Yield-bearing stablecoins look like cash and pay like a money-market fund, which is exactly the framing that gets holders in trouble. The wrapper rises in value steadily, the chart never blinks, and the position feels closer to a savings account than to a leveraged credit trade. In practice every yield-bearing stablecoin carries a stack of distinct risks, and the stacks differ enough between sUSDe, sUSDS, sDAI, and aUSDC that they should be evaluated as separate instruments rather than as variants of one product.
​

This article walks through five risk categories that apply to every yield-bearing stablecoin in some combination: depeg risk (anchored by the UST collapse of May 2022), smart contract risk, counterparty risk, yield-source sustainability, and regulatory risk under the GENIUS Act and parallel frameworks. Each section names the mechanism, the historical case study where relevant, and which wrappers carry the most exposure. A comparison table at the end maps the four largest yield-bearing wrappers to their primary risk vectors.
​

The point is not to argue against holding these instruments. Many DeFi treasuries hold sUSDe, sUSDS, sDAI, and aUSDC continuously and have done well. The point is to make sure the risk taxonomy is explicit before sizing the position, because the failure modes are real and the recoveries from past failures have ranged from total loss to full restoration.
​

A yield-bearing stablecoin is a token that targets a one-dollar peg and accrues yield to holders, either via a rising exchange rate (sUSDe, sUSDS, sDAI) or via a rebasing balance (aUSDC, aUSDT in Aave). The yield comes from one of several sources: T-bill interest passed through, lending demand on the underlying asset, perpetual funding rates captured via a delta-neutral hedge, or savings-rate distributions set by protocol governance.
​

The wrappers covered here are the four largest by supply as of Q2 2026: sUSDe from Ethena, sUSDS from Sky, sDAI from Sky's legacy DAI track, and aUSDC from Aave V3. Together they hold the majority of yield-bearing stablecoin TVL per DeFiLlama. Each one earns yield differently, so the risk profile differs even though the user experience (deposit a stable, watch the balance rise) feels similar.
​

One thing that does not count, for this article, is a plain fiat-backed stablecoin like USDC or USDT held idle. Those do not pay yield to holders. They sit on a balance sheet earning T-bill yield for the issuer, with none of that yield reaching the token holder. The risk profile of plain USDC is distinct and is covered in our reserves write-ups, not here.
​

Depeg risk is the possibility that the wrapper trades meaningfully below one dollar and either recovers slowly or does not recover at all. The largest stablecoin depeg in crypto history is the May 2022 collapse of TerraUSD (UST), an algorithmic stablecoin that briefly held more than $18B in supply before falling to a few cents and never recovering. The mechanism was a feedback loop between UST and its sister token LUNA: minting one required burning the other, and as confidence cracked the burn-mint loop accelerated the depeg rather than absorbing it.
​

Reuters reported on May 13, 2022 that UST had fallen to around 11 cents and LUNA had collapsed roughly 99% in a single week, wiping out an estimated $40B in market value across the Terra ecosystem per Reuters coverage. The SEC subsequently charged Terraform Labs and its founder Do Kwon with securities fraud, with the February 2023 SEC complaint documenting that the Anchor Protocol yield (around 20% on UST deposits) had been subsidized rather than earned, which is the proximate reason confidence cracked once the subsidy ran out.
​

UST is the worst-case scenario. None of the wrappers covered here use the same algorithmic mint-burn loop. Still, the case study is the right starting point because it teaches three lessons that apply to every yield-bearing stablecoin: yield that exceeds risk-free rates by a large margin usually reflects a subsidy or a risk that has not yet shown up; depegs can be reflexive rather than gradual, so the depeg compounds itself once it starts; and recovery is not guaranteed, even with a well-funded ecosystem behind the token.
​

Among current yield-bearing stablecoins, depeg risk shows up in different shapes. sUSDe depegs would most likely come from a stressed unwind of the perpetual hedge during a gap move or exchange failure. sUSDS and sDAI depegs would come from severe stress in the Sky CDP collateral, particularly if a large vault liquidates into thin onchain liquidity. aUSDC depegs would track USDC itself, which briefly dropped to roughly 87 cents during the Silicon Valley Bank weekend in March 2023 before recovering once Circle confirmed access to its reserves.
​

Smart contract risk is the possibility that a bug, exploit, or governance attack drains funds from the staking contract, the underlying lending market, or a dependency the wrapper relies on. Every major yield-bearing stablecoin has been audited multiple times by reputable firms. Audits reduce but do not eliminate the risk; they are point-in-time reviews of a static codebase and cannot anticipate every interaction with future deployments.
​

Ethena lists audits from Quantstamp, Pashov, Cantina, and Spearbit on its audit page, covering the minting, staking, and accounting contracts. Sky's USDS and sUSDS contracts inherit the audit history of the underlying MakerDAO codebase plus the new Sky governance modules. Aave V3 has been audited by OpenZeppelin, Trail of Bits, ABDK, and others, with all reports linked from the Aave V3 repo. Pendle, which many sUSDe holders touch via PT and YT positions, lists audits from Ackee, Spearbit, and others.
​

The limits of audits matter for two reasons. First, the wrappers compose. A holder of sUSDe deposited as Aave V3 collateral is exposed to bugs in Ethena's staking contract, in Aave's collateral logic, in the oracle that prices sUSDe, and in any liquidation module that may need to seize the position. A bug in any one layer can cascade. Second, governance can change code. Sky governance can change SSR parameters, can change collateral types, and can change the supply caps on individual vaults. Each change is a new surface that the original audits did not see.
​

The practical takeaway is that audit count is a weak signal of safety past a threshold. Three good audits and a bug bounty program is meaningfully better than zero. Ten audits is not meaningfully better than five for any wrapper that has been in production for more than a year, because the time-tested production exposure is doing most of the work. Holders should care more about whether the protocol has paid out bugs, whether it has a working bug bounty (Immunefi or equivalent), and whether the production codebase is in continuous use by other large protocols that would have caught issues.
​

Counterparty risk is the possibility that an entity holding assets on behalf of the protocol fails, defaults, or freezes the position. For yield-bearing stablecoins, the relevant counterparties differ by wrapper. sUSDe carries counterparty risk on the centralized perpetual exchanges where the hedge sits (Binance, Bybit, OKX, Deribit) and on the Off-Exchange Settlement custodians (Copper, Ceffu) that hold the bulk of collateral. sUSDS and sDAI carry counterparty risk on the Sky CDP vault keepers and on any centralized collateral the Sky Stability Module may use. aUSDC carries counterparty risk on Aave's liquidation venues and, indirectly, on Circle as the issuer of the underlying USDC.
​

Tokenized T-bill wrappers (OUSG, BUIDL, USDY) are not the focus of this article but carry the cleanest version of counterparty risk: the T-bill custodian. BlackRock holds the BUIDL fund's T-bills with BNY Mellon. Ondo uses StoneX and Clear Street for OUSG. Franklin Templeton holds BENJI's T-bills directly. A failure of any of these custodians would freeze redemptions even if the underlying T-bills are intact.
​

The FTX collapse of November 2022 is the closest precedent for centralized-venue counterparty failure that would have hit a delta-neutral hedge directly. Ethena did not exist in November 2022, but several funds running similar basis trades did, and many of them had open hedges on FTX that became unsecured creditor claims. The lesson built into Ethena's design is the Off-Exchange Settlement architecture, which keeps the bulk of collateral off the exchange balance sheet so a venue failure stops at the working-margin layer rather than the full hedge.
​

For Sky's CDP-backed wrappers, the closest historical stress was the March 2020 "Black Thursday" event in MakerDAO, where ETH price gap-moved through liquidation thresholds and a small number of zero-bid liquidation auctions left DAI under-collateralized for a period. Sky's current design (formerly MakerDAO) has been hardened with multiple stability modules, larger auction parameters, and a more diverse collateral base, but the structural risk that liquidations may not clear at fair prices remains.
​

Yield-source sustainability is the risk that the mechanism producing yield slows or reverses. This is different from a permanent depeg or a smart contract drain. The wrapper continues to function; the yield just goes away or goes negative.
​

The clearest case is Ethena's sUSDe and the funding-rate flip scenario. Funding rates on perpetual futures are positive most of the time because long-biased speculators dominate crypto perps, but they can flip negative when sentiment turns bearish. When funding flips negative, Ethena's short legs pay rather than receive, and sUSDe APY drops toward zero or below for the period. The Ethena Reserve Fund is sized to absorb short windows of negative funding, but a sustained regime change (months) would erode it. Ethena documents this in its research posts, and the audit firms that reviewed the contracts have flagged it as the headline mechanism risk in their reports.
​

Sky's SSR and DSR carry a different sustainability profile. The rates are set by governance against the income from Sky's diversified collateral pool (T-bills, RWA tokens, USDC reserves, ETH-backed vaults). If yields on those underlying assets fall, governance has to choose between cutting SSR to maintain the surplus buffer or holding SSR steady and drawing the buffer down. In 2026 the SSR sits at 3.75%, lower than its 2024 peak above 8%, which reflects this dynamic. The risk is not that yield goes negative; it is that yield falls below what a holder needs to make the position worth the lockup-equivalent friction.
​

Aave's aUSDC carries sustainability risk tied to utilization. The supply APY on USDC in Aave V3 is a function of borrow demand: when borrowers want to borrow USDC, the supply rate rises; when demand collapses, the supply rate falls. There is no negative scenario here; the floor is zero. The risk is that in a quiet market regime the realized yield is much lower than the headline rates from a peak period suggested.
​

Tokenized T-bill wrappers track the Fed funds rate. The yield is mechanical: roughly the four-week T-bill rate minus fees. If the Fed cuts rates, the yield falls in tandem. There is no protocol-level sustainability risk; the risk is just the rate cycle.
​

Regulatory risk is the possibility that a wrapper becomes restricted, delisted, or reclassified in a way that affects access or value. The most concrete near-term example is the GENIUS Act, a US stablecoin regulatory framework that explicitly excludes yield-bearing stablecoins from the category of "payment stablecoins" available to retail users in the United States.
​

The version of the bill that passed the Senate in 2025 and is moving through House reconciliation defines a permitted payment stablecoin as one that does not pay interest or yield to holders. Yield-bearing instruments would either need to register as securities or as money-market funds (with all the associated disclosure and custodial requirements) or restrict US retail access. The practical effect for sUSDe, sUSDS, sDAI, and aUSDC is that US-facing platforms would either need to gate access or work with the issuers to restructure the wrapper.
​

Ethena has already published jurisdictional restrictions on its app, blocking access from the United States and several other jurisdictions. Sky has not formally restricted access to sUSDS in the US but has signaled it will comply with whatever final framework emerges. Aave's aUSDC is a different case: the yield comes from lending and would likely be treated as deposit-equivalent income, which has a different regulatory path than a structured product like sUSDe.
​

The European MiCA framework has already taken a similar position. Yield-bearing stablecoins are excluded from the e-money token category and must register separately, with most issuers opting out of EU access rather than pursue a parallel registration. The same outcome (geographic gating) is the most likely near-term result of the GENIUS Act in the US.
​

Holders who care about long-term US access should track the final language of the bill closely. Holders outside the US face a different question: will the issuer maintain global access, or will it shrink to a smaller set of jurisdictions to simplify compliance? The current direction of travel for most yield-bearing issuers is geographic narrowing, not expansion.
​

The table below summarizes the four largest yield-bearing wrappers across the five risk categories. The intent is a fast read of which wrapper carries which exposure, not a single safety score. Cross-check current parameters against DeFiLlama Yields and each protocol's official documentation before sizing a position.
​

sUSDe carries the highest concentration of distinct risks because its yield source is unique and its counterparties are centralized exchanges. sUSDS and sDAI share most of their risk surface, with sDAI on the lower-yield, longer-track-record end. aUSDC's risk is closer to a money-market position with smart contract overlay; the yield is more modest and the failure modes are the most well-understood.
​

For a side-by-side that focuses on yield mechanics rather than risk, see our sUSDe vs sUSDS vs sDAI comparison. For the specific mechanism behind sUSDe's funding-rate yield, see Ethena USDe and sUSDe explained.
​

Position sizing for yield-bearing stablecoins should start from the worst-case drawdown rather than the expected APY. Institutional desks running this analysis typically assume each risk category can produce a 10-30% drawdown in isolation and a larger drawdown if two correlated risks fire at the same time (e.g., a perp hedge stress combined with a funding flip, or a CDP collateral stress combined with an audit-discovered bug).
​

A reasonable rule of thumb for retail holders: do not hold more in any one yield-bearing wrapper than you would be comfortable losing entirely. The probability of total loss is low for the four wrappers covered here, but it is not zero, and the historical record (UST, Iron Finance, several smaller depegs) shows that yield-bearing stablecoins can go from one dollar to zero in days rather than months when the right combination of stresses arrives.
​

Diversification across wrappers helps because the risk vectors are genuinely different. Holding equal slices of sUSDe, sUSDS, and aUSDC reduces concentration risk in any single failure mode. It does not reduce the systemic risk that a major regulatory action or a broad DeFi exploit could affect multiple wrappers at once, but it materially reduces the idiosyncratic risk of any one.
​

There is no single safest wrapper. sDAI carries the longest production track record and the most administered yield, which favors stability over upside. aUSDC carries the most well-understood risk surface because Aave V3 has been in production with deep audit coverage for years. sUSDS is similar to sDAI with a higher administered rate. sUSDe carries the highest yield and the most distinct risk concentration. Match the wrapper to your risk tolerance rather than chasing a single ranking.
​

Yes. The UST collapse of May 2022 is the precedent: an $18B yield-bearing stablecoin fell to a few cents in a week and did not recover. None of the current major wrappers use the same algorithmic mechanism, but a sufficiently severe combination of stresses (large smart contract exploit, exchange failure, collateral collapse) could push any of them to zero. The probability is low but not negligible.
​

No. None of the wrappers covered here carry FDIC insurance or any equivalent government-backed deposit guarantee. Some tokenized T-bill wrappers (BUIDL, OUSG) hold T-bills that are backed by the full faith and credit of the US government, but the wrapper itself is not insured against issuer or custodian failure.
​

The GENIUS Act is a US stablecoin regulatory framework that defines "payment stablecoins" as instruments that do not pay yield or interest to holders. Yield-bearing wrappers like sUSDe, sUSDS, sDAI, and aUSDC fall outside that definition and would need to register as securities or money-market funds for US retail access, or restrict US access entirely. Most current issuers are leaning toward geographic restrictions rather than US registration.
​

UST was an algorithmic stablecoin paired with the LUNA governance token. Maintaining the peg required arbitrageurs to mint and burn UST and LUNA against each other. The Anchor Protocol paid roughly 20% yield on UST deposits, subsidized rather than earned. When confidence cracked in May 2022, redemptions of UST minted new LUNA, which crashed LUNA's price, which further accelerated redemptions in a reflexive loop. Reuters reported UST at roughly 11 cents and LUNA down 99% within a week, with the SEC later charging Terraform Labs and Do Kwon for securities fraud and misrepresentation of the subsidy.
​

Sources and methodology. UST collapse figures from Reuters and the SEC complaint against Terraform Labs. Audit references from Ethena audits, Aave V3 audits, and Sky's published audit history. Supply, APY, and TVL figures from DeFiLlama on May 24, 2026. Sky Savings Rate at 3.75% per Sky governance. Figures refresh continuously; cross-check before deploying capital.
​