On-chain analysis lets you look where the stock market never lets you: at actual transfers of coins between wallets, exchanges and large holders. In this lesson we'll go over which metrics really tell you something about the market, where to find them and why on-chain data is so easy to misread.
- The blockchain shows how coins move, but not who owns them: analysts label addresses, and they can be wrong.
- The key metrics are exchange inflows and outflows, address activity, long-term holder behavior, MVRV and SOPR.
- Since the launch of exchange-traded funds and the growth of custodial services, many classic metrics have to be read with adjustments.
- On-chain data describes the market backdrop over weeks and months; it's better to look for your entry on the chart.
What you can and can't see on the blockchain
A public blockchain is essentially the network's complete ledger. Every transaction is out in the open: the amount, the time, the sender and recipient addresses, plus how many coins sit at each address. An exchange only shows you its order book and trades, while the blockchain reveals what happens to coins before and after they're traded.
The catch is that there are no names on the blockchain. You won't find a "Binance" or "whale" label under an address. Analytics firms tag addresses based on indirect clues: for example, if thousands of users send coins to addresses that are later swept into a single hot wallet, that wallet gets tagged as an exchange. Tagging is useful, but it's an educated guess, not a fact.
Another complication: one person can own hundreds of addresses, while a single exchange address can hold coins for millions of customers. So a metric like "number of addresses holding more than 1,000 BTC" doesn't tell you how many wealthy people there are. It tells you how coins are spread across wallets, and many of those belong to exchanges and custodians.
On-chain is everything recorded on the blockchain. Off-chain is what happens inside an exchange or service: trades between customers of the same platform never hit the blockchain, so you won't see them in on-chain data.
Key metrics and how to read them
There are hundreds of metrics, but a few are enough to get started. They answer three questions: is anyone actually using the network, which holders are getting ready to sell, and is the market in profit?
Network activity shows up in the number of active addresses and in fees. When they rise, there's demand for blockspace, but spam transactions or token airdrops can inflate the numbers artificially. So compare a network against its own history rather than comparing different networks with each other.
Holder behavior is read through coin age. When coins that have sat untouched for years suddenly move, it often means long-term investors are taking profits or moving their funds around. Metrics like Coin Days Destroyed measure exactly how old the coins that just started moving are.
Stablecoins deserve their own watchlist. The total supply of USDT and USDC and how much of it flows to exchanges show how much "dry powder" buyers are holding. A growing stablecoin supply doesn't guarantee rising prices on its own, but a steady decline often goes hand in hand with periods when money is leaving crypto.
| Metric | What it shows | How it's usually interpreted |
|---|---|---|
| Exchange netflow | The difference between coins deposited to exchanges and coins withdrawn | Inflow means possible selling pressure; outflow means coins are moving into storage |
| Stablecoin reserves on exchanges | How many dollar tokens are ready to buy | Growth means potential demand |
| Active addresses | How many addresses took part in transactions | Growth means a livelier network, but it could be spam |
| MVRV | The ratio of market cap to realized cap | High values mean the market is deep in profit; low values mean many holders are at a loss |
| SOPR | Whether spent coins are sold at a profit or a loss | Above 1 means selling at a profit; below 1 means selling at a loss |
| Hashrate | The computing power of Bitcoin miners | Reflects miners' costs and confidence |
Exchange flows and whales: where people get it wrong most often
The classic logic is simple: coins go onto an exchange to be sold and come off it to be held. Hence the alarm over large bitcoin inflows to exchanges, and the conclusion that a steady outflow means accumulation. With stablecoins it's the other way around: inflows to exchanges more often mean buyers are getting their money ready.
In practice, though, flows are noisy. Exchanges change addresses and shuffle coins from wallet to wallet, and if an analyst doesn't update the tags in time, an internal transfer looks like a massive deposit or withdrawal. On top of that, spot Bitcoin ETFs have been trading in the US since 2024, and huge amounts of coins now sit with custodians instead of in the usual exchange wallets. So falling exchange reserves no longer clearly mean coins are "moving to cold storage."
Same story with whales. Alert services post transfers worth hundreds of millions of dollars, and the crowd instantly hunts for a signal in them. But a large share of these transfers are routine operations by exchanges, funds and custodians. Before you react, check which tagged addresses the transfer went between and whether the pattern repeats.
A message saying "5,000 BTC just moved to an exchange" doesn't mean a crash is around the corner. The coins could be shuffled around inside the exchange, posted as collateral or sit there unsold for weeks.
Realized price, MVRV and SOPR in plain English
Market cap values every coin at the current price. Realized cap values each coin at the price it last moved on the blockchain, which gives you a rough total of what market participants paid for their coins. Divide that by the number of coins and you get the realized price: roughly the average purchase price across the market.
When the market price pulls far ahead of the realized price, MVRV climbs: most holders are sitting on big gains, and the temptation to cash out grows. When price drops to the realized price or below, the market as a whole is underwater, and people are less eager to sell. Historically, these zones have lined up with overheated markets and with bear market bottoms, but the exact thresholds have shifted from cycle to cycle.
SOPR looks at the same thing from another angle: are people selling at a profit or at a loss right now? In an uptrend it often dips back to 1 and bounces, because holders don't want to sell without a profit. A sustained stretch below 1 signals capitulation, when people sell even at a loss.
A market phase when holders sell en masse at a loss because they've given up on a recovery. You can often spot it by a sharp spike in volume combined with SOPR below 1.
Tools: where to find on-chain data
You don't have to pay for any of this. Block explorers show individual transactions and addresses, and aggregators turn that data into charts. Paid subscriptions give you more detailed metrics and address tagging, but for the big picture the free versions are usually enough.
The key is to match the tool to the job: Bitcoin calls for one set of metrics, DeFi and Ethereum tokens for another. There's a detailed breakdown of these services in the lesson on analytics platforms for crypto; below is a quick cheat sheet.
Keep in mind that different services can calculate the same metric differently, since each one tags exchange addresses and filters out internal transfers its own way. So don't compare Glassnode and CryptoQuant numbers head to head; follow the trend within a single source.
- Etherscan, mempool.space and other block explorers: check a specific transaction or address.
- Glassnode and CryptoQuant: Bitcoin and Ether metrics, exchange flows, MVRV, SOPR.
- Dune: open SQL dashboards built by the users themselves, especially useful for DeFi.
- Arkham and Nansen: address tags for funds, exchanges and well-known players.
- DefiLlama: total value locked in protocols, plus network revenue and fees.
How to build on-chain analysis into your trading
On-chain data moves slowly. It describes the state of the market well over weeks and months, but it won't pinpoint when to enter. So use it as a backdrop to tell whether the market is overheated or depressed, and find your entry point on the chart using technical analysis.
Another use is fact-checking news and projects. A team claims "explosive user growth"? Check it against active addresses and fees. Big team wallets started moving tokens to an exchange ahead of an unlock? That's a reason to be careful. This is how on-chain analysis complements fundamental analysis with facts.
To avoid drowning in charts, stick to a simple routine.
- Pick three or four metrics and check them regularly, for example once a week.
- Compare values with the history of the same coin, not with other coins' numbers.
- Look for several signals lining up rather than reacting to just one.
- Write down your conclusions and check a month later whether they held up.
If on-chain metrics point to overheating and the chart shows weakness at resistance, you've got two independent arguments. Different methods agreeing with each other are more reliable than any single metric.
Take one recent alert about a large BTC or USDT transfer (for example, from Whale Alert), open that transaction in a block explorer and look up the addresses in Arkham or another tagging service. Figure out whose addresses they are (an exchange, a fund, a custodian or an unknown wallet) and write down whether it looks like coins being sent to an exchange to sell or a routine internal transfer. This live example will show you why a single alert isn't a trade signal yet.
Entry is free and you trade on a demo account, so you need no money of your own. Free tournaments pay out Tour Passes, and those get you into tournaments with USDT prizes. The best go further — to a $5,000 funded account.
FAQ
Does on-chain analysis work for altcoins?
Generally yes, with a few caveats. Small coins have little data, their addresses are poorly tagged, and much of the trading happens inside exchanges. For tokens on Ethereum and other networks, it's more useful to look at how supply is spread among holders, upcoming unlocks and activity in protocols.
Can I do on-chain analysis for free?
Yes. For the basics, a beginner can get by with block explorers, DefiLlama, public Dune dashboards and the free tiers of analytics services. A paid subscription makes sense once you know exactly which metrics you need.
How is on-chain analysis different from technical analysis?
Technical analysis studies the price and volume of trades on exchanges; on-chain analysis studies how coins move on the blockchain. Technical analysis is better for finding an entry point, while on-chain data helps you see what large holders are doing and how overheated the market is.
Should I follow whale transfers in Telegram channels?
As a source of ideas, sure, but not as a trade signal. Most large transfers are routine operations by exchanges and custodians. Before you draw any conclusions, check the addresses in a block explorer or a tagging service.
Discussion
Comments from BuyCrypt Live. Sign in with Google or Apple — we create your BuyCrypt account on the way, the same one as in the app.
