On-chain analytics reads data directly from a blockchain’s transaction history to track capital movements, wallet behavior, miner and validator activity, and network usage patterns. Unlike price charts, which show what happened to price, on-chain data shows what participants did with their assets: whether large holders added positions, moved coins to exchanges, or held through a correction. It is a separate signal layer, not a replacement for technical or fundamental analysis.
How on-chain data is tracked
Every transaction on a public blockchain is visible and permanent. Analytics platforms like Glassnode, CryptoQuant, and Santiment index this data and calculate derived metrics: net exchange flows (are more coins entering or leaving exchanges?), realized profit and loss (what is the average cost basis of coins that moved today?), dormancy (how long have the coins that moved today been held?), and validator behavior (are validators accumulating or selling?)
Bitcoin’s UTXO model makes cost-basis tracking precise. Each UTXO records its value at the last time it moved, which is how metrics like MVRV, SOPR (Spent Output Profit Ratio), and Realized Cap are calculated. Ethereum’s account-based model makes equivalent calculations less precise, though the coverage has improved substantially through 2025.
What this means for traders
The metrics with the most consistent historical track records are exchange netflows (sustained net outflows, coins leaving exchanges to private wallets, have preceded bull markets; sustained inflows have preceded sell pressure), MVRV ratio (see the MVRV entry for detail), and long-term holder supply (when long-term holders start distributing, it has historically been a cycle top signal).
SOPR above 1 means sellers are in profit on average; below 1 means sellers are realizing losses. A sustained SOPR below 1 has historically coincided with capitulation bottoms. For how on-chain data fits within broader market analysis, see sentiment analysis tools.
A concrete example
In Q3 2022, Glassnode data showed that short-term Bitcoin holders (coins moved within the last 155 days) had realized losses for 67 consecutive days, the longest streak since the 2018 to 2019 bear market bottom. Long-term holders (coins unmoved for over 155 days) barely reduced their holdings during the same period. The combination of short-term capitulation and long-term holder conviction was a historically high-quality entry setup. BTC was at $18,000 to $22,000 during this period and began recovering in Q4 2022. On-chain data did not call the exact bottom, but it confirmed a favorable risk-reward zone.
Frequently asked questions
Is on-chain data available for all blockchains?
Bitcoin has the richest on-chain analytics ecosystem because its UTXO model makes cost-basis tracking precise. Ethereum and EVM chains have broad coverage on Glassnode, CryptoQuant, and Dune Analytics. Solana and other non-EVM chains have growing coverage on Solscan and Dune. The depth of analysis varies significantly across chains.
How do you access on-chain data for free?
Glassnode’s free tier shows lagged data (24 hours behind) for most Bitcoin metrics. Lookintobitcoin.com shows MVRV, SOPR, and several other metrics for free with full history. Dune Analytics allows free SQL queries against raw Ethereum chain data. CryptoQuant offers some real-time metrics free. None of the free tiers cover real-time professional-grade data across all metrics.
Are on-chain signals predictive or descriptive?
Both. Exchange netflows are descriptive (coins moved to exchange) and inferentially predictive (moving coins to exchange often precedes selling). The predictive value comes from historical patterns, not certainty. No on-chain metric has a consistent 100% hit rate; they shift probability assessments rather than determine outcomes.
