Ethereum holds 48.1% of the stablecoin supply behind Onchain Dollar Share.

• Largest network 48.1% Ethereum’s share of tracked supply
• Top two combined 78.0% The two largest networks together
• Network HHI 3,278 Sum of every named network’s squared percentage share. Higher means more concentrated; 10,000 would be a single network.
• Effective networks 3.05 How many equally sized networks would produce this concentration. Far below the raw count when a few networks dominate.

A further 3.81% of tracked supply sits in the record’s remainder, across networks it does not name individually. That share is excluded from the two measures above, so the concentration reported here is a lower bound. See classification and limitations.

At this record, USD stablecoin supply was 1.33% of US M2, or $1 of stablecoins for every $75 of US M2.

This compares the supply recorded on each network in the two frozen records named above. The differences are not observed transfers: the record contains no cross-chain flow data. Every row uses the same scale, set by the largest change in either direction.

• Networks expanded3of 10 named networks
• Networks contracted7of 10 named networks
• Unchanged0no recorded difference
• Net across the record+$933.5Mnamed networks and the remainder together

• Tron ▲expanded by +$822.9M
• Ethereum ▲expanded by +$119.8M
• All other networksremainder ▲expanded by +$89.8M
• Arbitrum ▲expanded by +$27.0M
• X Layer ▼contracted by −$100K
• Base ▼contracted by −$2.3M
• Polygon ▼contracted by −$5.2M
• BSC ▼contracted by −$14.1M
• Hyperliquid L1 ▼contracted by −$29.2M
• Avalanche ▼contracted by −$31.7M
• Solana ▼contracted by −$43.3M

Across all canonical records with a network breakdown, the largest network's share has fallen from 50.1% to 48.1%. The axis spans at least two percentage points, so a small change is not made to look large.

Where supply is recorded, not where it is used

Each figure is the stablecoin supply attributed to a network at the record’s cutoff. It is a stock, not a flow: it says nothing about transactions, settlement, payment activity or how often that supply moves.

No cross-network flows are measured

The record contains one daily supply figure per network. If one network expands while another contracts, those are separate changes in two balances between 12 August 2026 and the previous record. No cross-network movement is recorded.

Bridged and native representations are not distinguished

The record carries no native-versus-bridged classification, so this page makes none. Where an asset is issued on one network and represented on another, attribution follows whatever the source dataset reports, and the page cannot separate the two.

Network identity follows the source dataset

Names and boundaries are the source’s, and membership of the named rows can change between records: a network can fall into the remainder or rise out of it. When that happens the record states the crossing but no figure for it, and this page does the same.

The remainder is a bucket, not a network

All other networks is every network outside the 10 named rows, added together — $11.72B, or 3.81% of tracked supply. The record does not name its members, so they cannot be ranked, counted or charted individually, and their combined share is excluded from the HHI above.

The residual is the difference between the summed network breakdown and the record’s own total supply. It is not the remainder row: the remainder is supply attributed to networks outside the named rows, while the residual is supply the breakdown and the headline total account for differently. A small residual is expected, because network attribution and the headline total are captured from the same snapshot but reconciled independently. Tolerance is the threshold below which that difference is treated as immaterial.

See Methodology for definitions, the formula, and its caveats. Read the source figures in the Daily Tape record for 12 August 2026.