How Stablecoins Hold Their Peg: Redemption, Reserves and Liquidity

A stablecoin holds its peg when enough market participants can reliably exchange the token for its reference value, or profitably trade against deviations from that value. Reserves make the promise credible; redemption makes it usable; market liquidity transmits it to holders who cannot redeem directly. A weakness in any one of those connections can produce a price below one dollar even when the other connections still work.

The analytical mistake is to treat the peg as a property of the token alone. It is better understood as the result of a working transaction route. The useful question is not simply whether one dollar of assets exists somewhere. It is who can obtain that dollar, through which institutions, at what cost, and how quickly.

Separate the Target Price From the Available Exit

Consider three statements: a token targets one dollar, its issuer reports one dollar of reserves per token, and a particular holder can turn it into one spendable dollar immediately. These statements describe different things. The first is a design objective. The second is a balance-sheet assertion. The third describes an executable transaction.

That separation explains why a reserve statement and an exchange price can disagree without either necessarily being fabricated. One measures backing under specified valuation and reporting conventions. The other reflects what the next buyer will pay on a specific venue. Transaction size, settlement delays, eligibility and uncertainty sit between them.

The Federal Reserve's research on primary and secondary stablecoin markets distinguishes issuance and redemption with an issuer from trading among other participants. It identifies access to the primary market as an important part of the arbitrage mechanism connecting those markets.

For a holder, this suggests a practical exercise: write out the entire exit route. Selling on an exchange creates exposure to that exchange until withdrawal completes. Redeeming through an issuer creates a different route involving eligibility, settlement and banking arrangements. A wallet balance is the starting point, not evidence that every subsequent step is available.

How Arbitrage Connects the Markets

Suppose an eligible intermediary can purchase a token for $0.995 and redeem it for $1. The apparent spread is half a cent. Buying tokens tends to support the market price, while redemption reduces the outstanding supply. If the token instead trades above a dollar, an eligible participant may be able to create new tokens at par and sell them into that demand.

Neither direction operates for free. The intermediary needs capital, operational access, confidence in settlement and a sufficiently large spread. It may also face limits on how much it can process at once. A theoretical arbitrage can therefore exist while the actual market remains temporarily away from par.

An intermediary need not believe an issuer is insolvent to demand a discount. Waiting for cash consumes balance-sheet capacity. A larger position can exceed a risk limit even when a smaller position looks attractive. The price at which a route becomes usable depends on the participant operating it.

This is why the depth of the arbitrage network matters as much as the existence of a redemption button. Several independent, well-capitalized operators with working settlement routes create a different market from one where a single intermediary provides most of the connection. A long list of exchanges does not establish that independence.

A Worked Example of the Arbitrage Band

The following figures are hypothetical and describe a simplified transaction, not current fees or an investment opportunity. An intermediary considers buying 100,000 tokens at $0.995, spending $99,500. Redemption would produce $100,000 before costs, creating a gross spread of $500.

The modeled residual is positive, but it is not a risk-free profit. It excludes the possibility of failed execution, delayed redemption, a frozen account, changing fees and a loss on the backing assets. A professional intermediary may require an additional risk margin before committing capital.

Now reduce the discount to one tenth of a cent: a purchase at $0.999 creates only $100 of gross spread. Using the same cost assumptions produces roughly $268 of costs. The trade no longer works. A small persistent deviation can therefore be consistent with ordinary transaction frictions.

Under these assumptions, the approximate break-even price is $0.99732. This follows from setting redemption proceeds equal to purchase price plus proportional trading and financing costs and fixed charges. The number has no predictive value outside the example; the analytical value is the list of variables. A longer settlement delay or higher funding rate widens the required discount. A cheaper, faster route narrows it.

For a stressed token, the exercise changes from fee calculation to uncertainty assessment. Our guide to distinguishing liquidity shocks from solvency problems examines why the observed discount cannot identify the cause by itself.

Reserve Quality Has Several Dimensions

Backing is more informative when described as a portfolio with constraints than as a single percentage. Useful dimensions include credit quality, maturity, market value, concentration, custody arrangements and access to cash. A reserve asset can be valuable and difficult to monetize at the required moment. It can also be liquid in normal conditions and less liquid during a common shock.

Circle, for example, publishes reserve composition and assurance information for its stablecoins. Those disclosures illustrate the types of documents to inspect: categories of reserve assets, outstanding circulation, reporting dates and the scope of the accompanying assurance. They are issuer disclosures, not a substitute for independently evaluating every claim or forecasting future redemption performance.

A reserve review should ask what changes between reporting dates. Issuance changes liabilities; redemptions reduce both liabilities and available assets; asset prices and operational arrangements can change independently. Comparing figures from different times can manufacture a coverage ratio that never actually existed.

It is also necessary to distinguish asset visibility from the right to use those assets for token holders. A public wallet or a published reserve total does not resolve every question about obligations, restrictions and competing claims. The distinction is developed in proof of reserves versus a financial statement audit.

Redemption Rights Depend on the Holder and the Arrangement

The word redeemable often hides a conditional sentence. Eligible for whom? Through which entity? In which jurisdiction? To which bank account? Under which fees and processing rules? Those details should be checked in the applicable terms, rather than inferred from a logo or an exchange listing.

As one concrete example, Circle's published non-EEA USDC terms distinguish holders with Circle Mint accounts from other holders and attach conditions to direct redemption. The same page points EEA holders to different documentation. This is an illustration of why terms matter, not a claim that every issuer uses the same restrictions or that one region's rules apply worldwide.

An investor who cannot use the issuer's primary market is relying on someone else to connect their tokens to that market. That can work efficiently, but it introduces an additional dependency. It also changes what a successful small test establishes. Selling a few tokens on an exchange verifies that particular route at that time; it does not verify a large issuer redemption during stress.

The operational question becomes especially important for organizations with scheduled payments. A token may maintain a tight secondary-market price while a specific business cannot complete its required bank transfer. Price stability and cash availability are related outcomes, not interchangeable measures.

Different Designs Support the Peg in Different Ways

The same one-dollar target can be attached to different mechanisms. In a fiat-reserve model, the analysis emphasizes assets and issuer redemption. In a crypto-collateralized design, it emphasizes collateral values, liquidation rules and the reliability of pricing inputs. A design relying heavily on incentives to exchange against another endogenous token has a different stress path again.

The following comparison is an analytical map, not a ranking of products.

A hybrid is not automatically diversified. If several components ultimately depend on the same dollar token, bank or liquidity pool, their labels can exaggerate the number of independent supports. The task is to trace dependencies until they reach assets, institutions or enforceable mechanisms that are meaningfully distinct.

Similarly, a yield-bearing token that invests in government instruments deserves its own analysis. Tokenized Treasury products can have different ownership, valuation and redemption structures from a payment stablecoin. Similar price behavior does not make their legal or operational design identical.

A Peg Can Fracture Across Chains and Venues

A ticker is not a complete asset identifier. A native token issued on one chain, a bridged representation elsewhere and an exchange balance may expose the holder to different intermediaries. Two balances with similar names can therefore have different redemption paths.

Suppose the underlying issuer continues normal redemption, but a particular bridge stops releasing assets. The bridged representation can trade at a discount while the native token remains near par. Reading the event as an issuer-wide failure would confuse the broken link with the whole system.

The reverse error is also possible. A small venue can show a reassuring price while its order book has too little depth for a meaningful exit. To understand this distinction, map the additional claims introduced by wrapped tokens and bridges before treating different token representations as interchangeable cash.

Yield Is a Separate Promise

A stable price target says nothing about who receives income generated by the underlying assets. If an issuer earns interest on reserves, it does not follow that every token holder has a contractual right to that interest. If an application advertises returns on deposited stablecoins, identify the additional transaction: lending, market making, incentives or some other use of capital.

Combining a peg mechanism with a yield strategy can introduce exposures that were absent from the token itself. The correct comparison is between complete positions, including the application and exit route. The framework in where crypto APY comes from and who pays it helps separate the cash-flow source from the token's reference price.

Build a Peg Review That Can Be Updated

A reusable review can fit on one page. Record the exact asset and chain, the applicable issuer or protocol, the reserve or collateral description, the holder's exit route, its costs, and the documents supporting each statement. Add the date of each observation. Mark unavailable information as unknown rather than treating it as either reassurance or proof of failure.

Then specify what would change the assessment. Examples include a revised redemption policy, a new reserve concentration, loss of a banking route, a material change in collateral, or a persistent gap between quoted prices and executable exits. This creates a monitoring process based on mechanisms instead of reacting to every small movement in the last traded price.

There are limits to what an outside observer can establish. Public disclosures can be delayed, contractual interpretation can require jurisdiction-specific expertise, and a normal-market test cannot reproduce a system-wide run. The goal is not to certify that a peg can never break. It is to identify which links support it, which remain uncertain, and what evidence would reveal a change before the token is needed for a payment.