How Stablecoins Work Under the Hood
2026-06-13
A stablecoin is a crypto token engineered to hold a steady value, almost always one US dollar. They are the plumbing of crypto: most trading, lending, and on-chain payments settle in stablecoins because nobody wants to quote prices in an asset that swings 5% before lunch. But "one token, one dollar" is a promise, and the way each design keeps that promise determines how it behaves under stress.
The Peg Is Maintained by Arbitrage
A stablecoin holds its peg not by decree but because traders profit from correcting any deviation. If the token trades at $0.99 and can be redeemed for $1.00, arbitrageurs buy it cheap and redeem for a guaranteed cent. If it trades at $1.01, they mint new tokens for $1.00 and sell. This minting/redeeming pressure pulls the market price back. The credibility of redemption is therefore the whole game.
Fiat-Backed (USDC, USDT)
The simplest model: a company holds real dollars and short-term Treasuries in a bank, and issues one token per dollar held. Redeem a token and the issuer wires you a dollar and burns the token. It is custodial and centralized, you trust the issuer's reserves and attestations, and the issuer can freeze addresses. The trade-off is that this is the most robust peg in practice, as long as the reserves are real, liquid, and actually redeemable.
Crypto-Collateralized (DAI)
Here the backing is other crypto, locked in smart contracts, with no single fiat issuer. Because crypto is volatile, these systems are over-collateralized: lock $150 of ETH to mint $100 of stablecoin. If the collateral's value falls toward the debt, the position is automatically liquidated to keep the system solvent. The backing is transparent and auditable on-chain, but capital-inefficient and exposed to sharp crashes that can trigger liquidation cascades. In practice the "decentralized" label has frayed: DAI (now part of the Sky/USDS ecosystem) has come to lean heavily on centralized assets like USDC and tokenized treasuries for stability, a reminder that pure crypto-collateral is hard to scale.
Algorithmic (and Why Terra Failed)
Algorithmic designs try to hold the peg with supply mechanics and a sister token instead of hard collateral, expanding and contracting supply via incentives. The 2022 collapse of Terra's UST is the cautionary tale: its peg relied on confidence in a paired token (LUNA), and when redemptions spiked, the mechanism minted LUNA into oblivion in a "death spiral," wiping out roughly $40 billion. Purely algorithmic stablecoins have a structural reflexivity problem: the thing backing the peg loses value exactly when the peg is under attack.
The Risks to Watch
- Reserve risk: are the off-chain dollars real, liquid, and fully redeemable? (USDC briefly depegged in 2023 when reserves were stuck at a failing bank.)
- Smart-contract risk: a bug in a collateralized system can drain it.
- Liquidation risk: volatile collateral can cascade in a crash.
- Centralization risk: fiat-backed issuers can freeze, censor, or be shut down by regulators.
- Reflexivity risk: algorithmic pegs can spiral when confidence breaks.
Why They Matter More Every Year
Stablecoins now settle trillions of dollars annually and have become the killer app for crypto rails: instant, global, cheap dollar transfers. That scale is exactly why regulators moved fast, the US GENIUS Act and the EU's MiCA both set rules for reserves, redemption, and disclosure. A stablecoin is only as stable as the weakest link in its backing, and increasingly, the law treats it that way.