Every stablecoin yield strategy has a hidden second position: the peg itself. When you deposit USDC into Aave, you earn lending yield — but you're also long the promise that USDC redeems at $1.00 forever. Most of the time that promise holds so reliably you forget it's there. Then a bank fails, an attestation comes out wrong, or an algorithmic design meets a death spiral, and "stable" becomes the most ironic word in crypto.
The stakes are bigger than ever. The dollar stablecoin market has grown to roughly $300 billion in 2026 — up 30% in about a year, with the passage of the GENIUS Act in July 2025 marking its arrival as regulated financial infrastructure. Scale cuts both ways: more institutional reserve quality, but also a larger system to run on when confidence wobbles.
TL;DR. Stablecoins depeg for four reasons: reserve custodian failure (USDC/$0.87, March 2023), inadequate or fraudulent reserves (TUSD, UST), liquidity runs, and collateral spirals in crypto-backed designs. The GENIUS Act — effective January 2027 — forces reserve and redemption standards on "payment stablecoins" but doesn't cover algorithmic tokens and can't erase bank risk. Practical rules: diversify across issuers, prefer coins with frequent attestations and fast redemption, and treat any stablecoin APY far above the ~3.6–4% core lending rate as compensation for peg risk you're newly taking.
What actually keeps a stablecoin at $1.00
Three mechanisms hold pegs, and each one breaks differently:
- Fiat-backed (USDC, USDT, PYUSD): the issuer holds dollars or short-term Treasuries and promises 1:1 redemption. The peg lives or dies with reserve quality and access to redemption. Even pristine reserves fail if they sit in a bank that suddenly can't return your call.
- Crypto-overcollateralized (DAI/USDS): smart contracts lock more crypto than the stablecoins issued, liquidating collateral when prices fall. Breaks when collateral crashes faster than liquidations execute, or when the oracles feeding those liquidations fail.
- Algorithmic/synthetic (UST — RIP; USDe-style synthetic dollars): peg maintained by arbitrage, often against a hedged delta position. Breaks when the arbitrage reverses under stress — funding rates can go negative, and the "dollar" is only as good as the hedge.
The common thread: a stablecoin holds $1.00 because people believe they can exit at $1.00. Depeg events are belief-discovery moments.
Four depegs every yield farmer should know
| Event | Low | Cause | Outcome |
|---|---|---|---|
| UST (May 2022) | $0.02 | Algorithmic death spiral; no reserves | ~$40B destroyed; never recovered |
| USDC (Mar 2023) | $0.87 | $3.3B (8%) of reserves stuck at failed SVB | Recovered in ~48h after US backstop |
| USDT (May 2022) | $0.95–0.97 | Contagion panic during UST collapse | Recovered within days |
| TUSD (2023–26) | Redemption broken | Reserves misreported; ~$456M trapped in illiquid funds; SEC fraud settlement (2024) | Issuer bankruptcy (Feb 2026); redemption still uncertain |
The USDC case is the most instructive because nothing was "wrong" with the coin. Circle's reserves were real. But 8% of them sat in Silicon Valley Bank, and when the FDIC seized SVB on March 10, 2023, secondary markets repriced that uncertainty within hours — USDC hit $0.87 before the US government announced depositors would be made whole. Notably, Federal Reserve researchers documented that redemptions continued for weeks after the backstop: once trust cracks, it heals slower than prices do.
The TUSD case is the 2026-relevant cautionary tale. It looked like a normal fiat-backed coin, but attestations paused, the SEC's settlement revealed reserves had drifted into speculative offshore funds, and holders are still waiting for clean redemption after the issuer's bankruptcy. The lesson isn't "fiat-backed is unsafe" — it's that the quality of the attestation and the redemption mechanism matters more than the label.
Reserve quality tier list (2026)
Not all dollar tokens carry the same peg risk. A practical tiering based on reserve composition, attestation frequency and redemption track record:
| Tier | Coins | Profile | Residual risk |
|---|---|---|---|
| 1 — Regulated fiat | USDC, PYUSD | Mostly cash + short Treasuries; monthly attestations; proven crisis recovery | Bank custodian risk; regulatory action |
| 2 — Scale fiat | USDT | $100B+ Treasury holdings and unmatched liquidity, but ~24% of reserves in non-traditional assets per its latest attestation | Reserve opacity; jurisdiction risk |
| 3 — Crypto-backed / synthetic | DAI/USDS, USDe | On-chain collateral or hedged basis trade; transparent mechanics | Oracle failure; collateral crash; negative funding |
| 4 — Long tail | Small/new issuers | Thin liquidity, limited history, weaker attestations | Everything above, multiplied — treat as equity-like |
Tier 1 isn't "safe" — it's "survived a real test." USDC's own reserve fund shortened its average maturity after SVB precisely because bank-run risk got priced in. That's the system working, not risk disappearing.
What the GENIUS Act changes — and what it doesn't
The GENIUS Act, enacted July 18, 2025, creates the first US federal framework for "payment stablecoins": who may issue them, what reserves qualify, and what reporting and redemption standards apply, with the regime effective January 2027. Three practical consequences:
- Reserve standards get teeth. Issuers must back payment stablecoins with high-quality liquid assets and meet diversification rules — regulators are still writing the implementing regulations, which is where the details will live.
- Issuers can't pay interest. The statute bars payment stablecoin issuers from paying yield directly. In practice, third-party platforms keep offering "rewards" — which means the yield you chase in DeFi still comes from lending markets and incentives, not the issuer. That's a feature: it keeps the peg mechanism separate from the yield product.
- Coverage is limited. Algorithmic tokens, tokenized money-market funds and other yield-bearing instruments sit outside the framework. A "stablecoin" with a 15% APY attached is very likely not a GENIUS-Act payment stablecoin at all.
What it doesn't fix: the depeg mechanism that actually bit USDC was bank custody — and bank failures remain possible. Regulation raises the floor; it doesn't abolish the tail.
A depeg-risk checklist for stablecoin yield farmers
Before depositing into any stablecoin yield venue, run this five-point check:
- 1. Diversify across issuers, not just protocols. Splitting USDC across three lending protocols does nothing if your worry is Circle. Split across issuers (USDC/USDT/USDS) and mechanisms (fiat + crypto-backed).
- 2. Know your exit before you enter. Can you redeem 1:1 directly with the issuer, or must you sell on secondary markets? How deep is that liquidity at 3 a.m. during a panic? Tier-1 coins clear this; most long-tail coins fail it.
- 3. Read the yield source. Core stablecoin lending paid roughly 3.6–3.9% across major venues in September 2026. If a "stable" pool promises 10%+, the excess is emissions, incentives or mechanism risk — and often the peg itself is the collateral. (How to separate real yield from emissions.)
- 4. Watch for early warning signs. Sustained deviation beyond ±0.5%, delayed attestations, redemption queues, or a stablecoin trading at a discount on one venue but not others. These are your cheap exit windows.
- 5. Size for the tail. If a coin fully depegs, what do you lose? Keep any single issuer below the level where that loss ruins your year. A 4% APY never justifies a 100% principal loss — UST taught exactly one lesson, expensively.
Sources and further reading
- Federal Reserve Bank of New York — Stablecoins and (Non)Crypto Shocks: market size, USDC reserve composition after SVB.
- Federal Reserve Bank of Chicago — Stablecoins under the GENIUS Act: scope, reserves, effective dates.
- Federal Reserve Bank of Philadelphia — Stablecoin growth and run risk in historical context.
- rwa.xyz — live tokenized Treasury and stablecoin reserve data.
Frequently asked questions
What causes a stablecoin to depeg?
Four main causes: the bank holding fiat reserves fails (USDC/$0.87 during the SVB collapse), reserves are inadequate or fraudulent (TUSD, UST), redemptions exceed available liquidity during panic, or — for crypto-backed and algorithmic designs — collateral crashes faster than liquidations can respond.
Has USDT ever depegged?
Yes, briefly — it traded as low as $0.95–0.97 during the May 2022 Terra collapse and recovered within days. It has never broken structurally, but the episode showed even the deepest-liquidity stablecoins wobble under systemic stress.
Does the GENIUS Act make stablecoins safe?
It reduces but doesn't eliminate risk. Enacted July 2025 and effective January 2027, it imposes reserve and redemption standards on payment stablecoins. It doesn't cover algorithmic tokens, bars issuers from paying interest directly, and can't abolish bank-custodian risk — the exact mechanism behind the 2023 USDC depeg.
Which stablecoin has the lowest depeg risk?
Large regulated fiat-backed coins — USDC and PYUSD currently score best on reserve quality, attestation frequency and redemption. USDT offers unmatched liquidity but holds roughly a quarter of reserves in non-traditional assets. Diversification across issuers beats picking one winner.
Should I chase the highest stablecoin APY?
Treat excess yield as payment for extra risk. Core lending paid ~3.6–4% in September 2026; anything far above that comes from emissions, incentives or newer mechanisms — and often makes the peg itself part of your bet.
Rank stablecoin yields by risk, not hype
Live USDC, USDT, DAI and USDe APYs across blue-chip venues — with depeg-risk context.
Open the APY TrackerMore guides in the DifiCalc blog, or read Best Stablecoin Yield 2026, Tokenized Treasuries vs DeFi Lending and DeFi Yield Traps: 10 Red Flags.