A stablecoin is a token that aims to trade at a fixed value against some reference, usually a national currency. The word “stable” describes an intention. Whether it is achieved depends on a mechanism, and the mechanisms differ enough that treating all of them as one category is a mistake.
Redemption is what does the work
For a token backed by reserves, the peg holds because of arbitrage. If the token trades below the reference, someone can buy it cheaply and redeem it with the issuer for full value, and that buying pressure closes the gap. If it trades above, someone can create new tokens at par and sell them.
Notice what this depends on. Not the existence of reserves — the ability to reach them. A token fully backed by assets that cannot be redeemed promptly, or can only be redeemed by a small set of approved counterparties, has a weaker peg than its backing suggests. The reserve matters; the redemption path matters more.
What the reserve is made of
Reserves are not interchangeable. Short-dated government debt and bank deposits behave very differently under stress from commercial paper or secured lending, and the difference only becomes visible at exactly the moment you would rather it did not.
Two questions are worth more than the headline backing ratio: how quickly can these assets be converted to cash without a loss, and who is holding them. A reserve concentrated in a small number of institutions carries the risk of those institutions, and that risk transfers straight to the token. We track supply and reserve disclosures on the stablecoins page and the reserve watch.
Attestation is not audit
Most reserve reporting takes the form of an attestation — a professional firm confirming that particular assets were present on a particular date. That is useful and it is narrower than people assume. It is a snapshot, not a continuous guarantee, and it generally does not opine on whether the arrangement is sound.
Reading a monthly attestation as though it were a live solvency feed is one of the more common errors in this corner of the market.
Collateralised and algorithmic designs
Some tokens are backed by other crypto assets held in excess of the amount issued, with automated liquidation if the collateral falls too far. The over-collateralisation exists precisely because the backing is volatile. These systems are transparent, since the collateral is on-chain and verifiable, and their failure mode is well understood: a fast enough fall in collateral value can outrun the liquidation machinery.
Designs that rely mainly on incentives and supply adjustment rather than on assets have a structurally harder problem, because the mechanism that is supposed to defend the peg depends on confidence in the same system whose peg is being questioned. That circularity is the whole difficulty, and it is worth naming plainly.
The question to ask
Not “is it backed?” but “if holders wanted out at the same time, what exactly would happen, and who would be first in line?” The answer to that describes the risk far better than any ratio.
See our guide to what stablecoins are for the fundamentals. This is information, not advice, and stablecoin regulation varies by jurisdiction.