Digital Dollar Perils, Issue 4 of 6: A Stablecoin Is Only as Solvent as the Worst Asset in Its Reserve Stack

Every dollar-pegged stablecoin promises one-to-one redemption against U.S. dollars on demand. The promise is enforceable only to the extent that the reserves backing the stablecoin can actually be liquidated at par. The composition of those reserves, the legal structure that governs them, and the redemption mechanism that connects them to the secondary market all determine whether the promise holds in stressed conditions.

The 2023 USDC depeg, the 2022 Terra collapse, and the historical Tether disclosures each illustrated a different failure mode. The GENIUS Act addresses some of those failure modes and leaves others in place.

What the GENIUS Act actually requires.

Under the GENIUS Act, payment stablecoin issuers must maintain reserves on at least a 1:1 basis with outstanding stablecoins. Reserves must comprise liquid, highly rated assets, including U.S. Treasury bills, deposits at federally insured depository institutions, certain short-duration repurchase agreements, and other instruments specified by regulators. Issuers must publish monthly reserve composition disclosures and submit to examination by registered public accounting firms.

Issuers with more than $50 billion in stablecoins outstanding must submit audited annual financial statements. Issuers are prohibited from paying interest to stablecoin holders. The reserves can be used to redeem stablecoins and to serve as collateral in repurchase transactions, but other uses are restricted. The framework establishes federal supervision for larger issuers and a state regulatory pathway for smaller issuers.

What the Act does not require.

The Act exempts stablecoin issuers from the regulatory capital standards applied to traditional banks. A stablecoin issuer can operate with substantially less capital cushion than a bank holding the same dollar liabilities. The reserves are required to back the outstanding stablecoins on a 1:1 basis, but there is no additional capital buffer required to absorb losses if the reserves themselves lose value.

The Act does not require physical settlement of redemptions within a defined window. Issuers must establish and disclose redemption procedures, but the operational specifics, the speed of redemption, and the conditions under which redemption can be paused are determined by the issuer. In a stressed market, the difference between disclosed procedures and effective access can be substantial.

Most stablecoin users do not redeem directly with the issuer. They sell into the secondary market.

The Brookings Institution and other policy researchers have documented a structural property of the stablecoin market that matters for understanding peg stability. The vast majority of stablecoin users do not transact directly with the issuer. They buy and sell stablecoins on secondary markets through exchanges and arbitrageurs. Tether, the largest stablecoin issuer, averages about six arbitrageurs redeeming tokens on a given day, despite holding more than $160 billion in outstanding stablecoins.

What this means in practice is that the secondary market price is determined by supply and demand on exchanges, not by direct redemption pressure on the issuer. Arbitrageurs limit deviations from par by buying below $1 and redeeming for $1, or selling above $1 and acquiring at par. When arbitrage capacity is constrained, as it was in March 2023 during the SVB weekend, the secondary market price can deviate substantially from the redemption price. That deviation is the holder's actual exposure in a stressed scenario.

Every cash-equivalent instrument has reserve composition risk of some kind.

It is worth being honest here. Reserve composition risk is not unique to stablecoins. A money-market fund has reserve composition. A bank deposit is backed by the bank's loan book and capital structure, with FDIC insurance providing a secondary backstop up to defined limits. A Treasury bill is backed by the full faith and credit of the U.S. government, which is the lowest credit risk available in dollar-denominated instruments but is not zero. Even physical cash has reserve risk in the form of monetary policy that determines purchasing power over time.

What distinguishes stablecoin reserve risk is the combination of factors. The capital cushion is smaller than a bank's. The disclosure regime is younger and less battle-tested than the bank examination regime. The redemption mechanism depends on secondary market arbitrage rather than direct redemption. The legal structure that governs reserve priority in a default scenario is still being established under the GENIUS Act framework. None of those factors makes stablecoins uniquely dangerous. They make stablecoins differently dangerous than other cash-equivalent instruments.

CBDCs eliminate reserve risk by eliminating the reserve. That introduces other risks.

A central bank digital currency is a direct liability of the central bank. There is no reserve to fail because there is no reserve. The central bank can create or extinguish CBDC units through its own balance sheet operations, the same way it manages bank reserves under the current monetary system. From a reserve-risk perspective, a CBDC is structurally cleaner than a stablecoin.

The trade-off is that the holder has now substituted reserve risk for monetary policy risk. The same central bank that guarantees the CBDC also determines the supply of the currency in which it is denominated. The history of fiat monetary policy across jurisdictions makes clear that this is a non-trivial risk for long-duration holdings, even though it is a different kind of risk than reserve composition risk.

The portfolio responses available, and what each one costs.

A family office concerned about reserve composition risk has several responses available. The conventional approach is to diversify cash-equivalent holdings across multiple instrument types. FDIC-insured deposits below the insurance limits at multiple banks. Short Treasury bills held directly rather than through a fund. Money-market funds at custodians with strong risk management track records. Stablecoin exposure limited to transactional positions with explicit credit limits per issuer.

The cost of this approach is the same one that runs through this series. The conventional analog and lightly-digitized instruments work well for the functions they were designed for. As the surrounding monetary system digitizes, the family office maintains a bifurcated treasury structure that requires more operational management. That bifurcation may be the right answer for many families. It also has an ongoing cost.

Some families will consider an allocation to Bitcoin for the long-duration reserve portion of the portfolio, recognizing that Bitcoin is not a cash equivalent and should not be evaluated against that set. Bitcoin has no reserve composition because Bitcoin is the asset. The trade-off is volatility. Bitcoin's historical drawdowns above 70 percent from peak to trough are real and have to be managed within the portfolio framework. Downside-protected structures exist for this purpose. The instrument architecture is conventional even though the asset is unfamiliar.

The question to think through.

The treatment of stablecoins as cash equivalents on a family office balance sheet is a convention worth examining. The conventional treatment may not match the actual risk profile of the instrument. Whatever the family decides to hold, the position should be sized and structured with explicit recognition of what each instrument actually is.

Issue Five of the Digital Dollar Perils Series examines the geopolitical question. CBDCs as instruments of sanctions evasion and dollar containment. Stablecoins as instruments of dollar extension. The portfolio responses available for family offices with international exposure. Full piece in next week's newsletter.

How does your office evaluate reserve composition risk in cash-equivalent allocations?