Algorithmic stablecoins try to hold a $1 peg using code and market incentives rather than holding a full reserve of cash or crypto behind every token. Instead of being backed dollar-for-dollar, they expand and contract supply automatically, often paired with a second “balancer” token, to push the price back to target. The idea is elegant, but the history is brutal: the May 2022 collapse of TerraUSD (UST) destroyed roughly $60 billion in days. This guide explains how algorithmic stablecoins work, why they are so fragile, and what the Terra disaster taught the market.
Key takeaways
- Algorithmic stablecoins maintain a peg through algorithms and incentives, not full collateral backing.
- Many rely on a second volatile token to absorb price pressure, which creates a dangerous feedback loop.
- The May 2022 TerraUSD (UST) and LUNA collapse erased about $60 billion and is the defining example of the model’s fragility.
- They are vulnerable to “death spirals” when confidence breaks and the mechanism feeds on itself.
- Algorithmic stablecoins are far riskier than fiat-backed or overcollateralized coins; treat them as experimental.
What is an algorithmic stablecoin?
A stablecoin is a token designed to hold a steady value, usually $1. Most do this with reserves: fiat-backed coins like USDC hold cash and Treasuries, and crypto-backed coins like DAI lock up more crypto than they issue. Algorithmic stablecoins take a different path. Instead of holding enough assets to redeem every token for a dollar, they use an algorithm to manage supply and demand so the market price stays near the peg.
In short, the peg is defended by code and economic incentives rather than by a vault of assets. That makes them capital-efficient in theory, because they do not need to lock up large reserves, but it also means there may be little or nothing to fall back on when confidence fails. To compare against the safer models, see our stablecoin guide.
How algorithmic stablecoins try to hold the peg
There are a few designs, but most share the goal of adjusting supply to nudge the price toward $1.
Seigniorage / two-token models
The most famous design uses two tokens: the stablecoin and a separate, volatile “share” or balancer token. The system promises that you can always swap $1 of the stablecoin for $1 worth of the volatile token, and vice versa, regardless of the stablecoin’s market price. This arbitrage is supposed to keep the peg:
- If the stablecoin trades above $1: users mint new stablecoins by burning the volatile token, increasing supply and pushing the price down toward $1.
- If the stablecoin trades below $1: users burn stablecoins for $1 worth of the volatile token, reducing supply and pushing the price up.
On paper, rational arbitrageurs keep the price pinned. In practice, this depends entirely on the volatile token holding meaningful value, which is exactly what breaks under stress.
Rebase models
Some algorithmic coins adjust the number of tokens in every holder’s wallet automatically. If the price is above target, balances expand; if below, they contract. Your token count changes, but the protocol aims to keep the per-token price near the peg. These have generally struggled to achieve durable stability.
Partially collateralized hybrids
A few designs mix a partial reserve with an algorithmic component, holding some collateral while relying on the algorithm for the rest. This reduces, but does not remove, the core fragility, because the uncollateralized portion still depends on confidence.
The TerraUSD (UST) collapse: the defining failure
The clearest lesson in algorithmic stablecoins is TerraUSD (UST), the seigniorage stablecoin paired with the LUNA token. UST maintained its peg through the mint-and-burn mechanism described above: $1 of UST was always swappable for $1 of LUNA. The system grew rapidly, helped by the Anchor protocol, which advertised an unusually high, roughly 20% yield on UST deposits, drawing in tens of billions of dollars.
That yield was a magnet, but it also masked the structural fragility. In May 2022, large UST withdrawals and sell-offs pushed UST below its peg. As holders rushed to swap UST for LUNA, the protocol minted enormous amounts of new LUNA, crashing LUNA’s price. The falling LUNA made it even harder to restore the UST peg, which triggered more panic selling, which minted still more LUNA. This self-reinforcing collapse is known as a “death spiral.”
Within days, UST fell far below $1 and LUNA’s value was almost entirely wiped out. The episode erased roughly $60 billion in value, devastated retail holders, and contributed to a wave of failures across crypto lenders and funds that had exposure. It stands as the starkest demonstration that an algorithmic peg backed mainly by confidence can vanish almost instantly.
Why algorithmic stablecoins are fragile
The Terra collapse was not bad luck; it exposed structural weaknesses inherent to the model:
- Reflexive dependence: the stablecoin’s stability depends on the value of a volatile token, whose value depends on confidence in the stablecoin. When one falls, both can fall together.
- The death spiral: once the peg breaks meaningfully, the mechanism that is supposed to restore it can instead accelerate the collapse by flooding the market with the volatile token.
- Confidence is the real backing: with little or no hard collateral, belief in the system is the asset. Belief can disappear in hours.
- Yield as a crutch: many algorithmic systems used unsustainable yields to attract capital, which inflated them until the incentives broke.
- Bank-run dynamics: if everyone tries to exit at once, there is no reserve to satisfy redemptions at par.
These are not bugs that a better algorithm easily fixes; they stem from trying to create stability without sufficient assets behind it. Understanding the underlying market plumbing, including how decentralized exchanges and liquidity pools behave under stress, helps explain how quickly a depeg can cascade once liquidity dries up.
Algorithmic vs collateralized stablecoins
It helps to see where algorithmic coins sit relative to the alternatives.
| Type | Backing | Example | Relative risk |
|---|---|---|---|
| Fiat-backed | Cash and short-term Treasuries held in reserve | USDC, USDT | Lower; depends on issuer and reserves |
| Crypto-backed (overcollateralized) | More crypto locked than tokens issued | DAI | Moderate; depends on collateral and liquidations |
| Algorithmic | Algorithm and incentives, little or no full collateral | TerraUSD (failed) | Highest; vulnerable to death spirals |
For a broader look at how decentralized finance assembles these building blocks, our what is DeFi guide is a useful companion.
Warning signs to watch for
The Terra collapse and other failed projects share recognizable patterns. If you are evaluating any stablecoin and see several of these, treat it as a serious warning:
- Unusually high, “guaranteed” yields. Double-digit, fixed-sounding returns on a supposed dollar are almost always subsidized and unsustainable. Anchor’s roughly 20% rate on UST was the lure that masked the fragility.
- Little or no hard collateral. If the project cannot clearly show real assets backing the tokens, its stability rests on confidence alone.
- Dependence on a sister token. A peg defended by minting and burning a second volatile token contains the seeds of a death spiral.
- Opaque mechanics. If you cannot understand how the peg is maintained after a genuine attempt, that is itself a reason for caution.
- Reflexive growth. Rapid expansion driven mainly by yield incentives, rather than organic demand for payments or settlement, often unwinds just as fast.
None of these guarantees failure, but together they describe the profile of the designs that have collapsed. Healthy skepticism is the cheapest protection available.
What the market learned from the collapse
The fallout from Terra reshaped how the industry and regulators think about stablecoins. Several large crypto lenders and funds that had exposure to UST, LUNA or the broader leverage built around them failed in the following weeks, showing how a single uncollateralized design can transmit contagion across the ecosystem. Investors became far more focused on what actually backs a stablecoin, demanding transparency about reserves and redemption. Policymakers in multiple jurisdictions began drafting rules that specifically address, and often discourage, uncollateralized algorithmic designs. The episode also hardened a simple principle that experienced users now repeat: if you cannot identify the assets that would let you redeem your token for a dollar in a panic, you are holding confidence, not a dollar.
Do algorithmic stablecoins have a future?
After Terra, appetite for purely algorithmic stablecoins collapsed, and regulators in several jurisdictions moved to restrict or scrutinize uncollateralized designs. Newer projects tend to add real collateral or hedging rather than rely on confidence alone, blurring the line with collateralized models. Some experimentation continues, but the market has largely concluded that a stablecoin needs substantial assets behind it to survive a crisis. Anyone tempted by a high-yield algorithmic coin should remember that the yield is usually the bait, not the reward, and that “stable” is not a guarantee.
FAQ
What is an algorithmic stablecoin in simple terms?
It is a stablecoin that tries to stay at $1 using software rules and market incentives instead of holding a full reserve of cash or crypto for every token. It typically expands or shrinks supply, often via a second volatile token, to push the price back toward the peg. Because there may be little hard collateral, its stability depends heavily on continued confidence.
Why did TerraUSD (UST) collapse?
UST relied on swapping $1 of UST for $1 of its sister token LUNA to defend the peg. When UST lost its peg in May 2022, mass redemptions minted huge amounts of LUNA, crashing LUNA’s price, which made the peg harder to restore and triggered more selling, a death spiral. The mechanism meant to stabilize UST instead accelerated its collapse, erasing roughly $60 billion.
Are algorithmic stablecoins safe?
No. They are the riskiest category of stablecoin. Their stability rests on confidence and a self-reinforcing mechanism that can fail catastrophically and quickly, as TerraUSD showed. Even well-designed versions remain fragile under stress because there is often little hard collateral to fall back on. They should be treated as experimental and high-risk, not as a safe dollar substitute.
How are algorithmic stablecoins different from USDC or DAI?
USDC is fiat-backed, holding cash and Treasuries in reserve. DAI is crypto-backed, locking up more crypto than the DAI it issues. Both have real assets behind them. Algorithmic stablecoins instead rely mainly on an algorithm and incentives, with little or no full collateral, which makes them far more vulnerable to losing their peg permanently.
Crypto is volatile and risky; this is education, not financial advice. Do your own research.
