Algorithmic Stablecoin

What Is an Algorithmic Stablecoin?

An algorithmic stablecoin is a cryptocurrency designed to maintain a stable price (usually pegged to $1 USD) through software-driven supply adjustments rather than by holding equivalent reserves of fiat currency, commodities, or other crypto assets. Instead of each token being redeemable for a dollar in a bank account, the protocol itself expands or contracts the token supply to push the market price toward the target peg.

The appeal is straightforward: if you can build a stablecoin without reserves, you eliminate custodial risk, regulatory burden, and capital inefficiency. The protocol becomes self-sustaining, requiring no trusted issuer, no bank relationships, and no audits of reserve holdings. This vision of a decentralized, scalable stable currency has motivated dozens of projects since 2017.

The reality has been far less promising. Pure algorithmic stablecoins have consistently failed, with the most dramatic collapse being Terra/UST in May 2022, which destroyed roughly $40 billion in value. Understanding why these systems fail is essential for evaluating stablecoin risk more broadly, including the newer generation of regulated stablecoin frameworks that have emerged in response.

How It Works

Algorithmic stablecoins rely on a concept borrowed from central banking: seigniorage. In traditional economics, seigniorage is the profit a government earns by issuing currency at a cost lower than its face value. Algorithmic stablecoins apply this principle through smart contracts that manage token supply automatically.

The Seigniorage Model

The core mechanism operates in two modes depending on whether the stablecoin is trading above or below its target peg:

• Expansion (price above peg): when demand pushes the stablecoin above $1, the protocol mints new tokens and sells them on the open market. This increases supply, which should bring the price back down to $1. The newly created value (seigniorage) is distributed to protocol participants as a reward.
• Contraction (price below peg): when selling pressure pushes the stablecoin below $1, the protocol needs to reduce supply. This is typically accomplished through bond mechanisms, token burns, or incentivized buybacks that remove tokens from circulation, theoretically restoring the peg.

The asymmetry between these two modes is the fundamental vulnerability. Expansion is easy: minting new tokens costs nothing and generates profit. Contraction is hard: convincing holders to voluntarily lock up or burn tokens during a crisis requires offering future rewards, which only have value if confidence in the system persists.

Dual-Token and Multi-Token Models

Most algorithmic stablecoins use at least two tokens to separate the stable asset from the volatility-absorbing asset:

1. The stablecoin itself, which targets a $1 peg and is used for payments and savings
2. A governance or seigniorage token, which absorbs price volatility and captures the upside during expansion periods

In this model, the seigniorage token acts as a shock absorber. When the stablecoin needs to contract, the protocol issues seigniorage tokens or bonds that can be redeemed for stablecoins during the next expansion. Holders of the seigniorage token are essentially betting that the protocol will survive and expand again.

A simplified pseudocode representation of the core mechanism:

// Simplified algorithmic stablecoin rebalancing logic
function rebalance(currentPrice, targetPrice, totalSupply) {
  const deviation = currentPrice - targetPrice;
  const threshold = 0.01; // 1% band

  if (deviation > threshold) {
    // Price above peg: expand supply
    const mintAmount = totalSupply * (deviation / targetPrice);
    mintStablecoins(mintAmount);
    distributeSeigniorage(mintAmount);
  } else if (deviation < -threshold) {
    // Price below peg: contract supply
    const burnTarget = totalSupply * (-deviation / targetPrice);
    issueBonds(burnTarget, premiumRate);
    // Bonds redeemable 1:1 in next expansion
  }
}

Historical Examples and Failures

Terra/UST (2022)

Terra's UST was the largest algorithmic stablecoin, reaching a market cap of over $18 billion before its collapse in May 2022. The mechanism was conceptually simple: users could always swap 1 UST for $1 worth of LUNA (the seigniorage token), and vice versa. If UST traded below $1, arbitrageurs would buy cheap UST and redeem it for $1 of LUNA, burning UST and reducing supply. If UST traded above $1, they would mint new UST by burning LUNA.

The system appeared stable for over a year, partly sustained by Anchor Protocol, which offered ~20% yields on UST deposits. This artificial demand masked the underlying fragility. When large withdrawals from Anchor began in May 2022, UST lost its peg. As holders rushed to redeem UST for LUNA, the protocol minted enormous amounts of LUNA, crashing its price. This triggered a classic death spiral: falling LUNA prices meant each UST redemption produced more LUNA, diluting holders further, which drove more selling, which depegged UST further.

Within a week, UST fell from $1 to near zero, and LUNA went from $80 to fractions of a cent. The total value destroyed exceeded $40 billion, making it one of the largest financial collapses in cryptocurrency history.

Basis (Basecoin)

Basis, originally called Basecoin, raised $133 million in 2018 for a three-token algorithmic stablecoin: Basis (stable), Bond Tokens (contraction absorbers), and Share Tokens (expansion beneficiaries). During contraction, the protocol would auction Bond Tokens at a discount, redeemable for Basis during the next expansion. Share Tokens received seigniorage during expansion.

Basis never launched. The team returned funds to investors in December 2018 after concluding that its bond tokens would likely be classified as securities by the SEC, making the project legally unviable. The regulatory challenge highlighted a broader problem: algorithmic stablecoins that rely on speculative tokens to maintain stability face both technical and legal fragility.

Empty Set Dollar (ESD)

Empty Set Dollar launched in 2020 as a decentralized algorithmic stablecoin on Ethereum. It used a coupon system for contraction: when ESD traded below $1, holders could burn ESD for coupons redeemable at a premium during the next expansion period. Coupons expired after 30 days if no expansion occurred.

ESD experienced several depegging events. As confidence waned, fewer participants were willing to burn ESD for coupons with expiration risk. The coupon premium needed to attract buyers kept rising, making future expansions more expensive and less likely. ESD eventually stabilized well below its $1 target and became largely inactive.

The Death Spiral Problem

The consistent failure pattern across algorithmic stablecoins is the death spiral: a reflexive feedback loop where declining confidence accelerates the very collapse it fears. Understanding this mechanism is critical for evaluating any stablecoin that relies on market confidence rather than hard collateral.

The death spiral unfolds in predictable stages:

1. An external shock or large redemption pushes the stablecoin slightly below its peg
2. The contraction mechanism activates, requiring participants to buy bonds or burn tokens in exchange for future rewards
3. If confidence is shaken, fewer participants are willing to take the other side of contraction: why buy bonds in a system that might not survive?
4. Insufficient contraction means the peg is not restored, which further erodes confidence
5. Holders begin selling, increasing downward pressure on the price
6. In dual-token models, the seigniorage token also crashes as its value depends on future expansion that now seems unlikely
7. The system enters a reflexive loop: selling causes depegging, depegging causes more selling, until the stablecoin reaches near-zero value

This is structurally similar to a liquidation cascade in collateralized systems, but with a critical difference: collateralized systems have a floor (the value of the collateral), while purely algorithmic systems do not. When confidence evaporates, there is nothing backing the token, and its fundamental value is zero.

Hybrid Models: Partial Collateralization

The repeated failure of pure algorithmic stablecoins led to hybrid designs that combine partial collateral backing with algorithmic supply management. The idea is to provide a floor of real value while using algorithms to improve capital efficiency beyond 1:1 collateralization.

Frax was the most prominent hybrid model. It launched with a variable collateral ratio that started at 100% and could decrease as the system proved stable. At an 85% collateral ratio, each Frax token would be backed by $0.85 in USDC and $0.15 in algorithmic stabilization via the FXS governance token.

After the Terra collapse, Frax moved to 100% collateralization, effectively acknowledging that the market was unwilling to accept partially backed stablecoins. This trajectory illustrates a broader lesson: the capital efficiency gains from algorithmic stabilization are small compared to the trust deficit they create.

Other hybrid approaches have included using volatile crypto assets as partial collateral (overcollateralized models like DAI) or using delta-neutral derivatives positions to create synthetic stability. Each introduces its own risk profile, from oracle manipulation vulnerabilities to liquidation cascades during market stress.

Risk Comparison: Algorithmic vs. Backed Stablecoins

Understanding the risk profiles of different stablecoin models helps users and developers make informed decisions. Research into yield-bearing stablecoins like USDB provides useful context for comparing these approaches.

The core tradeoff is between decentralization and stability. Fully backed stablecoins like those classified as asset-referenced tokens provide reliable stability but require trusting an issuer. Algorithmic models offer trustlessness but have not demonstrated the ability to maintain stability under stress.

Risks and Considerations

Reflexivity and Confidence Dependence

The fundamental risk of algorithmic stablecoins is that their value depends entirely on collective belief in the system. Unlike collateralized stablecoins where each token has a claim on real assets, algorithmic stablecoin tokens have no intrinsic redemption value. They are worth $1 only because market participants collectively treat them as worth $1, and the mechanism works only as long as enough participants are willing to arbitrage deviations.

This creates a reflexive dynamic where confidence is self-reinforcing in both directions: confidence attracts more users and liquidity, which strengthens the peg, which builds more confidence. But loss of confidence triggers selling, which weakens the peg, which destroys more confidence. There is no external anchor to break this feedback loop.

Precision and Oracle Risks

Algorithmic mechanisms depend on accurate, timely price data to trigger supply adjustments. Oracle manipulation can trick the protocol into expanding or contracting supply inappropriately. Precision decay in price feeds can also cause the mechanism to drift, especially during volatile periods when accurate pricing matters most.

Governance and Upgrade Risks

Most algorithmic stablecoins include governance mechanisms that allow parameter changes: collateral ratios, interest rates, expansion limits, and contraction incentives. These governance decisions can introduce human error or manipulation. Poorly timed parameter changes during market stress have contributed to multiple depegging events.

Regulatory Uncertainty

Regulators worldwide have responded to algorithmic stablecoin failures with increased scrutiny. The EU's MiCA framework and other regulatory proposals distinguish between fully backed stablecoins and algorithmic designs, often imposing stricter requirements or outright restrictions on the latter. The classification of seigniorage tokens as securities remains an open legal question in most jurisdictions.

This glossary entry is for informational purposes only and does not constitute financial or investment advice. Always do your own research before using any protocol or technology.