Algorithmic stablecoins are the category that the market declared dead in 2022 and then quietly rebuilt from scratch. After the collapse of UST erased $40 billion in a week, survivors stopped calling themselves "algorithmic" for a while. Then the next generation — Frax v2, crvUSD, Ethena's USDe, and a reshaped Sky ecosystem around DAI — rebuilt the category around over-collateralization, dynamic liquidation, and basis-trade backing rather than pure seigniorage games. In 2026, algorithmic stablecoins are roughly 6-8% of the stablecoin market by supply but punch above that weight in DeFi integration and yield primitives.

This guide walks through the four algorithmic (and algorithm-adjacent) stablecoins that matter in 2026, how their peg mechanisms actually work, and where each fits in treasury and protocol design.

The term "algorithmic stablecoin" is fuzzier now than it was pre-UST. In 2021, it usually meant a pure seigniorage design — Terra-style, where peg was maintained by minting and burning a volatile sister token with no hard collateral underneath. That model is gone. What remains, and what the industry now groups under the algorithmic umbrella, is a broader category of stablecoins that use protocol-level mechanisms rather than traditional fiat reserves to maintain peg.

The surviving designs fall into three sub-categories:

All four designs share one property: peg is maintained by onchain mechanisms, not by a trust company holding Treasuries. That's the dividing line between "algorithmic" and "fiat-backed" in 2026 usage.

DAI is the original crypto-collateralized stablecoin, launched by MakerDAO in 2017 and now operating under the Sky ecosystem rebrand that introduced USDS as a parallel token. DAI's market cap sits around $5-6 billion in April 2026, with USDS adding another $3-4 billion when counted together.

The mechanism is the classic CDP model. Users deposit collateral — ETH, wBTC, USDC, real-world assets tokenized through MakerBurn's RWA dashboards — into vaults and mint DAI against it, typically at 150% or higher collateralization ratios depending on the asset. If the collateral value drops below a liquidation threshold, a keeper liquidates the vault, repays the DAI debt, and pockets a small penalty. The system remains solvent as long as liquidations can keep up with price drops, which is exactly where the March 2020 crisis broke and where subsequent Maker governance has spent years adding circuit breakers.

DAI's reserve composition has drifted sharply toward USDC and tokenized real-world assets over the last three years — which some critics argue makes DAI effectively a wrapper on Circle's reserves plus RWA exposure. Sky's response has been the USDS rebrand and the introduction of new collateral types designed to reduce USDC concentration. The MakerDAO and DAI enterprise explainer breaks this down in detail for treasury teams weighing CDP exposure.

DAI remains one of the most widely integrated stablecoins in DeFi — deep liquidity on Curve, Uniswap, and Balancer; accepted as collateral on Aave, Compound, and Morpho; and used as a base pair for many RWA protocols. For swap routing across DAI and its peers, the intent-based DEX alternatives comparison covers the venue landscape. The Maker vaults documentation is the canonical reference for anyone integrating DAI mint/burn mechanics directly.

Frax is the stablecoin that explicitly embraces the hybrid label. FRAX v1 launched in 2020 as a fractional-algorithmic stablecoin — partly backed by USDC, partly stabilized by the volatile FXS token, with the collateralization ratio adjusting based on market demand. After the post-UST market resetting, Frax v2 pivoted to a fully collateralized design, currently sitting near 100% collateralization with a mix of cash-equivalent assets and crypto collateral.

Frax v2 runs three integrated products: FRAX (the stablecoin), frxETH (a liquid staking derivative), and Fraxlend (the lending market that consumes FRAX as base liquidity). The design intent is a closed-loop stablecoin-native DeFi stack, and in 2026 that stack sits at roughly $600M in FRAX circulating supply — meaningfully smaller than DAI but with deeper protocol-level integration than most stablecoins at its size.

The Frax protocol documentation describes the Algorithmic Market Operations (AMO) framework that Frax uses to deploy reserves into DeFi while maintaining peg — a design that inspired several other stablecoin protocols and that Sky is now borrowing aspects of.

crvUSD is Curve Finance's native stablecoin, launched in May 2023, with a market cap around $400-500M in April 2026. What makes crvUSD notable isn't the CDP model itself — which is similar to DAI at a high level — but the liquidation mechanism: a Lending-Liquidating AMM Algorithm (LLAMMA) that continuously rebalances collateral into stablecoin as the collateral price drops, rather than waiting for a hard liquidation threshold.

The effect is a soft-liquidation zone instead of a liquidation cliff. If you deposit ETH and mint crvUSD, and ETH starts falling, LLAMMA gradually converts your ETH collateral into crvUSD across a price band. If ETH recovers, LLAMMA converts back. You pay a fee for this continuous rebalancing, but you're far less likely to be fully liquidated by a fast wick. That's particularly useful during volatile sessions and has made crvUSD popular with professional DeFi users who hold ETH and wBTC positions and want leverage without the full liquidation risk.

crvUSD is a smaller asset by supply but one of the most technically interesting designs in the category. For protocol teams thinking about collateral mechanics, the LLAMMA paper is worth reading.

USDe is the most prominent new entrant in the algorithmic stablecoin category since UST collapsed. Launched by Ethena Labs in early 2024, USDe reached $6 billion in circulating supply by April 2026, making it the largest non-fiat-backed stablecoin and the third-largest stablecoin overall after USDT and USDC.

USDe isn't collateralized by crypto in the CDP sense and isn't backed by Treasuries. Instead, USDe is backed by a delta-neutral basis trade: for every dollar of USDe minted, Ethena holds an equivalent long spot position in ETH (and increasingly BTC and liquid staking tokens) and an offsetting short perpetual futures position. The long and short cancel each other's price exposure, leaving the funding rate on the short perp as the net revenue stream — which is how Ethena's staked USDe (sUSDe) generates its yield.

The design is elegant and the yield has been strong — often 15-25% on sUSDe during bull-market funding regimes. The risk profile is unusual: USDe is exposed to exchange counterparty risk (Binance, Bybit, Deribit where the shorts are held), funding rate inversion (when perp funding goes negative, the strategy loses money), and basis dislocations. Ethena's transparency dashboards disclose exchange exposure, collateral composition, and yield attribution in near-real-time.

USDe is algorithmic in a different sense than DAI or Frax — it's not over-collateralized by crypto, it's hedged by synthetic shorts. The market has so far treated it as a credible dollar stablecoin, and it has become one of the most integrated assets on Ethereum DeFi in a remarkably short time. The best programmable stablecoin protocols guide covers where synthetic dollars fit in the broader stablecoin stack.

A side-by-side on how each asset maintains its dollar peg:

Every design trades something. DAI trades decentralization for USDC/RWA exposure. Frax v2 trades the pure fractional-algorithmic model for full collateralization. crvUSD trades hard liquidation for continuous rebalancing fees. USDe trades reserve simplicity for exchange dependency and funding-rate exposure. None of them are risk-free, and anyone holding meaningful size in any of them needs to understand the specific risk shape. For teams building onchain credit or treasury products against any of these, the execution-time compliance framing is worth reading alongside the risk analysis.

It would be incomplete to discuss algorithmic stablecoins in 2026 without acknowledging the category's history. The Terra/UST collapse in May 2022 is the reference event — an algorithmic stablecoin design that worked for almost two years and then unwound in a single week, wiping out tens of billions in user deposits and destroying confidence in the entire category for the rest of that year.

The surviving 2026 designs share a common lesson: they don't rely on reflexive mint/burn of a volatile sister token to maintain peg in a crisis. DAI, Frax, and crvUSD all have hard collateral underneath. USDe has a matched hedge rather than a reflexive loop. That's not a guarantee — new failure modes will emerge — but it's a materially different risk profile than UST's was.

The Coin Center post-UST analysis is worth reading for the policy framing, and for anyone operating stablecoin treasury at scale, the practical question is less about whether algorithmic stablecoins should exist and more about how to weight them in a reserve mix alongside fiat-backed alternatives.

At the Rail layer, algorithmic stablecoins move across chains using the same infrastructure as their fiat-backed peers: Hyperlane, LayerZero, Wormhole, and CCTP-adjacent paths where the issuer supports burn-and-mint. DAI is deployed on Ethereum, Optimism, Arbitrum, Base, Polygon, and a dozen other chains. USDe is expanding multichain rapidly.

At the Layer, orchestration platforms route between these rails based on cost, speed, and finality. Algorithmic stablecoins present an interesting wrinkle here: their onchain mint/burn mechanisms mean that in some cases, the cheapest path is to burn on one chain and mint fresh on another rather than to bridge liquidity. Eco Routes selects among these options at execution time, which matters for treasury teams moving between DAI and crvUSD across chains at size. The multi-source liquidity routing explainer covers why this beats sticking to any single pool.

At the App layer, algorithmic stablecoins dominate DeFi integration even when they trail fiat-backed stablecoins on raw market cap. Most structured yield products, most lending markets' interest-rate curves, and most collateral systems either use or are denominated against DAI, FRAX, USDe, or crvUSD. For developers building on this infrastructure, the stablecoin tools for developers comparison is the practical starting point.

A short decision framework:

For enterprise treasury teams, most of these assets are better held as a small portion of a broader stablecoin mix dominated by USDC and USDT, rather than as primary reserves. The top cross-chain liquidity protocols overview covers how orchestration layers handle that kind of multi-asset, multi-chain treasury in practice.

The category has rebuilt around over-collateralization, hard liquidation mechanisms, or matched hedge positions — materially different designs than UST's reflexive mint/burn loop. DAI, Frax v2, crvUSD, and USDe have all been live through multiple volatility cycles without peg failure. That's meaningful evidence, but not a guarantee. Each carries distinct risks that matter for sizing.

USDe from Ethena is the largest by circulating supply at approximately $6 billion, having surpassed DAI in 2024. Counting DAI and USDS together under the Sky ecosystem puts that combined supply near $9 billion. Frax and crvUSD are meaningfully smaller at $600M and $400-500M respectively but have strong integration inside their native DeFi ecosystems.

Yes. sUSDe (staked USDe) generates yield from funding rates on Ethena's short perp positions, typically 10-25% annualized depending on market conditions. DAI has the DAI Savings Rate (DSR) set by Sky governance, sitting around 4-7% in early 2026. Frax stakes into Fraxlend markets. crvUSD users earn trading fees from LLAMMA rebalancing.

Each uses a different mechanism. DAI and crvUSD use over-collateralized CDP vaults with liquidation mechanisms — if peg drifts, arbitrageurs mint or burn to close the gap. Frax v2 uses a hybrid of direct collateralization and AMO operations. USDe maintains peg through the delta-neutral basis trade, where the short perp hedge cancels the long spot position's price exposure.

Yes. DAI and USDe are deployed on most major EVM chains with canonical bridge paths. Frax and crvUSD are available on a narrower set. Cross-chain movement typically goes through Hyperlane, LayerZero, Wormhole, or issuer-specific burn-and-mint. Orchestration platforms like Eco Routes select the best path at execution time based on cost, speed, and finality.

Algorithmic stablecoins in 2026 are a narrower, more carefully engineered category than they were in 2021. DAI has earned its status as the benchmark over a decade of iteration. Frax v2 runs the cleanest closed-loop stablecoin DeFi stack. crvUSD brings a genuinely novel liquidation mechanism. USDe has rebuilt the synthetic dollar category around matched hedges rather than reflexive seigniorage. Together they represent roughly 6-8% of total stablecoin supply but a much larger share of DeFi integration, onchain yield primitives, and protocol-native collateral. For builders designing treasury systems or protocol mechanics in 2026, understanding this category isn't optional — it's where most of the interesting stablecoin infrastructure work is happening. The stablecoin workflow engines overview covers the execution-layer tooling that makes these assets composable at scale.