The Stablecoin Redemption Pipeline: Issuer Liquidity Mechanics

As of March 2026, the standard Circle Mint tier caps daily gross redemptions at $10 million, with a net redemption overage fee applying when net redemptions exceed $40 million during the evaluation window. A published quantitative limit on how much can leave through the front door per day is the single most useful disclosure in this market, and it points at the part of the pipeline that is usually discussed last.

The Pipeline Has Two Tiers

Primary redemption is a wholesale privilege. Direct one-for-one redemption with the issuer is restricted to institutional accounts that have completed business verification: a Circle Mint account for USDC, a verified direct customer relationship for USDT with a $100,000 minimum and a fee of the greater of $1,000 or 0.1 percent, plus a $150 verification charge. Everyone else transacts on secondary venues, and the structure is close to correspondent banking, where wholesale participants provide retail access.

That distinction is not a technicality, and the governing statute preserves it. The GENIUS Act requires payment stablecoins to be redeemable for a fixed monetary value and obliges issuers to honor redemptions within two business days, publish a redemption policy, disclose fees in plain language and give seven days notice before changing them. It does not require an issuer to offer redemption to every holder, so a platform can be fully compliant while restricting the primary window to institutional accounts.

The capacity consequence follows arithmetically. Against roughly $187 billion of USDT and $76 billion of USDC outstanding in mid-2026, the primary market can process only a fraction of supply per day even with dozens of large participants. The front door is narrow by design, and it is narrow in a way that is published.

Who Actually Bears the Gap

Follow a single exit. A holder without an issuer account sells into a secondary venue. Somebody buys. If the price is below par, the buyer's economic case is to redeem at par through the primary window and collect the difference.

Between those two events there is a position. The buyer owns the token from the moment of purchase until the issuer's wire arrives, which under the statutory standard can be two business days. During that interval the issuer's reserves are irrelevant to the price, because the reserves are not what is holding the quote up. What holds the quote up is a balance sheet willing to own the token while the redemption clears.

So the pipeline's real capacity is not the issuer's liquidity. It is the sum of intermediary balance sheet prepared to warehouse the gap, and that quantity is set by other people's risk limits, funding costs and willingness, none of which appear in any reserve disclosure.

What Breaks First

Not the reserves. The first thing that fails is the carry, and it fails on a schedule anyone can read in advance.

Consider the timing. A Friday evening stress event leaves primary redemptions queued until Tuesday or Wednesday under a two-business-day standard, while secondary markets reprice continuously through the weekend. An arbitrageur buying at a discount on Saturday is underwriting a four-day hold with no ability to hedge the issuer leg, and the fee schedule works against them at exactly that moment: redemption costs step up with volume, and Circle's March 2026 model added a monthly net flow adjustment under which sustained net redeemers face incremental charges while net minters may earn credits. A participant who has been redeeming all month faces a worse price for doing the thing that defends the peg.

The failure sequence is therefore legible. Gross daily caps bind first and push flow to secondary. The discount widens to compensate carriers for the wait. If the discount widens past what the carriers' risk limits tolerate, they stop buying, and the quote is then set by sellers alone. None of those steps involve a question about reserve quality, and none of them would be prevented by better reserves.

Detecting It Early

The useful observable is not the discount itself but whether it persists longer than arbitrage economics can explain. Compute the carrying cost of a redemption initiated now: the issuer fee at the relevant tier, funding for the expected settlement window, and any minimum size that excludes smaller participants. That number is the discount at which buying becomes rational.

A discount wider than that figure, sustained while the primary window is open, is the signal. It says that participants who could profitably redeem are declining to, which is a statement about their capacity or their willingness rather than about the issuer. That is the early warning, and it is computable from published fee schedules and an observed price.

Three other things are worth watching continuously, and a technical owner can instrument all of them: how close the firm's own redemption requests are running to the published daily cap, the time between request and wire receipt measured as a distribution rather than an average, and whether the discount appears at times that correspond to banking closures rather than to news.

Controls, Costs, and What Is Improving

The controls follow from the above and most of them are contractual. Hold primary accounts at more than one issuer so that a single queue is not the only exit. Know your own tier, cap and fee schedule as numbers rather than as a relationship. Size any position that depends on redemption to the amount that clears within one window, not to total holdings. And treat the net flow adjustment as a real cost of being a consistent redeemer, because that is what it is.

Costs sit in three places. The explicit fee schedule, which is published. The funding cost of the settlement window, which is the arbitrageur's and reaches holders as a discount. And the opportunity cost of maintaining accounts with multiple issuers to preserve optionality, which is the one most firms skip and then discover they needed.

What is improving is genuinely the disclosure. Published daily caps, tiered fees with notice periods before changes, a statutory maximum settlement time and a required published redemption policy together make this pipeline specifiable in a way it was not two years ago. A pipeline whose capacity and price are published can be modeled, and modeling it is now a reasonable expectation rather than guesswork. The remaining gap is that the one quantity that actually determines resilience, the balance sheet willing to carry the position between purchase and wire, is held by parties who publish nothing at all.

For informational purposes only. Not an offer to buy or sell any security. Available only to accredited investors who meet regulatory requirements.

‍