A stablecoin is a token intended to hold a fixed value against a reference, usually a major currency. The interesting question is never the intention. It is the mechanism.
The reserve-backed model
The most straightforward approach. An issuer holds assets equal to the tokens outstanding and undertakes to redeem tokens for those assets.
Which makes the token a claim on a pool of assets, and the value of that claim depends on three things — what the assets are, whether they are actually there, and whether redemption works when it is needed.
Each of those has been a live issue.
What the reserves contain
The composition matters enormously and has varied widely between issuers.
Short-dated government securities and cash at regulated banks are the conservative end.
Commercial paper, corporate debt, secured loans and holdings of other tokens sit at the other end, and have appeared in reserve disclosures.
The distinction is not academic. Assets that cannot be liquidated quickly at par cannot support redemption during a period of heavy withdrawals, which is exactly when redemption is demanded.
Attestation is not audit
A distinction that matters and is routinely blurred.
An attestation is a limited engagement where an accounting firm confirms reported figures at a point in time against records provided.
An audit is a substantially broader examination with an opinion on the financial statements as a whole.
Most reserve reporting in this sector has been attestation rather than audit.
Which is meaningfully weaker assurance, and the marketing language frequently does not distinguish them.
The banking dependency
A token backed by dollars requires a bank to hold those dollars.
Which reintroduces exactly the dependency the sector positions itself against, and it has been the transmission route for the clearest failures.
One major token lost its peg after a bank holding part of its reserves failed, and recovered only when the deposit position was resolved by authorities.
The episode demonstrated that a fully reserved token is only as sound as the institutions holding the reserves.
The algorithmic model
A different approach entirely, maintaining the peg through an arbitrage mechanism between the stable token and a second, floating token, rather than through reserves.
When the stable token trades below the peg, holders can exchange it for the floating token at par, reducing supply and pushing the price up.
This works while the floating token has value.
The failure mode is that a loss of confidence reduces the floating token's value at the same moment the mechanism requires issuing more of it, which reduces it further.
The largest such system collapsed to near zero within days, destroying an enormous amount of value, and the mechanism failed exactly as its critics had described in advance.
Overcollateralised models
A third approach backs tokens with volatile assets held in excess of the token value, with automatic liquidation when collateral falls below a threshold.
Which can work and is capital inefficient by design, since substantially more value must be locked than is issued.
Its vulnerability is that liquidations happen fastest when prices fall fastest, and mass liquidation into a falling market is its own problem.
The regulatory direction
Several jurisdictions have introduced or proposed frameworks specifically for these instruments.
Common elements are reserve composition requirements, segregation of reserve assets, redemption rights at par, and regular reporting.
Which converges on treating them as a payment instrument or as a form of deposit-like liability rather than as a general asset.
The unresolved question in most frameworks is the treatment of issuers outside the jurisdiction whose tokens circulate within it.
The reasonable summary
These are credit instruments. The token is a claim, and the quality of the claim depends on the assets, the legal structure, the verification and the redemption mechanism.
Which is exactly the analysis applied to any short-term credit instrument, and the fact that it is recorded on a distributed ledger changes none of it.
This is a description of how the mechanisms work and it is not a view on any particular instrument or on whether anyone should hold one. This is a volatile and evolving area, and anyone considering exposure should understand they may lose the entire amount.
Where the yield goes
A commercial detail that explains the business.
An issuer holding reserves in short-dated government securities earns the yield on them.
Token holders generally receive none of it, which means the issuer keeps the return on assets funded entirely by its customers.
At scale and at elevated interest rates, this is an extremely profitable arrangement, and it is why issuance grew so quickly once rates rose.
It also means the issuer's revenue collapses if rates fall, which introduces a business model risk distinct from the peg risk.
Redemption in practice
The right to redeem at par is the anchor, and its practical accessibility varies.
Some issuers redeem only for institutional clients above a minimum size, with retail holders relying on secondary market trading instead.
Which means the arbitrage that maintains the peg depends on a small number of participants having access, and the peg for everyone else rests on their willingness to use it.