Why Fiat-Backed Won: Algorithmic Stablecoins Hold Less Than 1% of Supply

The stablecoin market has delivered a clear verdict. Of the roughly $300 billion in stablecoin supply circulating in September 2026, fiat-backed designs account for approximately 93% of the total. Pure algorithmic stablecoins hold less than $500 million: a fraction so small it barely registers on market composition charts. The experiment in maintaining a dollar peg through code alone, without reserves, has been tried repeatedly and has failed almost every time.

This article traces how the market arrived at this point, what killed the algorithmic thesis, and why regulators, institutions, and builders have converged on fiat-backed stablecoins as the default model.

The Stablecoin Market in 2026

The numbers tell a stark story. Tether (USDT) commands over $183 billion in circulation, roughly 60% of the entire stablecoin market. Circle (USDC) holds approximately $74 billion, another 24%. Together, the top two fiat-backed issuers account for 84% of all stablecoins. The top five issuers, all fiat-backed, hold nearly 89%.

The remaining supply is mostly crypto-overcollateralized stablecoins like DAI/USDS from Sky (formerly MakerDAO) at roughly $13 billion, plus smaller protocols like GHO and crvUSD. These are not algorithmic: they require depositing more collateral than the stablecoins minted, which is the opposite of the algorithmic promise.

Key distinction: Crypto-overcollateralized stablecoins like DAI are sometimes grouped with algorithmic designs, but they function differently. DAI requires depositing $150+ of ETH to mint $100 of stablecoins. Algorithmic stablecoins attempt to maintain a peg with no backing or fractional backing, relying on arbitrage incentives and token mechanics instead of reserves.

How Algorithmic Stablecoins Were Supposed to Work

The algorithmic stablecoin thesis was elegant on paper. Instead of locking up real dollars in a bank account, a protocol would use economic incentives and smart contract mechanics to keep the token trading at $1. No banks, no counterparty risk, no reserves to audit. Pure DeFi money.

Several mechanisms were tried across the 2020 to 2022 era of algorithmic experimentation.

Seigniorage Models

Protocols like Basis Cash and Empty Set Dollar used a multi-token system inspired by central bank operations. When the stablecoin traded above $1, new tokens were minted and distributed to holders (expanding supply to push the price down). When it traded below $1, holders could purchase discounted "bond" tokens redeemable for stablecoins when the peg recovered. The system assumed that rational arbitrageurs would always step in to restore the peg. In practice, once confidence broke, no one wanted to buy bonds for a currency that might never return to $1.

Rebase Mechanisms

Tokens like Ampleforth automatically adjusted every holder's balance to target a price. If the token traded above $1, everyone's wallet balance increased. Below $1, balances decreased. This created a unit-of-account problem: your balance changed constantly, making the token impractical for payments, lending, or any form of commerce.

Mint-and-Burn with a Companion Token

This was the model that grew largest, exemplified by UST/LUNA. Users could always mint $1 of UST by burning $1 of LUNA, or redeem $1 of UST for $1 of LUNA. In theory, arbitrageurs would maintain the peg by buying underpriced UST and redeeming it for LUNA at a profit. In practice, this created a reflexive death spiral under stress.

The UST Collapse: The Inflection Point

On May 7, 2022, a large swap of approximately 85 million UST for USDC on Curve Finance triggered the beginning of the end for the largest algorithmic stablecoin ever created. UST had reached a peak market capitalization of $18.77 billion, making it the third-largest stablecoin behind USDT and USDC. Its companion token LUNA had peaked at $116 per token with a market cap exceeding $40 billion.

The depeg unfolded over approximately one week. As UST fell below $1, holders rushed to redeem for LUNA. Each redemption minted new LUNA, crashing its price. As LUNA's price crashed, the amount of LUNA needed to back each dollar of UST increased, creating a hyperinflationary spiral. LUNA went from $116 to $0.00008. UST went from $1 to pennies. Between $40 billion and $60 billion in combined value was destroyed.

Why this mattered beyond Terra: The UST collapse did not just kill one stablecoin. It triggered contagion across the crypto market, contributed to the bankruptcy of Three Arrows Capital, Celsius, and Voyager, and prompted regulatory action that would shape stablecoin policy for years. The Federal Reserve Bank of Richmond published a detailed post-mortem on the collapse, calling it a textbook example of a bank run applied to algorithmic money.

The Graveyard: Other Algorithmic Failures

UST was the largest failure, but it was far from the only one. A pattern emerged across multiple protocols: algorithmic stablecoins work during calm markets when arbitrage is profitable, then collapse suddenly when stress arrives.

The Federal Reserve published research on the IRON/TITAN collapse, describing it as "the world's first large-scale crypto bank run." The pattern they identified applies to every algorithmic design: when confidence breaks, the mechanism that is supposed to restore the peg instead accelerates the collapse.

Why Survivors Abandoned Algorithmic Design

The most telling evidence that algorithmic stablecoins failed is not the projects that died: it is the projects that survived by abandoning the model entirely.

Frax: From Fractional-Algorithmic to Fully Backed

Frax was the most prominent hybrid design: partially backed by collateral, partially by algorithmic mechanisms. At various points, its collateral ratio ranged from 85% to 100%, with the algorithmic portion absorbing the gap. In February 2023, three months after the FTX collapse and nine months after UST, the Frax community voted via governance proposal FIP-188 to move to a 100% collateral ratio, permanently abandoning the fractional-algorithmic model.

The original FRAX token has since shrunk to approximately $27 million in market cap. Frax Finance now focuses on frxUSD, a fully fiat-collateralized stablecoin with governance-approved custodians: a complete reversal of the original algorithmic thesis.

RAI: The Purist Experiment That Couldn't Scale

RAI from Reflexer Labs was perhaps the most intellectually honest algorithmic design. It was not pegged to the dollar but instead used a PID controller to maintain a floating redemption price, backed by ETH collateral. It explicitly avoided governance and aimed for pure algorithmic monetary policy. In November 2024, Reflexer announced the protocol would enter Global Settlement (an orderly shutdown) at the end of Q1 2025. RAI's market cap had shrunk to approximately $1.5 million with only about 527,000 RAI in circulation. Even the purest algorithmic approach could not attract meaningful adoption.

Why Fiat-Backed Stablecoins Won

The dominance of fiat-backed stablecoins was not a foregone conclusion. Early crypto philosophy favored trustless, decentralized alternatives. But several forces pushed the market toward reserve-backed designs.

Stress Resistance

Fiat-backed stablecoins have survived every major crypto crisis. USDT and USDC maintained their pegs through the 2022 bear market, the FTX collapse, and the 2023 banking crisis (USDC briefly depegged to $0.88 during the Silicon Valley Bank failure but recovered within 72 hours when the FDIC backstopped deposits). The reason is straightforward: if $1 of reserves exists for every token, the peg has a floor. Algorithmic stablecoins have no floor: once the mechanism breaks, the token can go to zero.

Institutional Requirements

As stablecoins moved from DeFi trading pairs to enterprise treasury management, cross-border payments, and payment rails, institutional buyers demanded clarity on what backs each token. A proof of reserves showing U.S. Treasury bills in a segregated account satisfies compliance teams in a way that "algorithmic mechanisms maintain the peg" never will.

Simplicity of the Trust Model

Fiat-backed stablecoins have a simple promise: one token equals one dollar in a bank account. The risks are knowable (issuer risk, custodian risk, regulatory risk) and manageable through standard financial controls like attestations, audits, and insurance. Algorithmic stablecoins require trusting game theory, smart contract logic, oracle accuracy, and the assumption that rational arbitrageurs will always show up: a complex and fragile chain of assumptions.

Regulation Sealed the Outcome

Even if algorithmic designs could solve their technical problems, regulation has now effectively closed the door in the two largest stablecoin markets.

The GENIUS Act (United States)

The GENIUS Act, signed into law on July 18, 2025, is the first U.S. federal law governing payment stablecoins. It explicitly requires 1:1 reserve backing in specified asset types: U.S. coins and currency, demand deposits at insured depository institutions, and U.S. Treasury bills with a remaining maturity of 93 days or less. Algorithmic mechanisms, crypto collateral, and overcollateralization do not qualify. There is no path for a pure algorithmic stablecoin to achieve legal status as a payment stablecoin in the United States.

MiCA (European Union)

The Markets in Crypto-Assets Regulation took effect for stablecoins on June 30, 2024, with full enforcement by December 30, 2024. MiCA requires that e-money token issuers maintain full reserve backing with liquid assets at a 1:1 ratio, publish regular transparency reports, and undergo mandatory audits. Algorithmic designs cannot meet these requirements. Major EU-regulated exchanges delisted non-authorized stablecoins throughout 2024 and 2025.

For a deeper analysis of global stablecoin regulation, see our stablecoin regulation tracker and GENIUS Act explainer.

Do Algorithmic Designs Have Any Future?

A small number of projects continue to experiment at the margins, but their trajectory points toward irrelevance for mainstream stablecoin use.

Crypto-Overcollateralized Stablecoins Survive, But They Are Not Algorithmic

DAI/USDS (Sky, formerly MakerDAO) at approximately $13 billion is the largest non-fiat-backed stablecoin. But calling it algorithmic is misleading. Users must deposit $150 or more of ETH, staked ETH, or other collateral to mint $100 of DAI. If the collateral value drops, positions are liquidated to protect the peg. This is closer to a secured loan than an algorithmic mechanism. Furthermore, roughly 40% of DAI/USDS collateral is now real-world assets (Treasury bills via Star allocators) and 35% is USDC via the Peg Stability Module: effectively making it partially fiat-backed.

Niche Experiments Remain Small

Protocols like LUSD (Liquity, ~$27 million market cap) and GHO (Aave, ~$584 million) continue operating as overcollateralized designs within DeFi. They serve specific niches: censorship-resistant borrowing, capital-efficient lending, or governance utility within their parent protocols. None has shown a path to challenging fiat-backed stablecoins for mainstream adoption.

Delta-Neutral Strategies: A Different Approach

A newer category of stablecoins like USDe (Ethena) uses delta-neutral hedging strategies: holding crypto collateral while shorting equivalent perpetual futures to maintain dollar value. At roughly $6 billion, this category has grown faster than any algorithmic design, but it carries its own risks including negative funding rates, counterparty exposure to centralized exchanges, and liquidation cascades in extreme volatility. These are not algorithmic in the traditional sense: they use real (though synthetic) economic backing rather than pure token mechanics.

What This Means for Stablecoin Users

The convergence on fiat-backed stablecoins has practical implications for anyone building or using stablecoin infrastructure.

• Reserve transparency matters more than mechanism design. Monthly attestations and segregated custody have become table stakes.
• Regulatory compliance is now a prerequisite for scale. Stablecoins that cannot meet GENIUS Act or MiCA requirements face delisting from major exchanges and exclusion from institutional flows.
• The risk profile has shifted from "will the peg hold" to "who is the issuer and custodian." For fiat-backed stablecoins, counterparty risk assessment replaces mechanism analysis.
• Yield-bearing stablecoins are the next frontier, but the yield comes from reserves (Treasury bill interest, lending revenue), not from algorithmic inflation or token emissions.

The USDB stablecoin on Spark reflects this market consensus. Issued by Brale, a FinCEN-registered Money Services Business with multi-state money transmitter licenses, USDB is backed 1:1 by U.S. Treasury bills and cash equivalents held in segregated, bankruptcy-remote accounts. Monthly audits by the independent accounting firm Abdo verify reserve adequacy. This is the model the market has chosen: verifiable reserves, regulated issuers, and transparent attestation rather than algorithmic faith.

For users who want to hold stablecoins on Bitcoin, General Bread provides a Spark-powered wallet where USDB is available with yield paid daily in BTC. For developers building stablecoin applications, the Spark SDK documentation covers integration with USDB and other Spark-native tokens. You can also explore how different peg mechanisms compare to understand the technical details behind reserve-backed stability.

This article is for educational purposes only. It does not constitute financial or investment advice. Bitcoin and Layer 2 protocols involve technical and financial risk. Always do your own research and understand the tradeoffs before using any protocol.