Golden Bitcoin coin standing among layers of glowing orange, turquoise, and blue waves, with Japanese candlesticks in the background
Original illustration generated with AI (Adobe Firefly) -- does not represent real data.

Bitcoin HODL Waves: What They Are and How to Read Them

HODL waves split all circulating Bitcoin by how long each coin has gone without moving. They show how much of the market is in the hands of long-term holders versus recent hands -- an angle of the cycle that price alone doesn't show.

HODL waves are a breakdown of all circulating Bitcoin by how long each coin has gone without moving addresses. It's an X-ray of market conviction: if a large share of supply has sat still for years, there are many long-term holders; if a lot moves constantly, speculative activity dominates. It's one of the hardest data points to get in all of on-chain analysis, and one of the ones we calculate directly from our own Bitcoin node.

What HODL Waves Are

This article complements our Score methodology, where this data point (combined with holder distribution by wallet size) forms part of the own-node component. You can also check out the Puell Multiple, another indicator that uses mining data.

The starting idea is simple: every Bitcoin coin has an "age," the time elapsed since it last moved on-chain. If someone buys Bitcoin and doesn't touch it for years, that's a signal of long-term conviction ("hodling," from a famous typo of "holding"); if coins change hands frequently, it's more a sign of trading or short-term speculation.

A useful analogy: imagine you could carbon-date every Bitcoin coin, the way a fossil is dated, and then stack them by age into geological strata. The deepest layers -- coins that haven't moved in years -- are the market's solid base: holders who've weathered entire cycles. The upper layers -- coins that moved this week -- are recent activity, more volatile. HODL waves are exactly that layered view, expressed as what percentage of total supply each age stratum occupies.

How to Read Them Across the Cycle

The value of HODL waves isn't in a single number, but in how the layers change over time:

  • Older bands fatten during bear markets. When price falls and general interest disappears, those who remain are conviction holders, still not moving their coins. "Old" supply grows as a share of the total. It's the quiet pattern of background accumulation.
  • Older bands thin out near tops. In full euphoria, coins that had sat still for years start moving: veteran holders take advantage of high prices to sell part of their stack. When the oldest supply starts to "wake up," it has historically been a signal of a late-stage cycle.
  • A spike in young bands signals fresh speculative demand. If a lot of supply suddenly moves into the "less than a month" band, fresh (or rotating) capital is coming in -- typical of heating-up phases.

None of these readings mark a specific day: they're slow, background shifts that locate the cycle phase, not time the turn. That's why HODL waves are used as context, not as a buy/sell trigger.

The Age Bands We Use

Our calculation groups every UTXO (unspent transaction output) by the blocks elapsed since it was created, into these bands:

  • Less than 1 day
  • 1 day - 1 week
  • 1 week - 1 month
  • 1 - 3 months
  • 3 - 6 months
  • 6 months - 1 year
  • 1 - 2 years
  • 2 - 3 years
  • 3 - 5 years
  • More than 5 years

Why It Requires Our Own Node

This is the hardest data point to get in the whole Score, and the reason is concrete: no free API (Coin Metrics, CoinGecko, Blockchain.com) offers HODL waves or a supply breakdown by age -- that data is reserved for paid plans at the big providers (Glassnode, CryptoQuant), or requires building it from scratch by indexing the age of every UTXO, which requires direct access to the blockchain.

Our process, in short: the Bitcoin node generates a full snapshot of the UTXO set (dumptxoutset, a binary dump of every unspent output at a given block), and our own parser walks through that file -- tens of millions of entries -- calculating the age of each one against the snapshot's height. The result is aggregated into the bands above. It's a heavy process (several minutes over a file several gigabytes in size) that runs automated once a week, not an on-demand calculation on every visit to the site.

A Related Metric: Holder Distribution by Wallet Size

From the same UTXO set snapshot we also calculate a different cut: instead of grouping by age, we group by how much BTC each address controls -- from "shrimp" (less than 0.01 BTC) to "whales" (more than 10,000 BTC). They're two different questions about the same data: "how long since this coin last moved?" versus "how much BTC does this address control?" Both feed the own-node component of the Score, and are shown in more detail with comparative history in our Supply and holders section.

Limits and Common Mistakes

HODL waves are powerful, but they have real limits worth knowing before drawing conclusions:

  • Lost coins look like eternal conviction. Millions of BTC with permanently lost keys never move, so they permanently fatten the oldest bands. Part of the "5+ years" supply isn't a patient holder: it's inaccessible Bitcoin that will never return to the market.
  • Institutional custody looks like an individual holder. Bitcoin held in an exchange's or fund's cold storage shows up as still coins, even though thousands of users may be buying and selling behind it without those coins moving on-chain. The "conviction" picture can be inflated by custody, not by individual holding.
  • A movement isn't always a sale. When an old coin moves, its age resets, but that can be a simple wallet change by the same owner, not a real sale. It's an estimate with a known bias, not an exact measure of intent.
  • It's a periodic snapshot, not real time. Since it's calculated at intervals over a full snapshot, it doesn't reflect what happened yesterday, but the photo of the latest dump. And while our node is still finishing its sync, the snapshots are partial: valid for illustrating the mechanism, not as a definitive state of the network.

The most common mistake is reading a single snapshot as a verdict. Its value is in the trend between snapshots, not in a single number -- and always combined with other indicators, never in isolation.

In Summary

HODL waves split all Bitcoin by holding age and show how much of the market is in firm hands versus recent hands: an angle of the cycle price doesn't reveal. Its strength is in the trend -- older bands fattening at bottoms, thinning near tops -- and its limit is that lost coins and custody can distort the picture. You can see it calculated from our own node, with comparative history, in the Supply and holders section.

Last updated: 2026-07-09