Algorithmic stablecoin

The best-known model is the dual-token or seigniorage design: a stable token plus a support token that absorbs volatility. The protocol lets users swap one unit of the stablecoin for an equivalent value of the support token through minting and burning. In theory, if the stablecoin falls below one, it is burned and support token is minted to remove supply and push the price up. The problem is that the same mechanism can invert and feed the decline.

It is the most fragile stablecoin type and the canonical study case for the systemic risk of models without full collateral. After the Terra episode, both MiCA in the European Union and the GENIUS Act in the United States lean toward 1:1 backing, leaving pure algorithmic designs largely outside the regulated payment-stablecoin framework. The practical takeaway is that real reserve backing is now the standard, not one option among equals.

TerraUSD (UST) was the reference collapse in May 2022, when holders swapped UST for its LUNA support token, hyperinflating its supply and feeding the fall of both to residual values within days; unsustainable yields near 20 percent a year had concentrated demand in something without sufficient collateral. As a nuance, FRAX began as a hybrid fractional-algorithmic model and over time migrated toward greater collateralization, so labeling it algorithmic is debatable.