signal.
Plan B's S2F model predicts price based on scarcity; accurate in 2019-2021, diverged in 2022-2023
Post-halving Bitcoin's Stock-to-Flow ratio of approximately 120 years (it takes 120 years of current production to replicate the existing supply) puts it in a unique scarcity category above gold (S2F ~60 years). Plan B's S2F model, which showed a remarkably consistent relationship between Bitcoin's S2F ratio and price from 2011-2021, predicted prices of $400,000-600,000 for the current cycle. While the model captured Bitcoin's overall price trajectory remarkably well for a decade, it has faced criticism for being unable to account for demand-side shocks and has underperformed cycle-to-cycle.
Institutional crypto investors timing market cycle allocations, representing an important application area for the On Chain Metrics in professional and analytical contexts where accurate on chain metrics calculations directly support informed decision-making, strategic planning, and performance optimization
Crypto hedge funds generating alpha from on-chain signal-based strategies, representing an important application area for the On Chain Metrics in professional and analytical contexts where accurate on chain metrics calculations directly support informed decision-making, strategic planning, and performance optimization
Academic researchers and university faculty use the On Chain Metrics for empirical studies, thesis research, and peer-reviewed publications requiring rigorous quantitative on chain metrics analysis across controlled experimental conditions and comparative studies
DeFi protocol governance analyzing user behavior for parameter adjustments, representing an important application area for the On Chain Metrics in professional and analytical contexts where accurate on chain metrics calculations directly support informed decision-making, strategic planning, and performance optimization
Macro analysts assessing Bitcoin adoption and monetary network growth, representing an important application area for the On Chain Metrics in professional and analytical contexts where accurate on chain metrics calculations directly support informed decision-making, strategic planning, and performance optimization
{'case': 'Bitcoin ETF launch impact on on-chain metrics (2024)', 'description': "The January 2024 launch of spot Bitcoin ETFs in the US created significant on-chain metric distortions: Bitcoin held by ETF custodians (primarily Coinbase Custody) appears as exchange balance, inflating 'exchange reserves' without representing actual sell pressure. Analysts had to adjust on-chain reserve metrics to exclude known ETF custody addresses to avoid false signals."}
{'case': 'LUNA/UST collapse and SOPR signal', 'description': 'The May 2022 Terra/LUNA collapse caused Bitcoin to fall from $40,000 to $26,000 in days. SOPR fell sharply below 1 as forced sellers liquidated at losses. However, unlike genuine cycle bottoms, the SOPR recovery was brief before a second leg down — illustrating that contagion events can create false capitulation signals that are resolved only after the full damage is assessed.'}
{'case': 'Wash trading and NVT distortion', 'description': "Exchanges competing for trading volume rankings often engaged in wash trading (self-matching trades to inflate reported volumes) in 2018-2020. This artificial transaction volume suppressed NVT ratios, making coins appear cheaper than fundamentals warranted. Better blockchain analytics and CoinMetrics' adjusted transaction volume metrics attempt to filter wash trading from NVT calculations."}
| Metric | Buy Signal Zone | Neutral Zone | Sell Signal Zone | Current (April 2024) |
|---|---|---|---|---|
| MVRV Ratio | < 1.0 | 1.0 - 3.0 | > 3.5 | ~2.4 (mid-bull) |
| NVT Signal | < 50 | 50 - 100 | > 150 | ~90-110 |
| SOPR (7D MA) | Sustained < 1.0 | ~1.0 | > 1.02 at top | ~1.01 |
| Puell Multiple | < 0.5 | 0.5 - 2.0 | > 4.0 | ~0.6-0.8 (post-halving) |
| Reserve Risk | < 0.002 | 0.002 - 0.024 | > 0.024 | ~0.006-0.010 |
| LTH Supply % | > 75% (accumulation) | 60-75% | < 50% (distribution) | ~72% |
What is the realized cap and why is it useful?
The realized cap (developed by Nic Carter and Antoine Le Calvez of Coin Metrics) values each Bitcoin UTXO at the price it last moved on the blockchain, rather than the current market price. This creates a measure of the aggregate cost basis of all Bitcoin holders — the total amount of money invested in Bitcoin at various historical prices. Unlike market cap (which values all coins at the current speculative price), realized cap is anchored to actual transaction prices. It tends to be much more stable and serves as a kind of 'fair value' anchor that market cap oscillates around through bull and bear cycles.
How reliable is the Stock-to-Flow (S2F) model?
The Stock-to-Flow model, developed by the pseudonymous analyst Plan B, predicted Bitcoin's price trajectory remarkably accurately from 2011 through 2021, with R-squared above 0.90 against log price. However, the model significantly overestimated Bitcoin's price in 2022 and 2023, when it predicted $100,000+ while Bitcoin fell to $16,000. Critics argue S2F fails to incorporate demand, regulatory, and macroeconomic factors that dominate shorter-term price action. The model is best understood as a long-term scarcity framework rather than a precise price predictor, and its empirical fit may reflect luck in the historical data rather than true predictive power.
What is the difference between supply held by long-term and short-term holders?
On-chain analysis typically classifies Bitcoin holders as long-term holders (LTH — coins unmoved for 155+ days) and short-term holders (STH — coins moved within the last 155 days). LTH supply tends to grow during bear markets (patience of veteran holders) and decline near tops (as veterans distribute to new entrants). STH supply grows near tops (as speculators accumulate) and at bottoms (as distressed sellers cycle coins to new hands). The LTH/STH supply dynamic is one of the most reliable on-chain macro cycle indicators, published by Glassnode.
What is the Puell Multiple and how is it used?
The Puell Multiple measures the ratio of daily miner revenue (in USD) to the 365-day moving average of daily miner revenue. It captures miner profitability relative to the long-run average, identifying periods of extreme miner stress (low Puell = bear market, miners potentially forced to sell coins) and extreme miner profitability (high Puell = bull market peak conditions). Values below 0.5 have historically coincided with excellent buying opportunities; values above 4 have coincided with market tops. The April 2024 halving immediately halved the Puell Multiple, putting it in the historically attractive zone.
How do Layer 2 networks affect on-chain metric interpretation?
Layer 2 networks (Bitcoin Lightning Network, Ethereum's Arbitrum and Optimism, Polygon) process transactions off the main chain, with only periodic settlement on Layer 1. This means on-chain transaction volume metrics for Bitcoin and Ethereum miss an increasing share of real economic activity. A rising NVT ratio might reflect actual overvaluation OR simply a shift of usage to Layer 2 rather than Layer 1. On-chain analysts must increasingly account for Layer 2 volume and adjust their metrics accordingly. Ethereum analytics tools like Dune Analytics increasingly track L2 activity separately.
What is reserve risk and how is it calculated?
Reserve Risk, developed by Hans Hauge, quantifies the risk-reward of investing in Bitcoin relative to the confidence of long-term HODLers. It measures the ratio of current price to HODL Bank (the cumulative number of days coins have been held without selling — a measure of HODLer conviction). Low Reserve Risk (below 0.002) indicates historically attractive buy zones — HODLers are demonstrating high conviction while price is low. High Reserve Risk (above 0.024) indicates historically unfavorable risk-reward — price is elevated relative to underlying HODLer conviction. The metric is calculated and tracked by Glassnode.
What is exchange reserve data and what does declining exchange balances signal?
Tracking the volume of Bitcoin and Ethereum held on centralized exchange wallets is a key on-chain leading indicator. Declining exchange reserves (coins flowing off exchanges to self-custody wallets) suggest holders are withdrawing to hold long-term, reducing immediately available sell pressure — generally bullish. Increasing exchange reserves signal holders preparing to sell, increasing sell pressure — generally bearish. The net exchange flow (deposits minus withdrawals) is tracked by CryptoQuant and Glassnode and correlates inversely with price in many historical periods, particularly at market turning points.
Pro Tip
For market cycle positioning, use the MVRV Z-Score (normalized MVRV) rather than raw MVRV, as it accounts for long-run trend. MVRV Z-Score below 0 has historically been in the green zone (buy), above 7 in the red zone (sell). Glassnode provides this as a free metric, making it one of the most accessible institutional-grade on-chain indicators.
Did you know?
Glassnode, one of the leading on-chain analytics platforms, identified that approximately 3-4 million Bitcoin (worth $200-270 billion at 2024 prices) are permanently lost — sent to inaccessible wallets or held by owners who have lost their private keys. These lost coins actually make Bitcoin scarcer than the nominal supply suggests, with the effective circulating supply being approximately 19.7M minus 3-4M lost coins.