Work Writing About Work with me

Stablecoins: Past, Present and FutureChapter 5 of 12

Which chains stablecoins live on, and who distributes them

A stablecoin is issued once but lives on many blockchains, and the chain it sits on changes what it is used for. This chapter maps where the dollars sit, explains why the same ticker on two chains can be two different claims, and shows why the firms that put a token in front of users hold more power than the firms that write its code.

What you'll learn

  • Name the chains that hold most stablecoin supply and say what each one is used for.
  • Explain why Tron, a chain few engineers admire, carries most of USDT.
  • Tell a native token from a bridged one, and say whose promise stands behind each.
  • Explain how Circle's CCTP and Tether's USDT0 move tokens between chains, and where the risk sits.
  • Describe distribution as the moat, using Coinbase's USDC deal and PayPal's PYUSD as cases.
In this chapter

Where the dollars sit

On 4 September 2026 DefiLlama counted about 312 bn USD of stablecoins across all blockchains. Two chains held almost four fifths of it: Ethereum with about 147 bn USD and Tron with about 94 bn USD. Solana, BNB Chain and Hyperliquid followed a long way behind.

Stablecoin supply by chain, 4 September 2026Ethereum held 147.15 bn USD (48.31%) and Tron 93.70 bn USD (30.77%). Solana, BNB Chain and Hyperliquid L1 make the visible tail. Shares divide by a chain-level total, not by the 312.0 bn USD all-pegs headline.Stablecoin supply by chain, 4 September 2026bn USD050100150Ethereum: 147.15 bn USD147.15EthereumTron: 93.7 bn USD93.7TronSolana: 16.21 bn USD16.21SolanaBNB Chain: 13.34 bn USD13.34BNB ChainHyperliquid L1: 6.99 bn USD6.99Hyperliquid L1Source: DefiLlama, stablecoin supply by chain, read 4 Sep 2026. Chain shares are not the same metric as Ethereum's ~65% of non-USD stablecoin supply (Dune/Visa).
Stablecoin supply by chain on 4 Sep 2026: Ethereum and Tron hold most of it, with Solana, BNB Chain and Hyperliquid far behind (DefiLlama).

Each chain earns its share for a different reason. Ethereum is where most lending markets, exchanges and trading protocols live, so a large stock of USDC and USDT sits there as collateral and working capital. Aave, the largest lending protocol, held about 17.7 bn USD of deposits on 4 September 2026. Tron is where people send USDT to each other. Solana is fast and cheap, and it has become a large venue for USDC payments and trading. Hyperliquid is a chain built for one exchange; almost all of its stablecoins are USDC posted as margin for futures trading.

Below them sit the layer 2 networks: chains such as Base and Arbitrum that run on top of Ethereum, batch many transactions together and settle them on Ethereum, which makes each transfer far cheaper. Base, launched by Coinbase, holds a large stock of USDC.

The practical lesson is that "which chain wins" depends on the question. Ethereum wins on supply and on depth of financial markets. Tron wins on person-to-person transfers. A chart of total supply alone hides that split.

Why Tron carries most of USDT

Tron holds about half of all USDT, more than Ethereum. It got there by being cheap and by being first in the places where people needed dollars.

The mechanism is Tron's fee model. Instead of paying gas on every transfer, a user can lock up, or stake, the chain's own token, TRX, and receive a daily allowance of network resources. A heavy user who stakes enough TRX sends USDT for close to nothing. A casual user who does not stake pays by burning TRX, roughly 1 to 3 USD per transfer, which is still less than a bank wire to many countries. Tether later added a program that lets users pay that fee in USDT itself, so a phone with only dollars on it can still send.

The result is a chain used mostly for payments between people. CoinDesk Research found on 21 July 2026 that 93% of Tron's stablecoin transfer volume was peer-to-peer, the highest share of any chain it tracked. For many users it works as a dollar account that needs no US bank.

Largest token's share of each chain's stablecoin supplyTron is 98% USDT. Hyperliquid L1 is 98% USDC. X Layer is 93% USDG. XRPL is 91% RLUSD. Ethereum, the largest chain by supply, is not on this chart because it is not a near-total monoculture.Largest token's share of each chain's stablecoin supply% of that chain020406080100Tron, USDT: 98 %98Tron, USDTHyperliquid, USDC: 98 %98Hyperliquid, USDCX Layer, USDG: 93 %93X Layer, USDGXRPL, RLUSD: 91 %91XRPL, RLUSDchain_monoculture.csv via the chapter brief, 4 Sep 2026. The Polish DefiLlama read the same day puts Tron at 97.9% USDT and Hyperliquid at 97.9% USDC; the CSV rounds both to 98%. Not averaged.
The largest token's share of each chain's stablecoin supply on 4 Sep 2026: Tron is almost all USDT, Hyperliquid almost all USDC, X Layer mostly USDG and XRPL mostly RLUSD.

Tron also shows chain monoculture, a chain whose stablecoin supply is almost entirely one token. About 98% of Tron's stablecoins are USDT. Hyperliquid is the mirror image with USDC. Monocultures are sticky: once every wallet, exchange and merchant on a chain uses one token, a rival has to win them all over at once.

Native and bridged tokens

The same ticker on two chains can be two different promises. The difference decides who owes you a dollar if something breaks.

A native token is minted directly by the issuer on that chain and counts in the issuer's reserve. Native USDC on Solana is Circle's own obligation, redeemable at Circle like USDC on Ethereum. A bridged token is a copy created by a third-party bridge: the bridge locks the original token on one chain and mints a stand-in on another. The stand-in is a claim on the bridge, not on the issuer.

The classic example is USDC.e on Arbitrum. Before Circle issued native USDC there in June 2023, the USDC on Arbitrum was a bridged copy backed by tokens locked on Ethereum. It did not appear in Circle's reserve reports, and redeeming it meant going back through the bridge first.

The danger of the bridged model is the pool of locked tokens. If a bridge is hacked or its operators disappear, the stand-ins lose their backing while the issuer's reserve stays intact. In July 2023 the Multichain bridge collapsed and more than 120 mn USD was drained from its Fantom side. Bridged USDC on Fantom fell to about 22 cents, while native USDC everywhere else stayed at one dollar. Circle froze some of the stolen funds, but it had never owed the bridged copies anything.

Burn-and-mint never builds a locked pool; lock-and-mint always doesNative USDC on a chain is Circle's own redeemable obligation. A bridged token is a claim on a bridge that holds the original elsewhere. Circle's CCTP burns on the source chain and mints on the destination, so no honeypot pool sits in between. USDT0 and historical lock-and-mint bridges lock the original and issue a wrapped claim against that pool.Burn-and-mint never builds a locked pool; lock-and-mint always doesBurn-and-mint (CCTP): no locked poolLock-and-mint (USDT0, old bridges): locked poolmint on Bmint wrappedNative USDC on A: Circle obligationNative USDC on ACircle obligationBurn on A: attestation, then mintBurn on Aattestation, then mintNative USDC on B: same issuer claimNative USDC on Bsame issuer claimLock USDT on A: Ethereum lockboxLock USDT on AEthereum lockboxLocked pool: the honeypotLocked poolthe honeypotWrapped token on B: claim on the bridgeWrapped token on Bclaim on the bridgeSource: Circle CCTP V2, 11 Mar 2025; Tether/LayerZero USDT0, 10 Feb 2026; CoinDesk on the Multichain exploit, Jul 2023
Burn-and-mint never creates a locked pool of tokens; lock-and-mint always does, and that pool is what attackers target.

How tokens move between chains

Issuers have answered the bridge problem in two ways, and the two largest chose different paths.

CCTP, Circle's Cross-Chain Transfer Protocol, launched in 2023 and works by burn-and-mint: a user burns USDC on the source chain, Circle's service confirms the burn, and native USDC is minted on the destination chain. No pool of locked tokens exists at any point, so there is nothing to steal in the middle. Version 2, released on 11 March 2025, added fast transfers that settle in seconds. Circle reported more than 110 bn USD moved through CCTP by 14 November 2025. Circle also created a standard, in November 2023, for converting a third-party bridged USDC into native USDC without disrupting holders, and native USDC grew from 11 chains in early 2024 to 28 in April 2026.

USDT0 is Tether's lock-and-mint answer: a version of USDT built on LayerZero's cross-chain token standard and run by a separate operator, Everdawn Labs. Real USDT sits locked in a contract on Ethereum, and USDT0 is minted on other chains against it. The model is popular: Tether reported more than 70 bn USD of transfers through USDT0 in under a year, and invested in LayerZero Labs on 10 February 2026. It keeps a locked pool, so a failure of the bridge contracts or its verifiers would hit USDT0 holders, not holders of USDT on Ethereum.

The new payment chains

Since 2025 a group of chains has been built specifically for stablecoins. A stablecoin-first chain uses a dollar token to pay transaction fees and aims for near-zero cost and settlement in under a second. The best-known are Plasma, backed by the Tether ecosystem; Tempo, incubated by Stripe; Arc, built by Circle; Stable, linked to Bitfinex; and Codex.

They are still small. Plasma, the largest by stablecoin supply, held about 0.9 bn USD on 4 September 2026, under 0.3% of the market. It launched in September 2025 with billions in deposits attracted by incentives, and most of that money left within months when lending protocols pulled out. Tempo opened its mainnet on 18 March 2026, and outside validators including Stripe and Visa joined in April; the network remains permissioned.

Arc is the one to watch. Circle opened its public mainnet on 16 September 2026, with a founding group of validators that includes BlackRock, DTCC, Mastercard, Standard Chartered and Visa. Fees are paid in USDC. These chains matter because each one is an issuer or payment company trying to own the rails its token runs on, instead of renting space on Ethereum or Tron.

Distribution is the moat

Anyone with a banking partner and a lawyer can issue a fully backed dollar token in a few weeks. The hard part is getting it into the apps where people already keep money. Distribution is that path from issuer to user: exchanges, wallets, fintech apps, payment processors, merchants, lending protocols and custodians, each of which wants a cut.

Nine paid channels between issuer and walletA fully reserved token can be copied in weeks. Placement on an exchange, a wallet, a PSP, a bank mandate or a custody desk cannot. Known cuts are labelled; a dash means the fee was not established in these sources.Nine paid channels between issuer and walletknown cutWallet, custodial or self-custody: Coinbase Wallet, exchange app, self-custody · —Wallet, custodial or self-custodyCoinbase Wallet, exchange app, self-custody—Centralised exchange: Coinbase, Binance, Kraken · rev-shareCentralised exchangeCoinbase, Binance, Krakenrev-shareFintech / neobank: Nubank, Revolut · embeddedFintech / neobankNubank, RevolutembeddedPayment service provider: Stripe, Bridge, PayPal · 1.5%Payment service providerStripe, Bridge, PayPal1.5%Merchant checkout: PayPal, Stripe merchants · vs 2.9%+$0.30Merchant checkoutPayPal, Stripe merchantsvs 2.9%+$0.30Bank partner: DBS for USDG · custodyBank partnerDBS for USDGcustodyBridge: CCTP, USDT0 · —BridgeCCTP, USDT0—DeFi listing: Aave, Morpho, Hyperliquid · listingDeFi listingAave, Morpho, HyperliquidlistingInstitutional custody / OTC: Coinbase, OTC desks · AUM feeInstitutional custody / OTCCoinbase, OTC desksAUM feeSource: Circle Internet Group, Form S-1 and 10-K (Coinbase collaboration and distribution costs), 2024–2025 filings. Not an invoice, 4 Sep 2026.
The paid channels between a stablecoin issuer and a user's wallet, from exchanges and fintech apps to payment processors and custodians, with the known cuts as of 4 Sep 2026.

The clearest case is the deal between Circle and Coinbase. Their agreement of August 2023 gives Coinbase all the reserve income earned on USDC held on Coinbase's platform, and half of the remaining reserve income on USDC held elsewhere. Circle paid Coinbase about 908 mn USD under that deal in 2024. The agreement was renewed in August 2026 through 2029 on the same structure. Circle mints the token and runs the reserve, yet the exchange that puts USDC in front of users takes a share of income close to what Circle keeps. The business chapter runs the numbers.

PayPal shows the other side. Its PYUSD token, issued by Paxos, grew fast while PayPal paid holders rewards, peaking at about 4.1 bn USD in early 2026. When the rewards campaign ended, supply fell to about 3.0 bn USD by September. The token did not change. The incentive in the distribution channel did.

Some distributors are becoming issuers. Revolut, with about 80 mn retail customers, launched its own euro stablecoin on 26 August 2026, issued through a Luxembourg entity of Bridge, the stablecoin firm Stripe bought in 2025. The same week it dropped USDT for customers in the European Economic Area because of EU rules. Nubank in Brazil went the partnership route with Circle. Both moves follow the same logic: whoever owns the customer relationship decides which dollar the customer holds.

Concentration is the result. On 4 September 2026 USDT and USDC together held about 83% of dollar stablecoin supply, according to DefiLlama. New issuers with strong backers grew quickly in 2026, but from a small base.

What to watch next

  • Arc's first months of real use: Circle's chain went live on 16 September 2026. Watch how much USDC settles there and whether its big-name validators route their own business onto it.
  • USDT0 versus native USDT: Tether expands across chains through a locked pool on Ethereum. Growth there raises the cost of any failure in the bridge contracts.
  • Tron's share of payment volume: Tron's lead in peer-to-peer transfers rests on cheap fees and habit. A cheaper, regulated alternative in its core markets would show up here first.
  • Distribution deals after US rules take effect: the US yield ban and a pending federal rule on affiliate rewards could force changes to reward programs like Coinbase's, which would shift where USDC is held.
  • Fintechs issuing their own tokens: Revolut's euro token is an early case. If more large apps issue rather than distribute, the two leaders lose shelf space.

Key takeaways

  1. Stablecoin supply is concentrated on two chains, Ethereum and Tron, and each serves a different use: financial markets on Ethereum, person-to-person transfers on Tron.
  2. Tron won USDT payments through cheap fees and early reach in emerging markets, which shows that technical elegance does not decide where money moves.
  3. A native token is the issuer's own promise; a bridged token is a promise from the bridge, and it can fail while the issuer's reserve is untouched.
  4. Burn-and-mint transfers, such as Circle's CCTP, avoid a locked pool of tokens; lock-and-mint transfers, such as USDT0, keep one, and that pool is the risk.
  5. Stablecoin-first chains are an attempt by issuers and payment firms to own their rails, but in September 2026 they held a small fraction of supply.
  6. Distribution is the moat: a token can be copied in weeks, while access to exchanges, wallets and fintech apps takes years and costs a large share of reserve income.

Glossary

Layer 2 network
a chain that runs on top of Ethereum, batches transactions and settles them on Ethereum. Base and Arbitrum are examples.
Chain monoculture
a chain whose stablecoin supply is almost entirely one token, such as USDT on Tron.
Native token
a token minted by the issuer directly on a chain and counted in its reserve.
Bridged token
a copy minted by a third-party bridge against tokens locked on another chain; a claim on the bridge, not the issuer.
CCTP
Circle's Cross-Chain Transfer Protocol, which burns USDC on one chain and mints native USDC on another.
USDT0
Tether's cross-chain version of USDT, minted on other chains against USDT locked on Ethereum.
Stablecoin-first chain
a blockchain designed around stablecoin payments, with fees paid in a dollar token.
Distribution
the paid path from issuer to user through exchanges, wallets, fintech apps, payment processors and custodians.

Go deeper

Sources

🟢 primary · 🟡 credible secondary · 🔴 tertiary (never used to cite a number)

All chapters of "Stablecoins: Past, Present and Future" →