Reading On-Chain Data
Every blockchain transaction is public and permanent, which turns wallet balances, transfer flows and contract activity into a research dataset no traditional market offers at anywhere near the same granularity.
Prerequisites: Blockchains and Consensus, Wallets, Keys and Custody
A stock exchange tells you the price and size of every trade, but not whose account traded, and never what any account's total holdings are. A public blockchain tells you all of that, permanently, for free — every transaction, every wallet balance, every smart contract interaction is sitting in a database anyone can query. That transparency is the raw material of on-chain analysis: reading market structure and behavior directly from the ledger rather than from prices alone.
On-chain data lets you observe the actual movement of coins between wallets and contracts, not just the resulting price — which turns questions like "are large holders accumulating or distributing," "is an exchange solvent," or "is a token's liquidity thin" into things you can measure directly instead of infer indirectly.
What's actually visible
An address's full transaction history and current balance are public by default (pseudonymous, not anonymous — the address itself isn't tied to a name unless linked elsewhere). From that raw data, analysts build derived metrics:
- Exchange flows. Coins moving into exchange wallets often precede selling; coins moving out often signal holders taking custody for the longer term.
- Whale wallets. Addresses holding unusually large balances can be tracked individually — a single wallet moving a large position is visible in real time, unlike an institutional trade in equities that settles opaquely.
- Smart contract activity. Interactions with a DeFi protocol — deposits, withdrawals, liquidations — are all individually visible transactions, not aggregated statistics.
Worked example
An analyst tracks a stablecoin's reserve wallet and notices $50 million moved out to an exchange deposit address over 24 hours, against $800 million total reserves.
- Share moved. $50m against $800m total reserves: , i.e. 6.25% of reserves shifted toward an exchange in a single day.
- Interpretation. This could be routine treasury management, or it could be an early signal of redemption pressure or reserve reallocation worth investigating before it shows up as a price move — the on-chain move is visible hours or days before any official disclosure would confirm the reason.
- Caveat. The data shows that funds moved and where, but not why — on-chain data answers "what happened" far better than "what it means," which still requires judgment.
What this means in practice
On-chain analysis firms (Chainalysis, Nansen, Glassnode) build businesses entirely around labeling addresses (this one belongs to an exchange, this one to a known fund) and packaging flow data into tradable signals — accumulation/distribution trends, exchange reserve levels, stablecoin mint/burn activity. It's also the primary tool for tracing illicit activity, since the same permanence that helps a researcher also means a hacker's stolen funds can often be followed step by step, even if the identity behind the wallet stays hidden.
Pseudonymous is not anonymous, but it's also not identified — don't over-read "we can see wallet X did Y" as "we know who did Y." Address clustering and labeling are probabilistic inference built on top of the raw chain data, not facts read directly off it, and mislabeled addresses are a real, ongoing source of bad on-chain research.
Related concepts
Practice in interviews
Further reading
- Chainalysis, The Chain Report (annual)
- Antonopoulos, Mastering Bitcoin (ch. on the blockchain)