Stablecoin Breakdown 2026 — USDC vs USDT vs DAI vs FRAX

Most people treat stablecoins like cash. They aren't. They're IOUs of varying quality, dressed up to look identical, and the gap between the safest one and the riskiest one is the gap between a federally insured deposit and a leveraged loan against thin air. This piece walks through the four stablecoins that hold ~95% of the market in 2026, what actually backs each one, and how to size a stablecoin allocation without ending up like a UST holder in May 2022.

The 30-second version

| Stablecoin | Issuer | Backing | Audit cadence | Risk tier | |---|---|---|---|---| | USDC | Circle (US-regulated) | Cash + short-term US Treasuries | Monthly attestation, quarterly | Lowest | | USDT | Tether (Hong Kong / El Salvador) | Mixed: cash, Treasuries, secured loans, gold, BTC | Quarterly attestation | Low–medium | | DAI | MakerDAO (decentralised) | Crypto collateral + USDC + RWAs | On-chain real time | Medium | | FRAX | Frax Finance | Algorithmic + collateral hybrid | On-chain real time | Medium–high |

That's the headline. The detail is where it matters.

USDC — the cleanest peg

USDC is issued by Circle, a US fintech that operates under New York's BitLicense, a Bermuda banking licence, and as of 2024 is regulated under MiCA in the EU. Each USDC in circulation is backed roughly 1:1 by cash held at US banks plus short-duration US Treasury bills managed by BlackRock.

Circle publishes a monthly attestation from a Big Four auditor (currently Deloitte) and a full quarterly financial report. The reserve mix is public, refreshed weekly, and the dollar value of reserves has stayed within 0.5% of supply for every reporting period since 2022.

That doesn't make USDC bulletproof. The March 2023 Silicon Valley Bank collapse stranded $3.3 billion of Circle's reserves for 48 hours, and USDC briefly traded down to $0.87. The peg recovered within three days once FDIC stepped in. The lesson wasn't that USDC is risky — it's that even the cleanest stablecoin carries banking-system risk and can break peg temporarily when the underlying banks wobble.

Best for: treasury management, lending collateral, long-term holdings, anyone who needs the regulator-friendly answer.

Watch for: counterparty concentration with US banks. If you're holding more than $250,000 worth, you're past the FDIC threshold of any single bank Circle uses.

USDT — the largest, most opaque, most useful

Tether is the original stablecoin and still the biggest by a wide margin — circulating supply north of $110 billion in 2026. It's also the one with the most uncomfortable history.

Tether claims 1:1 backing with reserves, and its quarterly attestations (now done by BDO) show the mix as roughly: 70% US Treasury bills, 15% cash and equivalents, 6% bitcoin, 5% secured loans, 3% gold, and a small bucket of "other investments". Tether earns north of $5 billion per year on those Treasury holdings, which is the real reason it can survive scrutiny that would have killed earlier stablecoins.

The problem isn't really the reserve quality today. It's:

1. No full audit, only attestations. An attestation says "the reserves are there on this date." A full audit verifies the entire flow of money in and out. The difference matters when stress comes.
2. History of misstatements. In 2021, Tether paid a $41 million CFTC settlement for misleading reserve disclosures between 2016 and 2019. The current reserve mix is much cleaner. The disclosure history isn't.
3. Jurisdictional opacity. Tether is incorporated in the British Virgin Islands and operationally based in Hong Kong with a treasury function in El Salvador. None of those jurisdictions can compel the kind of disclosure US regulators can.

USDT depegs more often than USDC. In normal markets it trades $0.998–$1.002. In a panic, $0.93–$0.97 isn't unusual. It always recovers because the redemption mechanism through Tether still works.

Best for: trading on non-US exchanges, fast on/off-ramp in emerging markets, anywhere USDT is the deepest liquidity pool.

Watch for: any sign of regulatory action in the US, where the FIT21 / GENIUS Act framework around stablecoins is tightening rapidly.

DAI — the survived-the-decentralised-thesis

DAI was supposed to be the answer to the centralisation problem with USDC and USDT. It's issued by MakerDAO (now rebranded Sky in 2024 but the DAI token still circulates), governed on-chain by MKR token holders, and backed by a basket of collateral types: ETH, wrapped BTC, USDC, real-world asset (RWA) positions in tokenised Treasury funds, and a few smaller positions.

The collateral mix has shifted dramatically over the years. In 2020 DAI was 80% ETH-backed. By 2024 it was over 60% backed by USDC and RWA, which makes the "decentralisation" thesis weaker than it sounds. The reality is that DAI is a wrapper for a portfolio of mostly-centralised assets, with the wrapper itself being decentralised governance.

DAI's peg holds via two mechanisms:

• Collateralised debt positions (CDPs). Anyone can mint DAI by locking 150%+ of its value in approved collateral. If the collateral drops below the liquidation ratio, the position is auto-liquidated.
• Peg Stability Module (PSM). A direct 1:1 swap between USDC and DAI with a tiny fee. This is what actually keeps the peg tight day-to-day.

DAI's peg is actually slightly more stable than USDT in normal conditions because the PSM keeps it pinned to USDC. The flip side: if USDC depegs, DAI usually depegs with it.

Best for: DeFi protocols where you want an on-chain settlement asset, MakerDAO ecosystem yield, anyone allergic to direct CEX-stablecoin exposure.

Watch for: the USDC dependency. DAI is not as independent of centralised infrastructure as marketing copy suggests.

FRAX — the algorithmic survivor

FRAX is the last meaningful algorithmic stablecoin standing after the UST disaster. Originally it was a fractional-algorithmic design: each FRAX was partly backed by collateral and partly by the algorithmic FXS token. Following community votes in 2023, FRAX moved to 100% collateral backing (USDC, on-chain, plus AMO-managed positions), so the "algorithmic" framing is mostly historical.

In 2026 FRAX behaves like a slightly more complex DAI: collateralised, on-chain, with the peg maintained through arbitrage and the FraxBP (Frax Base Pool) liquidity on Curve. The peg is good — usually within 0.2% — but FRAX has 1/40th the liquidity of USDC, so getting out of large positions costs more.

Best for: sophisticated DeFi users who want exposure to the FRAX ecosystem (sfrxETH staking, Fraxswap).

Watch for: any move back toward partially-algorithmic backing. Read every governance proposal. The whole "we're 100% backed" story can change quickly under stress.

The depeg history matters more than the reserve report

Stablecoin marketing pages all show ratios. They all look fine in calm markets. The thing that actually separates them is what happens under stress.

| Event | USDC | USDT | DAI | FRAX | |---|---|---|---|---| | Mar 2020 (COVID) | -3% | -1% | -8% | n/a | | May 2021 (China FUD) | -0.5% | -3% | -2% | n/a | | May 2022 (UST collapse) | -0.3% | -5% | -1% | -2% | | Mar 2023 (SVB) | -13% | -0.5% | -7% | -10% | | Aug 2023 (Curve hack) | -1% | -0.3% | -2% | -4% |

The pattern: USDC depegs hard when US banks wobble, USDT depegs hard when crypto-specific fear runs, and the smaller stables depeg the most because their liquidity is thinnest.

There's no "safe" answer. There's a "safe for your specific risk" answer.

The decision framework

For a stablecoin allocation in 2026:

If you need maximum stability and live in the US/EU: USDC. Split across two exchange custody accounts plus self-custody if the position is above $50k.

If you're trading in emerging markets, on Binance, Bybit, or any non-US exchange: USDT. It's the deepest liquidity pool and the on/off ramp is wider. Accept the depeg risk; size accordingly.

If you're earning DeFi yield: DAI or USDC. Both are accepted across all major lending protocols; pick the one with better rates on the chain you use.

If you want algorithmic exposure: small allocation only. FRAX is the only one that's earned the right to be in the conversation. Even then, keep it below 5% of stablecoin holdings.

For anyone: don't park more than 25% of liquid net worth in any single stablecoin, ever. Cross-collateralise across at least USDC + USDT if you're holding meaningful amounts. The cost of doing this is essentially zero. The benefit on the day someone depegs is enormous.

You can model what a 1% or 5% depeg costs on your specific position with our crypto fee calculator — same math applies to a depeg-cost as to a trading fee, since you're paying a haircut to exit. And for tax tracking on stablecoin moves, the crypto profit calculator handles the cost basis cleanly.

The most expensive mistakes

1. Treating stablecoins as cash for tax purposes. A swap from USDC to USDT is a taxable disposal in most jurisdictions. The dollar amount might be the same; the tax event isn't.

2. Bridging stablecoins. USDC on Ethereum is real USDC. USDC.e on some L2s is a wrapped version. Different smart contract, different risk profile, and historically different depeg behaviour. Read the asset name carefully.

3. Concentrating in one stablecoin because the yield is 0.5% higher. A 12% yield on UST in early 2022 was 100% higher than competitors. It became 0% over a weekend. The risk-adjusted return on slightly-higher stablecoin yield is almost never worth it.

4. Confusing audit and attestation. Audit = full verification. Attestation = a snapshot. Most stablecoins only have attestations. Adjust your trust accordingly.

5. Holding stablecoins on the exchange that issued them. Coinbase + USDC, Binance + BUSD (RIP). When the exchange wobbles, the stablecoin wobbles with it because withdrawals dry up first.

The shortcut

If you want a one-line rule: hold USDC for stability, USDT for trading, split across two custodians, and never trust a yield above 8% on anything claiming to be a stablecoin.

Stablecoins solved one problem (volatility) and created a different one (counterparty risk in a market that prefers not to talk about it). Most of the time, the system holds. The exceptions are predictable in shape and unpredictable in timing. Plan for the shape.