The core idea: a stablecoin is a cryptocurrency whose value is designed to track a stable reference asset — almost always the US dollar — so one token is supposed to be worth one dollar no matter what the broader crypto market is doing. That stability is the entire product. Without it, a stablecoin is just another volatile coin. The interesting question is not “what is it” but “what mechanism is actually keeping it at a dollar, and what happens when that mechanism is stressed?”
![]()
Image: Bitcoin on Laptop Keyboard, CC BY 2.0, via Wikimedia Commons (representative of the crypto asset class)
The Three Ways to Hold a Peg
There is no single stablecoin design. The mechanism behind the peg determines both how trustworthy it is and how it can fail.
Fiat-backed (off-chain reserve). The issuer holds actual dollars (or near-cash equivalents like short-term Treasury bills) in a bank and promises one token redeems for one dollar. Tether (USDT) and USD Coin (USDC) are the dominant examples. The peg is held by arbitrage: if the token trades below $1, buyers scoop it up and redeem it with the issuer for a dollar profit, pulling the price back; if it trades above $1, the issuer mints more. The weakness is trust in the reserve — you are relying on the issuer to actually hold the assets they claim, and on the banks holding those assets not failing.
Crypto-backed (on-chain, over-collateralized). Instead of dollars in a vault, these hold other cryptocurrencies as collateral, locked in smart contracts. Because crypto itself swings in price, the system demands more collateral than the stablecoin is worth — often $1.50 or more of ETH locked for every $1 issued — so a price drop still leaves the peg covered. MakerDAO’s DAI is the canonical example. The trade-off is capital inefficiency: you must lock up excess value, and the peg depends on the collateral staying liquid during a crash.
Algorithmic (seigniorage or hybrid). These try to hold the peg with code and incentive design rather than full reserves — expanding supply when price is high, contracting it when price is low. This is the most fragile design, because when confidence breaks, the mechanism that should defend the peg can accelerate the collapse instead. The 2022 failure of TerraUSD (UST) is the cautionary case: once its sister token lost value, the death spiral was fast and total.
Where Each One Breaks
Fiat-backed stablecoins break when the reserve is not what it claims to be — whether through poor auditing, fractional reserves, or a bank failure freezing the assets. The 2023 collapse of Silicon Valley Bank briefly took USDC below $0.88 because a chunk of its reserves were stuck in that bank, before recovery once the deposits were made whole.
Crypto-backed stablecoins break when collateral falls faster than the system can liquidate it, or when the smart contract has a bug. The over-collateralization buffer is the whole defense, and in an extreme, correlated crash it can be tested hard.
Algorithmic stablecoins break on confidence. They have the least hard backing, so the peg is a narrative as much as a mechanism. When holders doubt it, they sell, the mechanism issues more tokens to absorb the sell, supply balloons, and the price falls further — the death spiral. No external asset stops it.
Why Stablecoins Matter Beyond Trading
Stablecoins are not only for speculators. In countries with high inflation or restricted access to dollars, a dollar-pegged token can be a practical store of value and a way to send cross-border payments without a local banking system. They are also the settlement layer for most decentralized-finance activity: lending, swapping, and yield strategies are overwhelmingly denominated in stablecoins because nobody wants to borrow in an asset that might double or halve overnight.
The honest limitation: a stablecoin is only as stable as the weakest link in its mechanism and the institutions behind it. “Pegged to the dollar” is a design goal enforced by incentives and reserves, not a物理 guarantee.
The Takeaway
When you see a stablecoin advertised as “$1 always,” the real question is how. Fiat-backed relies on reserves you must trust; crypto-backed relies on over-collateralization you can verify on-chain; algorithmic relies mostly on belief. The first two have held through stress (with scares); the third has a documented body count. Understanding which bucket a token falls into tells you more about its real risk than any marketing claim.
Sources: Wikipedia — “Stablecoin” (mechanism taxonomy and TerraUSD/USDC case history); issuer disclosures for USDT/USDC reserve composition as published.