Yield-bearing stablecoins carry the same depeg, custodian, and smart-contract risks as plain stablecoins, plus the specific risks introduced by the yield engine. A 9% APY on sUSDe is not a free 9%. The headline rate is gross of expected loss from funding-rate inversions, exchange counterparty issues, and regulatory action. The expected loss is hard to quantify but visible in historical stress events: USDC's 13% depeg during the SVB weekend in March 2023, USDe's 3.5% wobble during the August 2024 funding inversion, the 7-day cooldown queues that lengthened to weeks during multiple stress windows.
​

This article enumerates the risk categories with named historical examples, the mechanism that produces each risk, and the trade-offs between yield engines. The goal is not to argue against yield-bearing stablecoins but to make the implicit costs explicit. Real yield equals headline APY minus expected loss from these tail risks. A holder cannot evaluate the trade without naming the risks.
​

Six risk categories cover the meaningful exposures. Most tokens face all six in some form; the relative weight differs by yield engine.

Each risk is enumerated below with named historical examples and the structural reason it applies.
​

Stablecoins target $1.00 (USDC, USDM) or a rising redemption value (USDY, sUSDS, sUSDe). A depeg is the token trading below the target. The size and duration of past depegs are useful priors.
​

Circle disclosed $3.3 billion of USDC reserves were held at Silicon Valley Bank when the FDIC took the bank into receivership. USDC traded as low as $0.87 over the weekend. The federal government's Sunday-evening backstop announcement returned the peg to $1.00 by Monday. The episode demonstrated that a 92% Treasury-backed token can lose 13% of its peg in 60 hours when one custodian is exposed (Circle's post-mortem).
​

Funding rates on Binance and Bybit perp markets inverted for nine consecutive days as the August 5 deleveraging cascade unwound long crypto exposure. Ethena's basis trade, which normally harvests positive funding, instead paid funding. USDe traded as low as $0.965 on secondary markets. sUSDe's redemption queue lengthened from minutes to days. The protocol's insurance fund absorbed the shortfall and the peg recovered within two weeks. A multi-quarter inversion would have produced a more severe depeg.
​

DAI has held within ~50 bps of $1.00 throughout its life, with two notable exceptions: March 12, 2020 (Black Thursday cascading liquidations to $0.95), and the March 2023 USDC depeg, when DAI's USDC-collateralized exposure pulled DAI to ~$0.93. The second event was specifically a contagion event: DAI held USDC, USDC depegged, DAI followed.
​

Liquidity risk is the cost of exiting a position at scale during stress. Headline secondary-market depth understates this cost: under stress, market makers widen spreads or pull entirely.
​

Tokenized-Treasury stablecoins generally redeem T+1 to T+2 through banking rails. USDY accepts redemptions from $100K up. USDM also has a $100K minimum. BUIDL minimum is $5M. The redemption queue is orderly under normal conditions; under stress (a Friday-evening event), settlement can stretch to T+3.
​

Curve, Uniswap, and DEX aggregators provide secondary liquidity. Pool depths are published in real time:

A holder needing to exit $50M of USDM during a stress window would either wait for primary redemption (T+1, $50M cleared in 1-2 days) or split across DEX pools and accept 30-100 bps slippage. Neither is catastrophic, but neither matches the same-day liquidity of a bank deposit at scale.
​

sUSDe imposes a 7-day cooldown for primary unstaking. During the August 2024 stress, secondary-market sUSDe traded at a 1-2% discount to the implied redemption value, reflecting the cooldown's opportunity cost during volatile windows.
​

Counterparty risk is a custodian, exchange, or RWA borrower failing in a way that affects the token's redemption ability or its underlying value.
​

USDM custodies Treasuries at BNY Mellon. USDY uses Morgan Stanley. BUIDL uses Bank of New York Mellon. A custodian failure would freeze redemption while the regulated process resolves the assets. BNY Mellon and Morgan Stanley are systemically important; their failure is a tail event but not zero.
​

USDe's short perpetual legs sit on Binance, Bybit, OKX, Deribit. The FTX collapse in November 2022 demonstrated that even large exchanges can fail in days. Ethena uses off-exchange settlement (Copper, Ceffu, Fireblocks) where possible to limit on-exchange asset custody, but the short positions themselves are inside exchange systems. An exchange freeze of Ethena's accounts would leave positions unhedged.
​

sUSDS holds 22% of reserves in tokenized real-world credit through partners like BlockTower and Centrifuge. Defaults are infrequent but not zero. Sky's RWA exposures are senior, short-dated, and overcollateralized in most cases, but the loss-given-default is not zero.
​

The token issuer itself can fail operationally. Mountain Protocol's Bermuda license is the foundation of its regulatory posture; a license revocation or operational failure of the company would freeze USDM redemption. Resolv's protocol-level failure would affect USR and RLP holders. The risk is small for established issuers but is non-zero.
​

Every token adds a contract surface. Stacked use cases multiply attack surface.
​

A holder of plain USDM takes Mountain's contract risk. A holder of plain USDC takes Circle's. The contracts have been audited (multiple firms) and have not been exploited as of April 2026.
​

sUSDe wraps USDe; sUSDS wraps USDS. Each wrapper adds a contract. The wrapper holds the underlying and exposes the holder to whatever bug exists in either layer.
​

A holder depositing sUSDe into a Morpho vault that lends against sUSDe collateral exposes themselves to: Ethena's contract, the sUSDe wrapper, Morpho's contract, the curator's vault contract, and any oracle the system uses to price sUSDe. Five layers of contract risk.
​

The Rekt News leaderboard (rekt.news) catalogs onchain exploits exceeding $1 million. Stablecoin-related categories include: oracle manipulation (Mango Markets, $114M, October 2022), reentrancy in lending markets, and bridge bugs (Wormhole, $326M, February 2022). Yield-bearing-stablecoin protocols have not had a top-tier exploit through April 2026, but the surface area is non-trivial.
​

The U.S. regulatory landscape for yield-bearing stablecoins is the largest single tail risk to the category.
​

The pending U.S. stablecoin legislation distinguishes payment stablecoins (which cannot pay interest to holders) from yield-bearing tokens (which become securities under SEC oversight). Under the current draft, USDC and USDT remain payment stablecoins, while USDY, USDM, sUSDS, and sUSDe become securities. The practical implications: registered offering exemptions, distribution restrictions, and qualified-purchaser limits.
​

USDY is non-U.S.-only. BUIDL is qualified-purchaser-only. USDM is broadly accessible from a Bermuda license. USDe restricts U.S. minting. The patchwork is administratively complex and creates the possibility that secondary-market acquisition by U.S. persons becomes legally precarious.
​

The SEC has signaled scrutiny of yield-distributing tokens through public statements and enforcement actions against tangentially related products (BlockFi, Voyager, Celsius — all centralized yield products that the SEC found to be unregistered securities). A direct enforcement action against a major yield-bearing stablecoin issuer would compress demand sharply.
​

The Markets in Crypto-Assets regulation, fully effective in late 2024, imposes specific requirements on stablecoin issuers serving EU residents. Tokenized-Treasury issuers like Ondo and Mountain have published compliance roadmaps; some smaller protocols have geo-blocked EU access pending regulatory clarity.
​

Each engine has a specific failure mode beyond the generic risks above.
​

Yield drops with the Fed funds rate. The September 2024 cut took USDM from 5.30% to 5.00% over 90 days. Direct rate cuts are not "risk" in the loss-of-principal sense; they are floor risk on the headline APY.
​

Aave, Morpho, and the Sky savings rate depend on borrower demand. A collapse in leveraged demand can push deposit rates near zero. The 2018-2019 bear market saw onchain lending rates compress to under 1%.
​

Funding rates can invert. The August 2024 inversion compressed sUSDe APY from 19% to 4% over 11 days. A multi-quarter bear market would force protocols to either reduce headline rates, dip into reserves, or restructure collateral.
​

A pool advertising 12% APY where 8% is the protocol's own token is exposed to that token's price. A 50% drop in the token cuts the effective yield to 8%.
​

The honest framing is: real yield = headline APY − expected annualized loss. Quantifying expected loss is hard because the tail events are infrequent. A reasonable framework:

The expected-loss numbers are estimates, not contractual. A holder's actual loss in any given year is path-dependent. The narrowing of the spread between headline and real yield as risk decreases is intuitive: T-bill-backed tokens give up little to risk; synthetic-dollar tokens give up more.
​

One of the underappreciated risks of yield-bearing stablecoins is operational. A team holding USDM on Ethereum that needs to settle a payment in USDC on Solana must manually unwind, bridge, and swap. Each step introduces slippage, settlement-time risk, and additional smart-contract surface. The unwinding process during a stress event can take days.
​

Eco Routes orchestrates the cross-chain settlement. A team holds yield-bearing stablecoins on the chains where they earn most natively, submits an intent to settle in any stablecoin on any of 15 supported chains, and Routes selects the best path. Solvers compete on price, finality, and route. The team retains yield-bearing positions and avoids the manual unwinding that would otherwise compress the realized yield. See stablecoin treasury APIs compared and stablecoin automation platforms for related context.
​

For T-bill-backed tokens (USDM, USDY): custodian failure. For lending-backed tokens (sUSDS): RWA defaults and governance changes. For synthetic-dollar tokens (sUSDe, USR): perpetual-funding inversions and exchange counterparty failures. For all of them: U.S. regulatory action against yield-distributing tokens. Read the digital dollars explainer for context.
​

Yes. USDe traded as low as $0.965 during the August 2024 funding inversion. DAI traded near $0.93 during the March 2023 USDC contagion. USDC itself depegged 13% during the SVB weekend. Tokenized-Treasury stablecoins like USDM and USDY have not had material depegs to date.
​

Real. T-bill engines pay from actual Treasury interest. Lending engines pay from real borrower demand. Basis-trade engines pay from real perpetual-funding flows. Some early protocols have subsidized yield with their governance token, which is not real; the four major engines (T-bills, lending, basis, RWA) are all economically real.
​

Yes. T-bill yield drops with the Fed funds rate. Lending demand can collapse. Funding rates can invert. Each engine has a specific zero-condition. The probability of all engines simultaneously zeroing is small; protocols can blend engines to smooth.
​

Depends on the structure. T-bill tokens with bankruptcy-remote LLCs (USDY, USDM) protect holders' claims through the legal entity even if the operating company fails. Onchain protocols (sUSDS, USR) have no entity-level isolation; recovery depends on the protocol's contract logic and any governance backstop.
​

Headline APY minus expected annualized loss from the relevant tail risks. T-bill-backed tokens have ~10 bps expected loss; synthetic-dollar tokens have ~125-250 bps. Real yield narrows the headline spread but does not eliminate the relative ranking. See automation platforms for related strategies.
​