Public blockchains record every transaction in a ledger that anyone can read. On-chain analysis is the practice of studying that data to learn how a network is being used and what its holders are doing. It adds context that a price chart alone cannot give. It also has clear limits, and treating any metric as a trading signal is a common mistake.
What On-Chain Data Contains
A blockchain stores transfers, balances and smart contract interactions. The basic tool for reading it is a block explorer, a website that lets you look up any transaction, address or block. The ethereum.org page on block explorers lists what they show.
The two largest networks record data differently.
- Bitcoin tracks unspent transaction outputs, known as UTXOs. Each coin can be traced back through earlier transactions, and the age of every output is known. The Bitcoin developer guide on transactions explains the model.
- Ethereum tracks account balances and contract state. Analysis there often focuses on token transfers, DeFi protocols and gas use.
Raw data has no names in it. A blockchain shows addresses, not people. Analytics firms attach labels such as exchange, miner or fund by clustering addresses with heuristics. Those labels are estimates, and they can be wrong or incomplete.
Exchange Flows
One widely watched metric is the flow of coins into and out of exchange wallets.
- Inflows: coins moving to exchanges can mean holders are preparing to sell, since selling usually happens there.
- Outflows: coins leaving exchanges for private wallets are often read as a sign of long-term holding.
- Stablecoin inflows: stablecoins moving onto exchanges are sometimes read as buying power arriving.
These are tendencies, not rules. Exchanges move funds between their own wallets. Custodians and ETF issuers shift coins for operational reasons. A deposit can be collateral for a derivatives position, not a sale. A single large transfer tells you very little without knowing who made it and why.
Large Holders
Addresses with large balances are often called whales. Analysts watch for long-dormant wallets becoming active, steady accumulation over time, and transfers to exchange deposit addresses.
There are caveats. Many of the largest addresses belong to exchanges and custodians holding coins for thousands of customers. One entity can control many addresses, and many people can share one. The direction of the price after a large transfer is not predictable from the transfer itself.
Network Activity
- Active addresses: the number of unique addresses that sent or received in a period. It is a rough measure of usage, but one person can create many addresses, and cheap chains attract bot activity.
- Transaction count: easy to inflate on networks with very low fees.
- Transfer volume: the value moved on-chain. It includes change outputs and internal transfers, so providers apply adjustments that differ from each other.
- Fees paid: what users actually spend to use the network. This is harder to fake than transaction counts, because it costs real money.
Supply and Holder Behavior
- Coin days destroyed: each coin accumulates one coin day for every day it sits unmoved. When it moves, those days are destroyed. A spike means old coins are moving, which may indicate long-term holders are selling or simply reorganizing wallets.
- Realized capitalization: values each coin at the price when it last moved, not at today's price. It approximates the total cost basis of all holders.
- MVRV: market value divided by realized value. A high ratio means holders are, on average, sitting on large unrealized profits. A ratio below one means the average holder is at a loss.
These metrics work best on Bitcoin, where the UTXO model makes coin age easy to measure. They describe past behavior. They do not forecast.
DeFi Metrics
On smart contract chains, analysts also look at:
- Total value locked: the value of assets deposited in a protocol. It rises and falls with token prices, so a change in TVL is not always a change in usage.
- DEX volume and pool liquidity: how much is traded, and how deep the pools are.
Limits of On-Chain Analysis
- Most trading happens off-chain. Trades inside a centralized exchange are entries in the exchange's own database and never appear on the blockchain.
- Derivatives, which make up a large share of crypto trading volume, are mostly invisible on-chain.
- ETF and custodian holdings sit in large pooled wallets that say little about the investors behind them.
- Activity on Layer 2 networks and across bridges is fragmented and easy to double count.
- Labels are heuristics. Different providers publish different numbers for the same metric.
- Patterns that held in past cycles rest on a small number of examples and may not repeat.
Common Mistakes
- Reacting to one large transfer reported on social media.
- Treating exchange inflows as a guaranteed sell signal.
- Comparing metrics across providers without checking their definitions.
- Ignoring off-chain factors such as interest rates, regulation and derivatives positioning.
How to Use It Sensibly
Use on-chain data to ask better questions, not to produce buy or sell signals. Check whether several independent measures agree. Combine them with price, liquidity and the wider economic picture. No single metric reliably predicts prices.
Nothing in this article is financial advice.
Key Points
- On-chain data shows addresses and transfers, not identities or intent.
- Exchange flows and holder metrics are context, not forecasts.
- Fees and liquidity are harder to fake than transaction counts and volume.
- Most trading volume happens off-chain and cannot be seen.
Track on CoinMarketGuy
CoinMarketGuy focuses on market data, which is useful alongside on-chain research. The DEX page shows active liquidity pools with their volume and liquidity. The Bitcoin price page shows the live price, market cap, supply and chart, and the homepage shows the Fear and Greed reading with total market cap and volume.