An algorithmic stablecoin attempts to maintain its peg through code-based supply expansion and contraction rather than holding real collateral. When the price rises above $1, new coins are minted (increasing supply, pushing price down). When it falls below $1, coins are burned or incentives redirect demand back. Most attempts have failed catastrophically, most notably Terra/UST in May 2022.
What Is an Algorithmic Stablecoin (And Why They Are Structurally Riskier)?
3 min read
The short version
Algorithmic stablecoins try to stay at $1 using only software and incentives, without backing each coin with actual dollars or crypto. The idea sounds elegant: the algorithm expands supply when demand is high and contracts it when demand drops. In practice, this works fine when confidence is high but collapses in a death spiral when confidence breaks, because there is nothing backing the coin except belief in the system.
How It Works
Common design (Terra/UST model): UST was paired with LUNA. To mint 1 UST, you burned $1 of LUNA. To redeem, you burned 1 UST and received $1 of LUNA (newly minted). The peg relied on arbitrage: if UST < $1, you could buy UST cheap and redeem for $1 of LUNA (profit). If UST > $1, you could mint new UST by burning LUNA (profit). The fatal flaw: in a confidence crisis, mass redemption of UST creates massive new LUNA supply, crashing LUNA's price, which reduces the value backing each UST, triggering more redemptions in a death spiral. May 2022: $18B of UST and $40B of LUNA value were destroyed in ~5 days. Other designs (Frax, partially algorithmic): use partial collateral + algorithmic mechanisms. These hybrid approaches are more resilient but still carry mechanism risk beyond what fully-collateralized coins face.
The Terra/UST death spiral (May 2022)
Starting state: 18B UST in circulation, LUNA at $80, total LUNA market cap ~$30B. Day 1: Large UST sells push price to $0.98. Arbitrageurs burn UST for LUNA and sell the LUNA for profit, expanding LUNA supply. Day 2: Confidence shakes. More UST holders rush to exit. UST hits $0.90. LUNA sell pressure from redemptions pushes LUNA to $50. Day 3: LUNA at $10. Now the total LUNA market cap ($6B) is less than the UST that needs backing ($18B). There is not enough value to back all the UST even in theory. Death spiral accelerates. Day 5: UST at $0.10, LUNA at $0.0001. Both are functionally worthless. $58 billion in combined value evaporated in under a week. No algorithmic mechanism could arrest the spiral once confidence collapsed.
What People Get Wrong
Algorithmic stablecoins are more decentralized so they are better
Decentralization does not equal safety. A decentralized system that collapses to zero is worse than a centralized one that maintains its peg. The decentralization of an algorithmic stablecoin gives you no recourse when the mechanism fails.
The Terra collapse was a one-time anomaly
Multiple algorithmic stablecoins have failed: Basis Cash, Iron Finance (partial collapse in 2021), Empty Set Dollar, and others. The structural vulnerability (reflexive death spirals under stress) is a known, repeating pattern, not a unique event.
New algorithmic designs have fixed the problem
Partial-collateral designs (Frax V2 is now nearly fully collateralized) have moved away from pure algorithmic approaches precisely because the pure model does not work in practice. The market has largely spoken: algorithmic-only pegs are not trustworthy for significant capital.
Keep Reading
Sources & Further Reading
- Terra Post-Mortem (Jump Crypto)
Detailed timeline of the UST/LUNA collapse from rise to crash
Questions People Also Ask
- Are any algorithmic stablecoins still operating?
- Most pure algorithmic designs are dead or have pivoted. Frax started as partially algorithmic but has moved toward full collateralization. USDD (Tron) claims algorithmic mechanisms but maintains large reserves. The trend is strongly toward collateralized models after Terra.
- Could a better algorithm prevent a death spiral?
- This remains an open research question, but the core challenge is economic, not technical. When confidence breaks and everyone tries to exit simultaneously, any system relying on faith rather than actual backing faces the same reflexive pressure. Better algorithms can delay the spiral but may not prevent it fundamentally.
- What should I do if I hold an algorithmic stablecoin?
- Understand the specific mechanism, the backing ratio, and what happens under stress. If the coin has less than 100% hard collateral, treat it as carrying higher risk than USDC or DAI. Keep only amounts you could afford to lose entirely. The 2022 collapse showed how fast "stable" becomes worthless.