Stablecoins: what the reported $891 billion in transfers actually measures

Adjusted volume rather than every movement

On 20 October 2026, Coinbase presented a study by The Block Pro Research on stablecoin usage. The report estimates adjusted transfer volume at $891 billion in the third quarter of 2026, compared with $216 billion in the first quarter of 2023. This increase concerns an indicator defined by the authors, rather than the raw total of movements recorded across all blockchains.

The data for this series comes from Allium. Its scope excludes activities associated with exchanges, decentralised finance, bridges and other infrastructure, as well as automated transactions and the issuance or redemption of stablecoins. These exclusions must accompany the figure whenever it is cited. Without them, two totals covering the same assets may appear contradictory even though they measure different activities.

The 73% outside banking hours comes from a different calculation

The report also states that 73% of filtered transfer volume occurs outside US banking hours. This measurement covers the twelve months ending on 31 July 2026. It uses Dune Analytics data and includes automated transactions, unlike the adjusted quarterly series. It excludes small transfers below one dollar, certain internal transfers, and minting or burning transactions, among other categories.

It would therefore be incorrect to multiply the $891 billion directly by 73% to obtain a quarterly amount transferred outside banking hours. The periods do not coincide and the filters differ. The first figure describes adjusted transfer value over a quarter; the second describes the distribution by time within a different set of observations.

An address is not a person

In its presentation, Coinbase also cites 293 million addresses holding stablecoins. The company specifies that this total does not equal the number of individual users. One person can control several addresses, while a custodial address can pool funds belonging to many customers. Comparisons with banking application users or cards issued therefore remain reference points for scale.

The same care is required when considering transfers. An asset moved several times can contribute several times to volume. Without additional data, a monetary total cannot reveal the number of merchants paid, recipient families or distinct customers. It provides information about observed circulation within the chosen scope.

Projections remain conditional

The study considers adjusted volume approaching $1,500 billion per quarter by the end of 2027 if the assumed growth rate continues. Coinbase presents this path as conditional. It is not an already observed measurement and does not guarantee that growth will remain unchanged over subsequent quarters.

This research was commissioned by a commercial participant in the ecosystem. That origin must remain visible to provide context for its results and presentation choices. It does not remove the need to examine the definitions, data sources and limitations set out in the report itself.

The study's contribution lies in making it possible to track series that are more precise than raw volumes. Comparing stablecoins with other payment methods still requires matching the periods, types of transaction and units being counted. A large volume, an active address and a commercial payment answer three different questions.