Shapes in these charts that are not yet explained
Some things visible in these series are unresolved. They are listed because a reader can see them, and a shape we cannot explain is worse unmentioned than admitted.
Argentine transfer size collapsed in January 2026: count up 6.9 times on flat dollar volume. Cause unidentified.
The nBRL spike is now explained, and has been removed rather than annotated. Two isolated weeks near $70m in late 2024 were one issuer passing a round 400,000,000 nBRL between two of its own wallets, part of a 1,000,000,000 deployer mint that was burned down in stages over eight months. Both wallets are now registered, which takes peak nBRL supply from 900,031,397 to 1,946,954. Every other movement in that token's history is three digits.
The week of 2026-01-26 holds $113m of BRZ volume against a roughly $9m baseline, and it is not treasury movement. The treasury exclusion leaves it unchanged.
BRZ treasury activity stops in 2026: 67.5% of 2025 BRZ volume ran through the three registered treasury addresses, then 0.1% in 2026. Either those operations ceased or the issuer moved to wallets not yet identified. Until that is settled, treat 2026 BRZ volume as possibly still containing treasury flow.
The treasury register was incomplete for BRZ itself until August 2026, and that is the honest way to describe its limits. Three addresses were registered; a fourth on BNB Chain held 98.2% of that chain's BRZ, and tracing it backwards found five predecessors, because the issuer rotates the reserve between wallets. Registering only the current one made BRZ supply read a flat R$100m for two years and then fall off a cliff. All nine are now registered.
That sweep has only been done for BRZ. Other issuers likely have equivalent wallets that have not been identified, and ranking present-day holders finds only the wallet in use today. Absence from the register is not evidence that a token is clean.
Coverage is not complete, and one case is measured. MXNB is ingested on two of the five chains it runs on, so every MXNB figure here is a lower bound rather than a measurement. It was found only because an external benchmark existed to compare against; no equivalent check has been run on the other tokens.