On-chain analysis 101: reading the blockchain like a trading desk

Imagine if every bank transfer in the world were posted on a public noticeboard. You couldn't always see whose names were on the accounts, but you could see every amount, every timestamp and every account balance, live. That's basically what a public blockchain is.

Traditional finance runs on information that companies choose to share: quarterly reports, investor calls, the occasional leak. Public blockchains flip that. Essentially every transaction is visible in real time to anyone who wants to look, and it stays there permanently. On-chain analysis is the skill of turning that raw firehose of data into an actual read on what's going on in the market.

The good news: you don't need to code to get started. The bad news: it's very easy to misread, and plenty of confident social media threads do exactly that. This guide covers both sides.

The building blocks

Before any metric, you need to know what's being measured.

Addresses

The public identifiers that send and receive assets, like a long string starting with "0x" on Ethereum or "bc1" on Bitcoin. One person can control thousands of addresses, and one address (say, an exchange's hot wallet) can hold funds for millions of users.

Transactions

Individual transfers or interactions with smart contracts. Each one is timestamped, includes the fee paid, and can be traced forever.

Blocks

Batches of transactions. Each block references the one before it, which is what lets analysts rebuild exact history and timing.

Smart contract events

Logs created when someone interacts with a contract, like swapping tokens, depositing into a lending pool, or getting liquidated. This is how analysts track activity inside DeFi, not just simple transfers.

Labels

Analytics firms tag known addresses, such as exchange wallets, project treasuries or famous funds. Labels are what turn "address X sent coins to address Y" into "a fund just moved coins to an exchange". They're educated guesses, not facts, and they're sometimes wrong.

How to read a single transaction

The best way to start is to look up one transaction on a block explorer (a free website that shows blockchain data, covered in our toolkit). Here's what you'll usually see and what it means:

Once you can read one transaction, you can follow money: click the receiving address, see where funds went next, and start to understand how assets move between people, exchanges and protocols.

The metrics analysts actually watch

Exchange flows

This tracks coins moving onto and off known exchange addresses. Big inflows to exchanges are often read as a possible sign of selling, because you usually need coins on an exchange to sell them. Big outflows are often read the other way, as accumulation, because people moving coins into their own wallets usually aren't planning to sell immediately.

Neither reading is certain. Coins might move to an exchange to be used as collateral, or move off because an exchange is reorganizing its wallets. Treat flows as a probability signal, not a statement of intent.

Exchange reserves and stablecoin balances

Instead of daily flows, some analysts look at the total held on exchanges over time. A long, steady decline in coins sitting on exchanges suggests holders are moving to self-custody. Rising stablecoin balances on exchanges are sometimes called "dry powder": money parked and potentially ready to buy.

Holder cohorts: long-term vs short-term

By grouping addresses by how long coins have sat without moving, analysts can separate long-term holders from short-term traders. A common rule of thumb treats coins unmoved for about 155 days or more as "long-term holder" supply. When long-term holders are steadily adding while short-term holders are panicking, that divergence is closely watched. When long-term holders start selling heavily into a rally, it can be a sign of distribution (see our market cycles guide).

Realized value and MVRV

Normal market cap values every coin at today's price. Realized cap values each coin at the price it last moved on-chain, then adds them up. That roughly approximates what the whole network paid for its coins, an aggregate cost basis.

MVRV is simply market value divided by realized value. When it's high, the average holder is sitting on big unrealized profits, which historically has coincided with overheated markets. When it's below 1, the average coin is worth less than it was bought for, which has historically shown up in deep bear markets. It's a sentiment thermometer, not a timer.

SOPR: are people selling at a profit?

The spent output profit ratio looks at coins that moved on a given day and compares the price when they moved to the price when they were acquired. Above 1 means, on average, coins are being sold at a profit; below 1 means at a loss. In bull markets, dips toward 1 sometimes act as a floor, because people don't like selling at a loss. In bear markets, rallies toward 1 often stall, because people who were underwater sell as soon as they break even.

Network activity

Active addresses, transaction counts and fees paid are basic but useful signs of real usage, as opposed to pure price speculation. A rising price with falling activity is a divergence worth noting; the two don't usually stay apart for long. Be careful, though: activity can be inflated by bots, spam or airdrop farmers making lots of tiny transactions.

DeFi and protocol health

Inside DeFi, on-chain data shows total value locked in a protocol, how much is being borrowed, and where liquidation levels sit. That lets you check a protocol's collateral directly instead of trusting a dashboard the protocol itself runs. If you're thinking about depositing anywhere, this is worth learning. Our DeFi yield guide explains why.

Whale watching

Plenty of accounts on social media post alerts whenever a huge amount moves. These can be interesting, but they're often noise: a big transfer may just be an exchange shuffling coins between its own wallets. Before reacting to a whale alert, check whether both addresses belong to the same entity.

On-chain vs chart analysis

Technical analysis studies price and volume patterns on a chart, mostly without caring what causes them. On-chain analysis studies the activity underneath: who's moving coins, how concentrated ownership is, and what's happening inside protocols. They work best together.

For example, a price breakout that comes with growing active addresses and coins flowing off exchanges tells a different story from the exact same breakout with falling activity and coins flowing onto exchanges. The chart looks identical; the context doesn't.

Also note that some popular indicators, like funding rates and open interest, come from derivatives exchanges rather than the blockchain itself. They're useful, but they're "off-chain" data and depend on what exchanges report.

The traps that catch beginners

Address count isn't user count

One person can run thousands of addresses; one exchange address can represent millions of people. Metrics built on raw address counts need careful reading.

Correlation isn't causation

Exchange inflows rising while price falls doesn't prove the inflows caused the drop. Both might be reactions to the same news.

A lot of activity is invisible

Trades inside centralized exchanges happen on their internal books and never touch the public chain. On-chain data sees a real but partial picture.

Labels can be wrong

Address labels are educated guesses. A wallet tagged as "a fund" might be something else entirely.

Smart players hide their tracks

Large holders often spread funds across many addresses and intermediary wallets precisely to make their activity harder to read.

Cherry-picking

With hundreds of metrics available, it's easy to find one that confirms whatever you already believe. Decide what you'll look at before you look.

A simple routine to get started

1. Pick one network

   Bitcoin or Ethereum are easiest because they have the most free tools and educational material.

2. Learn to read transactions

   Look up a few on a block explorer until the fields feel familiar.

3. Track three metrics, not thirty

   For example: exchange flows, long-term holder behavior and MVRV. Check them weekly and write down what you see.

4. Compare against price and news

   Over a few months you'll start to see how the data behaves in different conditions, including when it's wrong.

5. Stay humble

   Use on-chain data to ask better questions, not to feel certain. No metric has a perfect record.

This article is general education about how things work. It is not financial, investment or tax advice, and it doesn't consider your circumstances. Do your own research and talk to a licensed professional before making decisions with money.