Algorithmic stablecoins are cryptocurrency tokens that try to hold a steady price, usually $1.00, without keeping cash or Treasury bonds in reserve for every token. Instead, software rules expand or shrink the token supply as market demand shifts. The approach is ambitious: replace human central bankers with code. It is also fragile. When confidence cracks, the same feedback loops that stabilize the price can reverse violently, as the roughly $40 billion TerraUSD collapse in May 2022 showed.1Federal Reserve Bank of New York. The Financial Stability Implications of Digital Assets
How the Peg Is Supposed to Work
Every algorithmic stablecoin borrows the same idea from monetary economics: if you control how many tokens exist, you can influence the price. When demand pushes the price above $1.00, the protocol creates new tokens to bring the price back down. When the price slips below $1.00, the protocol removes tokens from circulation to make the remaining ones scarcer. The software watches the market price continuously and runs these adjustments automatically.
The key difference from collateralized stablecoins like USDC or Tether is what sits behind each token. A fully collateralized stablecoin lets you redeem your token for actual dollars or short-term Treasury bills held in reserve. An algorithmic stablecoin offers you economic incentives instead: arbitrage opportunities, bonus tokens, or governance rights designed to make rational traders push the price back toward the peg on their own. The system works as long as enough people believe it will keep working. That is both its efficiency and its fundamental vulnerability.
The Main Designs You’ll Encounter
Rebase Tokens
Some protocols adjust every holder’s wallet balance at the same time. If the token trades at $1.05, the protocol mints additional tokens and distributes them proportionally across all wallets. If the price falls to $0.95, the protocol reduces the token count in every wallet to create scarcity. Your percentage of the total supply stays the same, but the raw number of tokens you hold changes constantly.
The experience for holders is disorienting. Watching your wallet balance shrink during a contraction feels like losing money, even though each remaining token is supposed to be worth more. In practice, contractions often set off panic selling, which drives the price down further and triggers another contraction. Protocols using this model, such as Ampleforth, have spent extended periods trading well below their target.
Seigniorage and Two-Token Systems
A more common architecture uses two tokens working together. The primary token is the stablecoin itself, and a secondary “share” or “governance” token absorbs the volatility. When the stablecoin drops below $1.00, users can burn stablecoins in exchange for newly minted share tokens, reducing stablecoin supply and theoretically lifting the stablecoin price. When the stablecoin rises above $1.00, the protocol mints new stablecoins and distributes them to share token holders as profit.
The design borrows from how governments earn seigniorage, the profit from issuing currency that costs less to produce than its face value. The share token is a claim on future seigniorage: if the stablecoin grows in adoption, share token holders collect the newly minted supply. The problem is that the share token’s value depends entirely on faith in the system’s future growth. Unlike a bond backed by tax revenue, a share token is backed by the expectation that strangers will keep buying the stablecoin tomorrow.
Fractional Models
A hybrid approach tries to split the difference. Fractional designs back a portion of each token with external reserves like USDC or Ethereum, while the remaining portion is stabilized algorithmically through a native utility token. At a 75% collateral ratio, seventy-five cents of each stablecoin is redeemable against real reserves and the remaining twenty-five cents is managed through minting and burning the utility token.
The collateral ratio typically adjusts with market conditions. During calm periods the protocol gradually lowers the ratio, relying more on the algorithm. During stress the ratio increases to reassure holders. This flexibility makes fractional models more capital-efficient than fully collateralized stablecoins, but the uncollateralized portion still carries the same death-spiral risk as pure seigniorage systems. When Frax, the most prominent fractional design, moved to a fully collateralized model in early 2023, it was an implicit acknowledgment that the market would not trust the algorithmic portion during stress.
The Death Spiral and the Terra Collapse
The catastrophic failure mode is self-reinforcing. When the stablecoin loses its peg, the protocol mints large quantities of the share token to absorb selling pressure. Flooding the market with share tokens tanks their price, which destroys confidence in the system’s ability to defend the peg, which triggers more stablecoin selling, which forces even more share token minting. Each loop makes both tokens worth less.
TerraUSD (UST) and its companion token LUNA showed this at scale in May 2022. UST had roughly $18 billion in market capitalization when it lost its dollar peg. As holders rushed to exit, the protocol minted trillions of LUNA tokens trying to absorb the supply, hyperinflating LUNA from roughly $80 to fractions of a cent within days. The collapse erased approximately $40 billion in combined value and cascaded through the broader crypto ecosystem, contributing to the failure of hedge fund Three Arrows Capital and the bankruptcies of lending platforms Voyager and Genesis.1Federal Reserve Bank of New York. The Financial Stability Implications of Digital Assets The SEC later charged Terraform Labs and its CEO Do Kwon with violating securities registration and anti-fraud provisions, alleging they marketed UST as a yield-bearing instrument paying up to 20% interest. The resulting enforcement action produced a settlement of roughly $4.5 billion in disgorgement, interest, and civil penalties.2U.S. Securities and Exchange Commission. SEC Charges Terraform and CEO Do Kwon with Defrauding Investors in Crypto Schemes
Liquidity depth matters here. The arbitrage mechanism that is supposed to restore the peg only works if traders can execute large trades without moving the market against themselves. A $5 million sell order in a pool holding $50 million might push the price down a few cents; the same sell in a pool holding $5 million could crash the price by 20% or more. Fragmented liquidity across trading venues makes it harder for arbitrageurs to close the gap.
Flash Loan and Oracle Attacks
Every algorithmic stablecoin depends on price oracles, the external data feeds that tell the smart contract what the token is currently trading for. Smart contracts cannot observe markets on their own. They need oracles to aggregate price data from multiple exchanges, calculate a volume-weighted average, and deliver it on-chain. The contract then checks whether the price is above or below $1.00 and executes the supply adjustment.
That dependency on external data is an attack surface. Flash loans let anyone borrow enormous sums of cryptocurrency with no collateral, as long as the loan is repaid within the same blockchain transaction. The sequence is familiar: borrow millions, use the borrowed funds to execute massive trades that distort the price in a specific liquidity pool, trigger a vulnerable smart contract that reads the distorted price as real, profit from the contract’s miscalculation, then repay the loan. The whole attack happens in one transaction and costs little more than gas fees.
In one notable case, an attacker used a flash loan to manipulate a liquidity pool’s token balance, causing a burn function that relied on that pool’s price feed to credit the attacker double the legitimate amount of tokens. For algorithmic stablecoins, oracle manipulation can trigger artificial supply expansions or contractions that destabilize the peg while the attacker profits at other holders’ expense. The CFTC’s anti-manipulation rules prohibit using deceptive devices in connection with commodity transactions.3eCFR. 17 CFR Part 180 – Prohibition Against Manipulation
Where U.S. Law Stands
Algorithmic stablecoins sit in a regulatory gray zone, and several federal agencies are actively trying to clarify it. Whether a particular token falls under securities law, commodities law, or banking regulation depends on how it is structured, marketed, and used.
The SEC evaluates digital assets under the Securities Act of 1933 to decide if they qualify as investment contracts. If a token meets the Howey test, meaning buyers invest money in a common enterprise expecting profits primarily from the efforts of others, the issuer must register the offering or qualify for an exemption.4U.S. Securities and Exchange Commission. Framework for Investment Contract Analysis of Digital Assets Pure payment stablecoins may not meet this standard, but share tokens and governance tokens associated with algorithmic systems frequently do, especially when marketed as profit-generating.
The CFTC has taken the position that certain digital assets qualify as commodities, giving it jurisdiction over fraud and manipulation in spot markets and authority over derivatives and leveraged trading. A protocol that facilitates leveraged transactions can trigger registration requirements under the Commodity Exchange Act.5U.S. Securities and Exchange Commission. Application of the Federal Securities Laws to Certain Types of Crypto Assets and Certain Transactions Involving Crypto Assets FinCEN separately requires entities that transfer funds on behalf of the public to register as Money Services Businesses, which brings Bank Secrecy Act obligations including an anti-money-laundering program, currency transaction reports, and detailed fund-transmittal records.6Financial Crimes Enforcement Network. BSA Requirements for MSBs Operating as an unlicensed money transmitter is a federal crime carrying up to five years in prison.7Office of the Law Revision Counsel. 18 USC 1960 – Prohibition of Unlicensed Money Transmitting Businesses
What the GENIUS Act Changes
The most significant development is the GENIUS Act (Guiding and Establishing National Innovation for U.S. Stablecoins), which creates a federal licensing framework for stablecoin issuers. Permitted issuers must maintain reserves backing every outstanding stablecoin on at least a one-to-one basis, using only high-quality assets: U.S. currency, demand deposits at insured banks, Treasury bills with 93 days or less remaining maturity, certain repurchase agreements, and qualifying money market funds.8U.S. Congress. S.1582 – GENIUS Act
For algorithmic stablecoins, the implications are stark. A purely algorithmic design that holds no external reserves cannot satisfy a one-to-one reserve requirement. The Act does not ban algorithmic stablecoins by name, but it effectively excludes them from the “permitted payment stablecoin issuer” category. Stablecoins issued outside that framework cannot be treated as cash or cash equivalents for accounting purposes, used as margin or collateral by regulated financial firms, or accepted as settlement assets by banking infrastructure.8U.S. Congress. S.1582 – GENIUS Act The Act directs the Treasury Secretary to study “endogenously collateralized” stablecoins, the technical term for tokens backed primarily by their own ecosystem’s tokens rather than external assets. Anyone issuing a dollar-denominated stablecoin without approval faces civil penalties of up to $100,000 per day.
What You Actually Get as a Holder
Algorithmic stablecoins fall outside most of the safety nets that protect holders of traditional financial products. Knowing what you don’t have is as important as knowing how the technology works.
No Deposit or Investment Insurance
Stablecoins are not bank deposits and are not covered by FDIC insurance. The GENIUS Act makes this explicit: payment stablecoins “shall not be backed by the full faith and credit of the United States” and issuers are prohibited from implying otherwise.9Federal Register. GENIUS Act Requirements and Standards for FDIC-Supervised Permitted Payment Stablecoin Issuers and Insured Depository Institutions The reserves an issuer holds at an insured bank are insured up to $250,000 as corporate deposits belonging to the issuer, not as pass-through coverage for individual stablecoin holders. SIPC, which protects brokerage accounts, does not cover digital assets unless they are securities registered with the SEC.10Securities Investor Protection Corporation. What SIPC Protects
Bankruptcy Risk
If an issuer or the platform where you hold tokens goes bankrupt, your recovery depends on the legal structure of your relationship with that entity. If the arrangement was custodial and tokens were held on your behalf in a segregated manner, those tokens may be treated as your property rather than part of the bankruptcy estate. If the arrangement was a debtor-creditor relationship, common when platforms commingle user funds, you become a general unsecured creditor behind secured lenders and priority claims. Even in favorable custodial arrangements, identifying and recovering digital assets from bankruptcy is slow and uncertain, especially when assets sat in omnibus wallets mixing multiple customers’ holdings.
Tax Tracking
The IRS treats cryptocurrency as property, so virtually any disposition or change in holdings can trigger a taxable event. For rebase tokens, this is a headache: every time the protocol adjusts your wallet balance, you may owe taxes on the difference. The IRS has not issued guidance specifically addressing rebases, but under its general framework, receiving additional tokens through an automatic supply expansion looks like taxable income, and a supply contraction that removes tokens could be treated as a disposal.
Short-term capital gains on crypto held less than a year are taxed as ordinary income. For 2026, federal income tax rates run from 10% to 37% depending on filing status and income.11Internal Revenue Service. IRS Releases Tax Inflation Adjustments for Tax Year 2026 Gains on tokens burned in a seigniorage system or redeemed through a fractional protocol are also reportable. Underreport, and the IRS can impose an accuracy-related penalty of 20% on the underpaid amount.12Internal Revenue Service. Accuracy-Related Penalty A rebase protocol that adjusts balances daily can generate hundreds of potential taxable events a year.