The Ledger Is Visible, the Economy Is Not
The BTC–Ethereum whale comparison now has a measurement problem to solve. The study observes wallet activity and price returns around public alerts, but a public ledger is only one layer of a crypto market: exchange trading can settle internally, addresses do not name their owners, and the same transaction count can represent very different economic value. This chapter uses the BIS Project Atlas model to trace how raw records become indicators and to show why coverage and attribution limits should change how we interpret a muted Ethereum response or a broad Bitcoin response.
A public ledger is only one layer
A blockchain is unusually rich as a data source. It can preserve a record of transactions, amounts, timestamps and addresses that anyone can inspect. For Bitcoin-like ledgers, historical records can reveal inputs, outputs, transaction sizes, holdings and network relationships. That makes on-chain data valuable for studying large transfers: a researcher can locate an event, connect it to a public alert and examine nearby activity without asking a central bank or exchange for a private transaction file.
But the ledger does not record every economic action that surrounds a cryptoasset. When two customers trade inside the same centralised exchange, the exchange can settle the trade on its own books. That internal movement does not need to appear as a new public blockchain transaction. The same is true of other activity carried out in venues that use the blockchain only for deposits, withdrawals or periodic settlement. A ledger can therefore show the exchange's balance movements while hiding the customer-level trading that produced them.
That distinction matters directly for whale alerts. The Philadelphia Fed study measures non-whale wallet activity around selected on-chain events and notes that its analysis does not include transactions with crypto exchanges. A large transfer involving an exchange may still be meaningful, but it does not offer the same view of who traded inside the exchange or why the exchange moved its inventory. On Ethereum, the gap can be wider because activity also passes through smart contracts, tokens and layer-2 venues. A retail response may be real and economically relevant while failing to appear as a direct retail wallet transfer in the measured window.
Project Atlas was built around this gap between abundant raw records and reliable economic indicators. Its first principle is to combine on-chain data gathered from public ledgers or blockchain nodes with off-chain data reported by exchanges and other market actors. The two sources answer different questions. On-chain records show what was written to the ledger. Off-chain records can add trading volumes, exchange activity or other events that were settled privately. Neither source is complete by itself, and disagreement between them is not automatically an error.
Atlas turns this combination into an analytical pipeline. Incoming raw records are retained, then cleaned and normalised so that data from different sources can be compared. Granular and aggregated data can then support research or dashboards.