Ethereum Stablecoins Explained

Stablecoins are the settlement layer of the Ethereum economy: dollar-pegged tokens that let traders price assets, park capital, and move value without leaving the chain. But "stable" is an engineering claim, not a guarantee — and the four major designs achieve their peg in fundamentally different ways, with fundamentally different failure modes. This guide maps the full taxonomy and the history of what happens when pegs break.

Why stablecoins dominate Ethereum activity

Almost everything in Ethereum DeFi is priced, borrowed, or settled in stablecoins. They are the quote asset in most DEX pairs, the dominant borrowable asset on lending markets like Aave, and the unit in which traders take profit without off-ramping to a bank. They also serve users far beyond trading: in high-inflation economies, dollar stablecoins on Ethereum and its Layer 2 networks function as accessible dollar accounts.

That centrality is why stablecoin design matters so much. When a major stablecoin wobbles, everything priced in it wobbles too — collateral values, liquidation thresholds, LP positions. You can see how much of the ecosystem's value sits in stablecoins and stablecoin-adjacent protocols on the DeFi bubble map.

The stablecoin market in 2026: a snapshot

Before the taxonomy, the scale. The total stablecoin market sits at roughly $300 billion as of mid-2026 — around $303 billion in July 2026, per CoinLaw's market statistics — up from about $124 billion at the end of 2023, per Bessemer Venture Partners. The striking feature of 2026 is not growth but a plateau: after the vertical expansion of 2024 and early 2025, total supply has flatlined in the $300–305 billion range since late 2025. Concentration remains extreme:

Supply figures per CoinLaw, DailyCoin, and American Banker (as of mid-2026). Two structural notes: first, the two fiat-backed giants control roughly 85% of all supply, so the "decentralized stablecoin" story remains a niche within a centralized market. Second, the decentralized leaders swapped places — Sky's USDS (about 94% of it on Ethereum) grew to roughly twice the size of Ethena's USDe after the yield-driven flows of 2024–2025 cooled. Ethereum remains the dominant settlement chain for stablecoin supply, ahead of Tron, per CoinLaw's chain-share data.

The four families of stablecoin

Every stablecoin answers the same question — "why is this token worth $1?" — with one of four mechanisms:

Fiat-backed: USDT and USDC

The two giants of the category are structurally similar — a centralized issuer holds dollar assets and issues tokens redeemable 1:1 — but differ meaningfully in transparency, jurisdiction, and reserve composition.

USDC (Circle)

Circle describes USDC as backed 100% by highly liquid cash and cash-equivalent assets. The majority of reserves sit in the Circle Reserve Fund, an SEC-registered government money market fund managed by BlackRock, holding cash, short-dated U.S. Treasuries, and overnight Treasury repurchase agreements, with the remainder in cash at large banks. Circle publishes monthly third-party attestations verifying that reserves exceed USDC in circulation, and Deloitte serves as its independent auditor — all per Circle's transparency page (as of August 2026).

USDT (Tether)

Tether is the largest stablecoin by circulation and publishes quarterly attestations prepared by the accounting firm BDO rather than monthly ones. Its reserves are dominated by U.S. Treasury exposure — Tether reported a record of roughly $141 billion in direct and indirect Treasury exposure across its 2025 attestations — alongside smaller allocations that have included gold and bitcoin, per Tether's attestation announcements (as of late 2025). Tether's inclusion of non-traditional reserve assets and its offshore domicile are exactly the points its critics raise; its scale and profitability are the points its defenders raise.

The freeze function

Both issuers deploy token contracts with blacklist capabilities and have frozen addresses at the request of law enforcement. A blacklisted address can no longer move its tokens. For most users this is invisible; for the system it is fundamental: fiat-backed stablecoins are censorable by design, which is the main philosophical trade against decentralized alternatives. It also means "your" USDC or USDT is ultimately a claim on a company honoring redemption — a very different object from ETH itself, whose supply and transfer rules live entirely on-chain (see our ETH asset page for the base asset by contrast).

USDT vs USDC at a glance

The split is increasingly geographic: USDT is the offshore dollar, USDC the regulated one. The same trade-offs — reach and liquidity versus disclosure and legal recourse — repeat in nearly every stablecoin decision.

Crypto-collateralized: DAI, USDS, and the vault model

DAI, launched by MakerDAO, answers the centralization problem with overcollateralization. To mint DAI, a user locks collateral — ETH, staked ETH, or other approved assets — into a vault (historically called a CDP, collateralized debt position) and may borrow DAI against it only up to a fraction of the collateral's value. Every DAI in existence is backed by more than a dollar of assets held in smart contracts anyone can inspect.

The peg is defended mechanically:

• Liquidations. If a vault's collateral value falls below its required ratio, the protocol auctions the collateral to buy back and burn the vault's DAI debt, restoring full backing before the position becomes insolvent.
• Interest rates. Borrowing costs (stability fees) and the savings rate paid to DAI holders are governance levers that expand or contract DAI supply to steer the price toward $1.
• The Peg Stability Module. A facility that swaps DAI 1:1 against other stablecoins such as USDC, which hardens the peg but — importantly — reintroduces centralized collateral into a nominally decentralized system. Over the years, a substantial share of DAI's backing has come from USDC and real-world assets, a trade-off DAI's own community debates constantly.
• Emergency shutdown. A last-resort mechanism that settles the system and lets holders redeem collateral directly.

In August 2024 MakerDAO rebranded to Sky and introduced USDS, an upgraded stablecoin that DAI holders can convert to 1:1, per Blockworks. The rebrand was contentious precisely along the censorship axis discussed above: USDS is upgradeable and designed to support compliance features such as a freeze function, while the original DAI contract is immutable and continues to exist unchanged. The two-token outcome is a live experiment in whether users prefer regulatory compatibility or immutability.

The cost of the vault model is capital inefficiency — you must lock more than a dollar to mint a dollar — and its tail risk is a collateral crash faster than liquidations can clear, as happened during Ethereum's "Black Thursday" crash of March 2020, when network congestion (see our gas fees guide for why congestion matters) disrupted Maker's auctions and left the system briefly undercollateralized.

Synthetic dollars: Ethena's USDe

Ethena's USDe, launched in 2024, is the most consequential new design since DAI — and Ethena itself is careful to call it a "synthetic dollar" rather than a stablecoin. Instead of holding dollars or overcollateralized crypto, USDe holds a delta-neutral position: for every dollar of USDe, roughly a dollar of crypto collateral (largely liquid staking tokens such as stETH, plus other assets) paired with an equal-sized short position in perpetual futures on derivatives exchanges, per Ethena's documentation. If ETH falls 10%, the spot collateral loses 10% and the short gains 10%: the dollar value holds.

The design generates yield from two sources: staking rewards on the collateral, and the funding rate that leveraged long traders pay to shorts in perpetual futures markets — a flow that is usually positive in crypto because speculative demand skews long. Holders can stake USDe into sUSDe to receive that yield.

The risks are equally structural:

• Funding risk. Funding rates can go and have gone negative during risk-off periods. A sustained deeply negative regime would turn the yield engine into a cost that erodes the reserve buffer.
• Counterparty and custody risk. The short legs live on centralized derivatives exchanges, with collateral held via off-exchange settlement custodians. An exchange failure mid-position is the nightmare scenario — the hedge would need to be rebuilt elsewhere in a stressed market.
• Liquidity mismatch. Direct mint/redeem is limited to approved participants; everyone else exits via DEX pools, so secondary-market depth determines how well the peg holds in a rush — the same dynamic our liquidity guide describes for any pegged asset.

USDe scaled to a multi-billion-dollar supply within its first year and has held its peg through several market shocks, but its risk profile is closer to a hedge fund strategy wrapped in a token than to a bank deposit. Its yield is compensation for real risk, not free money.

Case study one: the UST death spiral (May 2022)

TerraUSD (UST) was the flagship of the fourth family: the algorithmic stablecoin, backed not by assets but by an arbitrage loop with a volatile sister token, LUNA. Burn $1 of LUNA, mint 1 UST; burn 1 UST, mint $1 of LUNA. As long as LUNA had value and holders believed in the mechanism, arbitrageurs kept UST at $1 — and Terra's Anchor protocol paid roughly 19.45% yield on UST deposits, which pulled in billions of demand, per MIT Sloan's analysis.

In early May 2022, large UST withdrawals and sales broke the peg. Redemptions then minted enormous quantities of LUNA, hyperinflating its supply; LUNA's price collapse destroyed the backing UST depended on, which triggered more redemptions — the "death spiral." LUNA fell from an April peak above $116 to fractions of a cent by May 13, and roughly $45 billion of combined market value across UST and LUNA evaporated within about a week. The collapse propagated through lenders and funds and set off the 2022 crypto credit crisis. It remains the definitive demonstration of why a stablecoin backed chiefly by confidence in its own ecosystem cannot survive a genuine run.

Case study two: USDC and Silicon Valley Bank (March 2023)

The counterpoint case shows that full reserves do not eliminate depeg risk — they relocate it. On March 10, 2023, Silicon Valley Bank failed. Circle disclosed that $3.3 billion of USDC's cash reserves — nearly 8% of the cash portion — was held there, and over that weekend USDC traded as low as roughly $0.87, per CNBC. DAI, then heavily backed by USDC via the Peg Stability Module, depegged in sympathy — a reminder that stablecoin risks are correlated through shared collateral.

When U.S. regulators announced that all SVB depositors would be made whole and Circle committed to covering any shortfall, the peg recovered within days. Two durable lessons: first, a fiat-backed stablecoin is only as sound as the banks and instruments holding its reserves; second, depegs are amplified by market structure — redemptions were closed for the weekend, so all price discovery happened in DEX pools and on exchanges, where thin weekend liquidity exaggerated the move.

Case study three: Stream Finance's xUSD and the yield-token contagion (November 2025)

The most instructive recent failure involved neither an algorithmic design nor a bank. Stream Finance's xUSD marketed itself as a yield-bearing stable asset: deposits were deployed into external trading strategies, and the token's value was meant to track a growing pool of assets. On November 4, 2025, Stream disclosed that an external fund manager had lost roughly $93 million of the platform's assets. Within 24 hours xUSD fell from $1 to about $0.26, and by the end of the week it traded between $0.07 and $0.14, per KuCoin's crisis coverage.

The contagion mattered more than the initial loss. Elixir had lent Stream $68 million — roughly 65% of the reserves backing its own deUSD stablecoin. When Stream froze withdrawals, Elixir could not honor redemptions; deUSD collapsed about 98% and the project subsequently shut down, per Yahoo Finance. Because lending markets including Euler, Morpho, Silo, and Gearbox had accepted these tokens as collateral, researchers traced roughly $285 million of interconnected debt through the affected vaults, and curators froze markets across DeFi.

Three lessons. First, a token that deploys its backing into opaque, off-chain trading strategies is a fund share with a $1 sticker, not a stablecoin — the "family" framework above only works if you can verify which family you are actually in. Second, stablecoins that hold other stablecoins as reserves inherit each other's failure modes, exactly as DAI inherited USDC's in 2023, but with leverage added. Third, the yield is the tell: xUSD paid double-digit returns in a market where T-bills paid a fraction of that. As with UST's 19.45%, the premium was the risk premium.

Yield-bearing stablecoins and tokenized T-bills

A structural shift since 2023: the yield on reserves increasingly flows to holders, not just issuers.

• Savings wrappers. sUSDe (Ethena) passes through basis-trade yield; sDAI and sUSDS (Sky) pass through the protocol's savings rate. These are stablecoin balances that grow — with all the underlying protocol's risks attached.
• Tokenized money market funds. BlackRock's BUIDL, launched on Ethereum in March 2024 with Securitize as transfer agent, tokenizes a fund of T-bills, repo, and cash, paying daily dividends on-chain; it became the largest tokenized Treasury fund, crossing $1 billion in assets by March 2025, per The Block. Products like this blur the line between stablecoins and securities — BUIDL is restricted to qualified investors, not freely tradeable.

By 2026 this category has become an industry of its own. Tokenized US Treasury products held roughly $16 billion as of July 2026, the largest category of tokenized real-world assets, per rwa.xyz data summarized by MetaMask's RWA overview. BUIDL grew past $2.8 billion and has distributed over $100 million in dividends since inception, Franklin Templeton's BENJI reached about $2.4 billion, and Circle's tokenized money fund USYC — around $3 billion in mid-2026 — overtook BUIDL as the largest tokenized Treasury fund. The competitive logic is straightforward: under the GENIUS Act, payment stablecoin issuers cannot pass reserve interest to holders, so yield-seeking balances migrate to tokenized funds and DeFi savings wrappers that sit just outside the payment-stablecoin perimeter.

When you see a stablecoin advertising yield, always ask where the yield originates: T-bill interest, derivatives funding, lending markets, or token emissions. The source defines the risk. Our guide on reading TVL applies the same source-of-truth skepticism to protocol metrics.

The regulatory landscape, briefly

Stablecoins are now squarely inside the regulatory perimeter in the two largest Western markets:

• United States — the GENIUS Act. Signed into law on July 18, 2025, it establishes a federal framework for "payment stablecoins": only licensed entities may issue them, reserves must be 100% backed by liquid assets such as cash and short-term Treasuries, and issuers must publish monthly reserve disclosures, per Covington's legal analysis.
• European Union — MiCA. The stablecoin provisions of the Markets in Crypto-Assets regulation have applied since June 30, 2024, requiring fiat-pegged tokens to be issued by authorized e-money institutions. Circle obtained authorization for USDC through its French entity; Tether declined to seek authorization, and MiCA-compliant exchanges progressively delisted USDT for European Economic Area users from late 2024, per Scorechain's MiCA overview.

GENIUS implementation: the 2026 rulemaking wave

Signing the GENIUS Act was the start, not the end. The statute gave federal agencies until July 18, 2026 to issue implementing regulations, and the first half of 2026 brought a coordinated wave of proposals, per Morgan Lewis's implementation tracker:

• AML and sanctions. On April 8, 2026, FinCEN and OFAC jointly proposed rules requiring permitted issuers to run full anti-money-laundering and sanctions compliance programs — including the technical capability to freeze or block tokens when legally required, per the US Treasury's announcement. The freeze function discussed earlier is becoming a legal mandate, not just an issuer policy.
• Licensing pipelines. The OCC proposed its framework for approving national payment stablecoin issuers, per OCC Bulletin 2026-3, and the FDIC and NCUA proposed parallel application processes for banks and credit unions.
• State regimes. Treasury proposed the principles for deciding when a state regulatory regime is "substantially similar" to the federal one — the gate through which smaller state-chartered issuers must pass.

Direction of travel: fiat-backed stablecoins are converging on regulated, T-bill-backed, disclosure-heavy models, while decentralized and synthetic designs occupy the space regulation doesn't cleanly cover.

The new entrants: banks, fintechs, and Tether's American answer

A clear legal framework invited a new class of issuer, and 2025–2026 delivered them:

• Tether's USAT. USDT itself does not fit the GENIUS framework, so Tether launched USAT on January 27, 2026 — a separately issued, US-regulated stablecoin distributed through Anchorage Digital Bank, the first federally chartered crypto bank, per Kavout's 2026 landscape review. The world's largest issuer now runs a two-coin strategy: offshore USDT for global markets, onshore USAT for the regulated one.
• PayPal's PYUSD. Launched in 2023, PYUSD broke out after GENIUS passed: circulation ran near $3.5 billion by May 2026 — more than five times its level a year earlier — with access extended to roughly 70 markets, per American Banker. It is the first stablecoin with a built-in consumer distribution network (PayPal and Venmo).
• The bank consortium. Major US banks including JPMorgan Chase, Bank of America, Citigroup, and Wells Fargo have been in discussions to launch a shared stablecoin, per the same American Banker reporting — the clearest sign that incumbent finance sees regulated stablecoins as payments infrastructure it cannot ignore.
• Circle, the public company. Circle went public in 2025, making USDC's economics — reserve interest income against distribution costs — visible in quarterly filings and cementing its position as the institutional default, with MiCA authorization in the EU and monthly Deloitte attestations, per Insights4VC's analysis.

For holders, more regulated issuers mean more redemption guarantees and disclosure — and more entrenchment of the freeze-and-comply model. The censorship-resistance end of the spectrum is now served almost entirely by the decentralized and synthetic families.

The mechanics of a depeg: how pegs break and heal

The case studies above share an anatomy worth making explicit. A stablecoin actually has two prices: the primary price (what the issuer or protocol redeems at — $1, by construction) and the secondary price (whatever DEX pools and exchanges say right now). The peg is nothing more than arbitrageurs keeping the two aligned: if the token trades at $0.99, they buy it and redeem at $1.00; if it trades at $1.01, they mint at $1.00 and sell. Every depeg is a story about that loop breaking somewhere:

• Redemption is closed or slow. USDC's March 2023 depeg happened over a weekend precisely because bank-wire redemption was shut until Monday — all price discovery was forced into secondary markets. Coins with narrow redemption windows or approved-participant-only redemption (like USDe) lean harder on secondary liquidity.
• The $1 claim itself is doubted. If backing is impaired — SVB's exposure, Stream's $93 million hole — arbitrageurs won't buy at $0.95 to redeem for what might be $0.90 of assets. Price then discovers the market's estimate of the actual backing, which is why xUSD went to $0.10, not $0.98.
• Secondary liquidity is thinner than the exit. Even with sound backing, a rush of sellers into shallow pools moves the price exactly as AMM math dictates — the dynamic our liquidity guide computes step by step. Depeg severity is a function of exit volume divided by pool depth.

Recovery runs the film backwards: restore confidence in the $1 claim (the SVB deposit guarantee), reopen redemption, and arbitrage closes the gap — usually within days. When the claim cannot be restored (UST, xUSD, deUSD), there is no arbitrage to do, and the "depeg" is simply repricing to reality. This is the single most useful lens for judging a wobbling stablecoin in real time: ask whether redemption at full value is still credible, not whether the chart looks scary.

How to assess a stablecoin's risk

A practical checklist before holding size in any stablecoin:

1. Identify the family. Fiat-backed, crypto-collateralized, synthetic, or algorithmic? This alone tells you the dominant failure mode.
2. Read the backing evidence. Attestations (who, how often, what standard?) for centralized coins; on-chain collateral dashboards for decentralized ones. An attestation is a point-in-time snapshot, not a full audit.
3. Map the redemption path. Who can redeem at par, how fast, and with what minimums? If you can't redeem, you depend entirely on secondary-market liquidity.
4. Check peg history. How did it trade in March 2020, May 2022, November 2022 (FTX), March 2023, November 2025 (Stream/Elixir)? Survived stress is the best evidence available.
5. Check freeze and upgrade powers. Can the issuer blacklist you? Can governance change the contract under you?
6. Measure real liquidity. Depth on major DEX pools and exchanges determines the exit price in a panic — the mechanics in our liquidity and slippage guide apply directly.
7. Question the yield. If a "stable" asset pays double-digit yield, the risk is priced in somewhere. UST's 19.45% was the warning, not the reward.
8. Watch the flows. Large redemptions and whale movements often precede visible stress — our guide on tracking Ethereum whales covers the tooling.
9. Check regulatory status. Is the issuer licensed under the GENIUS Act framework or MiCA, or does it operate outside both? Licensing does not eliminate risk — SVB proved that — but it defines who guarantees redemption, what backs the token, and which disclosures you can actually read.

Related guides

• Ethereum DeFi Explained — the lending and DEX infrastructure stablecoins power.
• Token Liquidity & Slippage Explained — why depeg severity is a function of pool depth.
• How to Read DeFi TVL — interpreting the metrics stablecoin protocols report.
• How to Track Ethereum Whales — spotting large stablecoin flows before they move markets.

Frequently Asked Questions

Sources & further reading

• Circle — USDC Transparency & Reserves
• Tether — 2025 Attestation Reports announcement
• Ethena docs — USDe Overview
• Blockworks — Maker rebrands as Sky, introduces USDS (August 2024)
• MIT Sloan — Anatomy of a Run: The Terra Luna Crash
• CNBC — USDC breaks dollar peg after $3.3B SVB exposure revealed (March 2023)
• Covington — The GENIUS Act Becomes Law (July 2025)
• The Block — BlackRock's BUIDL crosses $1 billion (March 2025)
• CoinLaw — Stablecoin Market Cap Statistics 2026
• CoinLaw — Stablecoin Market Share by Chain (2026)
• DailyCoin — Sky USDS Overtakes Ethena USDe in Stablecoin Ranking
• Bessemer Venture Partners — Stablecoins: From DeFi Primitive to Global Financial Infrastructure
• Morgan Lewis — GENIUS Act Implementation: Key Proposals (April 2026)
• US Treasury — Proposed Rule to Implement the GENIUS Act's Illicit-Finance Requirements (April 2026)
• OCC — GENIUS Act Regulations: Notice of Proposed Rulemaking (2026)
• Kavout — How the Stablecoin Landscape Is Shifting in 2026 (USAT launch)
• American Banker — PayPal Boosts Its Stablecoin as Banks Prepare a Shared One (July 2026)
• Insights4VC — Inside Circle's Stablecoin Economics
• KuCoin — 2025's Defining Crisis: Stream Finance Collapse and Elixir Depeg
• Yahoo Finance — Elixir Shuts Down deUSD After Stream Finance's $93M Loss (November 2025)
• MetaMask — Real-World Asset Tokens in 2026 (rwa.xyz data)

Disclaimer: Stablecoins carry risks of de-pegging, regulatory action, and smart contract failure. This information is for educational purposes and is not financial advice.