What Realized Cap Can and Can’t Tell You
Realized cap is a genuinely clever idea — valuing each coin at the price it last moved, not at current spot. It is also built on assumptions that…
Active addresses is one of the most-cited on-chain metrics, and one of the most routinely over-interpreted. Here is what it can support as evidence, and what it cannot.
Not necessarily. Active addresses counts unique addresses transacting per day, but one person can generate many addresses and one exchange can serve millions of customers from a handful — so the metric is a coarse activity trend, not a user count.
Active addresses is probably the single most commonly cited on-chain metric, appearing in headlines, research reports and social media threads as shorthand for “Bitcoin adoption.” It deserves closer scrutiny than it usually gets, because what it counts and what people claim it shows are two different things.
Active addresses counts unique Bitcoin addresses that appeared in a transaction — sending or receiving — within a given period, typically a day. It is directly observable and verifiable: anyone can independently confirm the count from public chain data, which is a genuine strength relative to metrics built on proprietary heuristics.
The gap between address and user runs in both directions. Modern wallet software, following the widely-adopted HD wallet standard (BIP-32/44), generates a new receiving address for every incoming transaction by default — this is good privacy practice, since it prevents an outside observer from trivially linking all your payments together. One person, transacting normally, can generate dozens or hundreds of addresses over time. That inflates the count relative to actual user numbers.
In the other direction, custodial exchanges — where a very large share of retail Bitcoin activity happens — can serve millions of customers while touching the chain only occasionally, through batched, consolidated transactions from a small number of hot-wallet addresses. Internal transfers between customers of the same exchange never touch the chain at all. This deflates the count relative to actual economic activity.
None of this makes the metric worthless. A large, sustained, multi-month rise or fall in active addresses reflects a real shift in something — on-chain settlement demand, wallet software adoption patterns, or genuine new usage — even if it cannot tell you precisely which. It is a legitimate coarse activity trend, and worth watching for that reason.
What it cannot do is serve as a stand-in for “how many people use Bitcoin,” a number nobody can currently measure with precision from public data alone. Any headline translating an active-address chart directly into a user-growth claim is making a jump the data does not support.
On The Bitcoin Signal, we show the current active-address figure against its 30-day average, explicitly labelled with the caveat above. We treat directional, sustained multi-month changes as worth investigating further, and we do not report day-to-day noise in this series as meaningful on its own.
If each user consistently mapped to roughly one active address, a rising address count would be reasonable evidence of more users. But address generation is address-per-transaction in common wallet designs, not address-per-user, which breaks that mapping in one direction.
In the other direction, exchange-internal activity — the majority of retail Bitcoin economic activity by volume — is largely invisible to this metric, since it happens off-chain inside a custodian's database and only touches the chain in occasional, batched, address-light settlement transactions.
Both effects mean the same raw count can rise for reasons that have nothing to do with new users: a single whale consolidating UTXOs across many addresses, a service migrating address schemes, or genuinely new user growth, all look identical in the aggregate series.
This does not mean the metric is meaningless — a large, sustained, multi-month rise or fall is still a real signal of some kind of shift in on-chain behavior worth investigating further.
It does mean the metric cannot, on its own, distinguish "more people are using Bitcoin" from "existing users and services are generating more addresses," and treating a chart of this metric as a proxy for adoption without that caveat is not supported by what the data actually measures.
Realized cap is a genuinely clever idea — valuing each coin at the price it last moved, not at current spot. It is also built on assumptions that…
A sharp move in either direction always arrives with confident explanations attached within the hour. Most of them are stories fitted to the candle…
Learning to read a chart is a genuinely useful literacy. Believing the chart predicts the future is a different claim entirely, and the evidence for…
The week’s Bitcoin news, one reasoned take, and the on-chain number that mattered. Free, weekly.
Signal, not noise. Unsubscribe any time.