The Bank for International Settlements says Bitcoin on-chain transfer estimates can differ by a factor of six, and the reason is not magic, it is methodology.
- UTXO handling drives the gap in how Bitcoin transfers are counted.
- Public blockchain data is not self-explanatory; interpretation changes the result.
- Bitcoin is transparent, but not simple and neither are on-chain dashboards.
The BIS working paper, Hidden by complexity? Measuring stablecoin, crypto and decentralised finance ecosystems, examined roughly 100 billion blockchain records across Bitcoin, Ethereum, and Tron. Its main lesson is blunt: you can have public data and still end up with wildly different answers if you do not define what you are measuring.
For Bitcoin, the big headache is UTXOs, or unspent transaction outputs. Bitcoin does not work like a bank account with one running balance that simply moves around. Instead, it uses chunks of bitcoin that are created, spent, and then re-created as new outputs.
That matters because Bitcoin wallets usually spend entire UTXOs. If a wallet holds a 1 BTC UTXO and needs to send 0.5 BTC, the transaction typically consumes the full 1 BTC output and creates two new outputs: 0.5 BTC to the recipient and 0.5 BTC back to the sender as change.
Now ask a simple question: what should count as “transfer volume” in that transaction? The 0.5 BTC sent out? The full 1 BTC input that was consumed? Something adjusted in between? Different answers produce different totals, and the BIS says those choices can push Bitcoin transfer estimates apart by a factor of six.
That is not a minor disagreement. It is the difference between one dashboard and another telling two very different stories while both sounding authoritative. Numbers with no methodology are just pretty wallpaper.
The useful part of the BIS finding is not that Bitcoin data is unreliable. It is that blockchain data is easy to misread. Transparency is a real strength, but transparency does not remove the need for careful measurement. A public ledger can still be mangled by sloppy definitions, cherry-picked filters, and the usual crypto habit of treating one metric like it proves everything.
That distinction matters for more than Bitcoin. The BIS paper also looked at Ethereum and Tron, where the same problem shows up in different clothing. Ethereum activity is tangled up with smart contracts, token transfers, swaps, lending, approvals, and other actions that can turn one user move into a chain of on-chain events. Tron has its own patterns, especially around stablecoin usage. Same broad industry, different plumbing, different measurement traps.
The paper also found stablecoins are a major driver of trading activity across the chains studied. On Ethereum, stablecoins often sit inside smart contract ecosystems such as decentralized exchanges and lending protocols. On Tron, stablecoin holdings more often appear outside smart contracts, which points to a more direct transfer-and-hold pattern.
That is a good reminder that “crypto activity” is not one neat thing. A transfer, a payment, a swap, an exchange reshuffle, and a contract interaction are not interchangeable. Yet they are often lumped together by dashboards and commentators who want one clean number to fit a tidy narrative. Reality, as usual, refuses to cooperate.
Supporters of Bitcoin often point to on-chain data as proof of adoption and network strength. Critics use different measurements to argue the opposite. The BIS paper cuts through both camps’ favorite tricks: if you do not say exactly what you are counting, the headline number is mostly theater.
That does not make on-chain metrics useless. It makes them conditional. If you want to understand Bitcoin activity, you need to ask whether you are looking at settlement, transfer volume, exchange churn, wallet reshuffling, or actual economic payments. Those are not the same thing, and pretending they are is how people end up sounding confident while being wrong.
Even if you have no love for the BIS, and plenty of Bitcoiners have good reasons to treat the institution with skepticism, the methodological warning still lands. Counting blockchain activity is harder than bragging about it. A lot harder.
Key takeaways
-
Why can Bitcoin transfer estimates differ so much?
Because different methods treat UTXOs, change outputs, and transaction flow differently. The BIS says those choices can produce estimates that are six times apart. -
Does that mean Bitcoin on-chain data is useless?
No. It means the data must be interpreted carefully. Bitcoin’s blockchain is transparent, but transparency does not remove the need for clear measurement rules. -
What does UTXO mean in plain English?
It is the chunk-based accounting model Bitcoin uses. When a UTXO is spent, the full output is consumed and new outputs are created, including change back to the sender. -
Does this issue affect other blockchains too?
Yes. The BIS also examined Ethereum and Tron, showing that on-chain metrics depend heavily on each network’s structure and common user behavior. -
What should investors and analysts take from this?
Do not trust raw on-chain numbers without asking how they were built. If a dashboard does not define what it counts, the conclusion may be just as fuzzy as the metric.
The clean takeaway is simple: Bitcoin’s ledger is public, but the story told from that ledger can change dramatically depending on the rules used to count it. For a space obsessed with transparency, that should be a useful wake-up call.
Further reading
A few related pieces that add more context to the measurement debate, payments trends, and policy battles around Bitcoin and stablecoins:
