Stablecoins Under Stress Do Not Behave Like One Safe Haven

Stablecoins do not behave like a single safe haven when markets break. Recent crises show that design and plumbing matter as much as brand. Verified: during the March 2023 Silicon Valley Bank failure, USDC traded as low as $0.8774 on Coinbase before recovering, while operational backlogs cleared only after U.S. banking reopened and resolution steps were announced. Verified: during the May 2022 Terra collapse, the algorithmic UST lost its peg entirely, wiping out roughly $40–45 billion of market value across UST and LUNA. Verified: even the largest fiat-backed stablecoin, USDT, briefly traded off-peg around $0.94–$0.97 amid heavy redemptions during the same 2022 contagion.

These episodes are not footnotes. Circle disclosed that SVB failed to honor a pre-failure withdrawal request for $3.3 billion of USDC reserves, roughly 8% of USDC’s backing at the time, spotlighting bank counterparty risk inside fiat-backed models. U.S. policy snapshots now cite these stress events to assess stablecoin fragility and interconnectedness. The data shows one family resemblance across designs: when stress hits, redemptions can be large and fast. But how those outflows translate into price dislocations, user experience, and systemic spillovers differs sharply by model and market microstructure.

What the 2022–2023 shocks changed about stablecoin safety

Verified: Circle’s SEC filing states that SVB did not honor the withdrawal of $3.3 billion in reserve deposits requested before the bank’s failure. That single counterparty exposure equaled roughly 8% of USDC’s reserves at the time, demonstrating that even fiat-backed stablecoins face off-chain banking risk (Circle S-1). Verified: as the panic spread, USDC’s secondary-market price fell to $0.8774 on March 11, 2023 before recovering alongside regulatory actions and bank-resolution developments (VSC CEX timeline).

These facts reframed two assumptions. First, “cash and equivalents” are not abstract. Where, how, and with whom reserves are held can become the stress point. Second, secondary-market pricing can diverge from redemption value when banking rails are constrained, liquidity fragments across venues, or market makers widen spreads. Reasonable inference: weekend closures and uncertainty about reserve access amplified the USDC discount even if ultimate credit losses were unlikely.

Policy thinking also shifted. U.S. Treasury and TBAC materials now reference the SVB/USDC and Terra episodes when evaluating concentration, market share shifts, and spillovers in digital money markets (Treasury/TBAC). Verified: analyses since 2019 document that the largest stablecoins have experienced single-day redemptions exceeding about 4% of market cap, magnitudes that would be considered material in traditional banking run dynamics (Washington University Law Review).

Evidence from three recent stress episodes

The clearest way to see how heterogeneous “stable” can be is to line up outcomes across designs and venues.

Verified: the USDC dislocation persisted through a weekend when banks were shut. Verified: Circle later said it had “cleared substantially all” mint/redemption backlogs by the close of U.S. banking hours on March 15, and that it redeemed $3.8 billion and minted $0.8 billion since Monday morning of that week (Circle blog). Verified: U.S. policy summaries and market analyses highlight that such single-day redemption magnitudes, across issuers, have exceeded about 4% of market cap in prior episodes (Washington University Law Review).

Reasonable inference: these data points suggest that price stability and redemption capacity hinge on factors outside on-chain mechanics, including custodian concentration, timing of banking access, and the willingness and ability of intermediaries to make markets when uncertainty spikes.

Implications for users, liquidity providers, and venues

When stress hits, “stable” splits into operational reality and market reality.

• Operational reality. Verified: banking access determines redemption pace. Circle cleared backlogs only after U.S. banking hours resumed and a resolution path emerged in mid-March 2023 (Circle blog). Opinion: users relying on instant exit liquidity should assume that fiat-backed redemptions can slow when banks are shut or counterparties are uncertain.
• Market reality. Verified: USDC traded at $0.8774 on March 11, 2023 on at least one major exchange (VSC CEX). Reasonable inference: fragmented order books, wider spreads, and risk-off behavior can push secondary prices away from par even if ultimate redemption at par later resumes.
• Liquidity management. Verified: during the Terra contagion, USDT briefly traded off-peg while processing multi-billion redemptions, as reported by market analyses (Capital.com). Opinion: cross-stablecoin diversification does not guarantee uniform behavior under stress; funding desks should plan for basis risk between quote assets.

Market narrative: exchange operators and DeFi protocols that hardwire assumptions of a single “risk-free” dollar tend to inherit basis risk when one stablecoin deviates and another does not. That risk shows up as volatile collateral values, liquidation cascades, or impaired swap routes, particularly on weekends.

Policy and reserve design takeaways

Verified: U.S. Treasury and TBAC materials reference the March 2023 USDC and May 2022 Terra episodes when assessing stablecoin fragility, concentration, and spillovers (Treasury/TBAC). Opinion: the common policy thread is to reduce uncertainty around reserves, counterparties, and redemption rights so that asset quality, not rumor, sets the price in a stress window.

Reasonable inference: reserve composition and custodian concentration are first-order design choices. The SVB case shows that even low-risk assets can be temporarily unreachable if a key bank fails. The Terra collapse shows that algorithmic feedback loops can erase the peg entirely when market confidence breaks, independent of banking access. Together, they argue for clearer disclosures, tested redemption arrangements, and contingency playbooks that account for weekend liquidity and custody risks.

Counterargument: short-lived depegs can be microstructure, not solvency

The strongest alternative view is that short-lived fiat-backed depegs reflect market microstructure under uncertainty. Verified: USDC’s price recovered after regulators acted and banking resolution developments emerged, and Circle reported clearing backlogs and processing $3.8 billion of redemptions by March 15 (Circle blog). Verified: USDT met large redemptions during the Terra contagion while its peg wobble remained intraday (Capital.com). Opinion: in these cases, the peg discount may have been an opportunistic risk premium for immediacy and venue risk rather than a judgment of insolvency.

Downside scenario: the same heterogeneity that limited contagion from UST to fiat-backed tokens in 2022 can also fragment liquidity when multiple tokens wobble at once. If large issuers share concentrated custody or banking partners, stress can transmit through the off-chain layer. Verified: policy and academic reviews emphasize that single-day redemptions exceeding about 4% of market cap have already occurred in major stablecoins, a sign that run dynamics are plausible if confidence becomes correlated (Washington University Law Review).

What would confirm or weaken this thesis

• Redemption intensity. Verified threshold: single-day redemptions exceeding ~4% of market cap have precedent in major stablecoins. Confirmation would be recurrence across multiple issuers; a weakening sign would be lower and smoother outflows in the next risk event (Washington University Law Review).
• Custodian concentration. Confirmation: issuer filings or disclosures showing heavy reliance on a small set of banks or custodians. Weakening: diversified custody footprints and tested contingency lines. See Circle’s SVB disclosure for how single-counterparty exposure can matter (Circle S-1).
• Secondary-market pricing on weekends. Confirmation: repeated, venue-specific discounts during bank closures. Weakening: tighter spreads and shallower discounts despite headlines, indicating better market-making and access to fiat ramps.
• Issuer operations in stress. Confirmation: visible mint/redeem backlogs or pause notices during shocks. Weakening: rapid processing updates akin to Circle’s March 15 report that backlogs were “substantially” cleared with billions redeemed (Circle blog).
• Market share shifts. Confirmation: policy snapshots noting accelerated rotation between USDT and USDC after stress headlines. Weakening: stability in market shares despite episodic news, per ongoing policy monitoring (Treasury/TBAC).
• Design-specific failures. Confirmation: renewed stress in endogenous-collateral or algorithmic models. Weakening: durable peg performance across non-fiat designs through multiple volatility spikes, which would challenge the current hierarchy of perceived safety.

Editorial conclusion: the record shows that stablecoins are plural, not monolithic. Verified events across 2022–2023 demonstrate that under stress, each token’s peg behavior reflects its reserve design, counterparty map, and the state of banking rails at that moment. Opinion: a workable operating assumption for users and policymakers is to treat basis risk between “dollars” as a feature, not a bug, and to plan liquidity, disclosure, and oversight accordingly.

Disclaimer: This article is provided for informational purposes only. It is not offered or intended to be used as legal, tax, investment, financial, or other advice.