How to Read On-Chain Data

On-chain data is the public information recorded on a blockchain: transactions, addresses and balances. On open networks such as Bitcoin and Ethereum, anyone can see the amount of every transfer, the sending and receiving addresses, the fee paid and when it happened. On-chain analysis aggregates these records to measure how much a network is used, whether coins are flowing to exchanges or to private wallets, and how long-term holders are behaving.

The strength of this data is its transparency. Its limit is that you usually do not know who owns an address or why a transfer was made. Good on-chain analysis keeps what is measured separate from what is inferred.

What you can and cannot see on-chain

Visible: the asset and amount sent, addresses, fees, block time, smart contract calls and token transfers. Our guide on tracking a blockchain transaction shows how to look up a single transfer.

Not visible: the identity behind an address, the purpose of a transfer and the internal ledgers of centralised exchanges. Most trades on an exchange are settled in its own database and never touch the blockchain, so on-chain transaction counts cannot be compared directly with exchange volume.

The most widely used on-chain metrics

• Active addresses: the number of addresses transacting in a period, a rough gauge of network usage.
• Exchange inflows and outflows: coins moving into and out of addresses known to belong to exchanges.
• Whale movements: transfers by addresses with very large balances; see our entry on crypto whales.
• Fees and hash rate: indicators of demand for block space and, on Bitcoin, of mining power.
• Stablecoin supply and exchange balances: clues to how much dollar liquidity is sitting ready in the crypto market.
• MVRV and SOPR: whether coins are in profit or loss relative to the price at which they last moved (explained below).

Why address counts are not user counts

One person can control dozens of addresses, while a single exchange address may represent thousands of customers. Bitcoin wallets also generate a fresh change address with many transactions, which inflates address counts naturally. Providers such as Glassnode try to cluster related addresses into entities belonging to the same person or institution. These entity-adjusted metrics are more meaningful, but they are still estimates based on heuristics.

How to read exchange inflows and outflows

The common reading is that large inflows to exchanges may signal intent to sell, while outflows suggest coins are being moved to private wallets for longer-term storage. But a deposit does not prove a sale. The user may be posting collateral, funding a derivatives account, or the exchange may simply be shuffling coins between its own hot and cold wallets.

Net flow is inflows minus outflows. If 5,000 BTC flow in and 4,200 BTC flow out on the same day, net inflow is 800 BTC. The net figure can hide the scale of activity: 100 BTC in and 100 BTC out produces the same zero as 100,000 in and 100,000 out. Look at gross flows and at which exchanges are covered too.

What are realized cap, MVRV and SOPR?

Realized cap values each coin at the price when it last moved on-chain rather than today’s price, giving a rough estimate of the market’s aggregate cost basis. MVRV is market cap divided by realized cap; well above 1 suggests holders are on average in profit, below 1 that they are on average at a loss. SOPR (spent output profit ratio) compares the value of coins when spent with their value when they last moved; above 1 means coins moving that day changed hands at a profit on average.

These metrics have a blind spot: the “last moved” price is not necessarily anyone’s real purchase price. Someone who bought BTC on an exchange at $40,000 and withdrew it when the price was $45,000 appears on-chain with a $45,000 cost basis. Transfers between a holder’s own addresses also distort the picture.

On-chain data on Ethereum and in DeFi

On smart contract networks, on-chain data goes well beyond transfers. DEX trading volume, collateral and debt on lending protocols, gas usage and the number of newly deployed contracts can all be read from the chain. Checking a token’s largest holders, movements from team wallets or changes in a liquidity pool in a block explorer gives you hard evidence to set beside a project’s announcements. Here too, it pays to verify that address labels are accurate.

What dormant coins can tell you

Analysts group coins by how long they have sat unmoved, and call holders of long-dormant coins “long-term holders”. Old coins moving is sometimes read as profit-taking. Yet a dormant coin may simply have lost keys, and a move may just be a change of custody. Bitcoin’s UTXO model is what makes these age calculations possible.

Where to find on-chain data

For individual transactions, use block explorers for Bitcoin and Ethereum, such as Etherscan. For ready-made metrics and charts, analytics platforms such as Glassnode and CryptoQuant are widely used, and tools like Dune let you write your own queries. For assets locked in DeFi protocols, sites that publish TVL data are the usual reference. Platforms label addresses and define metrics differently, so the same metric can look different from one site to another.

Four steps for reading an on-chain chart

1. Read the definition: is the unit coins or dollars, a daily total or a moving average?
2. Check the coverage: which exchanges and address labels are included?
3. Explain sudden changes: genuine usage, price action, a new address label or a data revision?
4. Separate data from interpretation: “net exchange inflows rose” is data; “investors are about to sell” is an inference.

Avoid ranking networks by raw transaction counts as well. A Bitcoin transfer, an Ethereum contract call and a layer-2 transaction do not carry the same economic weight. Tracking changes within one network’s own time series is far more informative.

Sources

• Glassnode: address clusters and adjusted metrics
• Glassnode: holding time classification
• Bitcoin developer documentation: transactions and outputs