The headline answer first

USDC is cleaner on paper. USDT is older and has survived more stress events. Both have broken peg. Neither has done so permanently. The question "which is safer" only has an honest answer once you specify safer against what, and the two tokens have asymmetric exposures that map to different risks. The desk holds both; we'll explain the split at the end.

Short version. If your primary risk is a single US bank failing, USDT is structurally less exposed. If your primary risk is the issuer disappearing one weekend without warning, USDC has the cleaner balance sheet and the SEC reporting obligation. Both risks are real. Hold both.

This piece is the spoke companion to our six-dimension USDT vs USDC report. The cornerstone covers everything — issuer, reserves, regulator, liquidity, cross-chain, failure history, twelve scenarios. This piece narrows the question to safety alone and gets into the comparisons headline writers usually skip.

What "safer" actually means here

The word does too much work in casual conversation. We mean five specific things, in order of how often the desk gets asked about them.

RISK 01

Permanent loss of peg

The token never returns to one dollar. The clearest example in the four-family canon is Luna's UST in May 2022, which lost peg and never recovered. Neither USDT nor USDC has experienced this. Both have come close once or twice. The base rate is low but not zero.

RISK 02

Reserve shortfall

The issuer cannot honour redemptions because reserves are smaller than circulating supply. The NYAG 2021 investigation into Tether is the closest thing the industry has to a documented case here, though it was resolved before reaching the limit. The risk applies to any fiat-backed stablecoin and depends on reserve composition and auditor relationship.

RISK 03

Counterparty failure

A bank or money-market fund holding part of the reserves fails. This is the risk that USDC took in March 2023 with Silicon Valley Bank. Resolution depends on the political appetite for a bailout. USDC's depositor base was politically attractive enough to get the FDIC's discretionary backstop; that outcome was not guaranteed.

RISK 04

Address freeze

The issuer freezes specific token addresses, usually at the request of US OFAC or for criminal-investigation cooperation. Both Tether and Circle do this. The risk to a normal holder is near zero. The risk to a sanctioned counterparty or a wallet that received funds traceable to sanctioned activity is real.

RISK 05

Regulator-forced wind-down

The token is delisted from major venues or wound down by regulator action. BUSD is the precedent: NY DFS ordered Paxos to stop minting in February 2023, and the token was effectively dead within a year despite no balance-sheet problem. The risk is jurisdiction-specific. MiCA has produced a softer version of this with USDT on EU venues since 2024.

How USDC and USDT score against each risk

USDT · the case

R1 permanent loss of peg. Not experienced in eight years across multiple severe stress events. Cumulative redemptions processed during 2022 Luna contagion: roughly 7 billion. Returned to peg every time.

R2 reserve shortfall. NYAG 2021 settlement disclosed historical reserve gaps. Since 2021, BDO-signed quarterly attestation has shown reserves at or above circulating supply. The attestation is point-in-time, not a financial audit.

R3 counterparty failure. Reserves diversified across cash equivalents (mostly short-dated US T-bills), corporate bonds, precious metals, digital assets and secured loans. No single US bank concentration above the rough thresholds Circle had with SVB.

R4 address freeze. Tether freezes addresses at law-enforcement request. Public count is in the high hundreds since 2017. Process is slower than Circle's.

R5 regulator wind-down. USDT has been progressively delisted from EU exchanges under MiCA since 2024. No US wind-down threat to date; Tether sits in BVI and El Salvador, outside US state perimeters.

USDC · the case

R1 permanent loss of peg. Not experienced in seven years. One severe depeg (March 2023) recovered within forty-eight hours.

R2 reserve shortfall. No documented case. Monthly attestation signed by Deloitte. The 2024 IPO added quarterly SEC filings, which is the strongest disclosure regime any stablecoin issuer currently sits inside.

R3 counterparty failure. Experienced in real time during the SVB weekend. 3.3 billion of USDC reserves were at SVB out of roughly 40 billion. Recovery was fast but depended on the FDIC's discretionary backstop, not a structural feature of USDC's design.

R4 address freeze. Circle freezes addresses faster than Tether, typically within hours of an OFAC update or court order. Process is more visible and better documented.

R5 regulator wind-down. Lower risk than USDT for US- and EU-domiciled holders because Circle is regulated where the volume sits. Higher exposure to any future US legislation that constrains stablecoin design directly (the 2025 GENIUS Act discussions, for example).

Reserve composition, read carefully

USDC reserves in 2026

Circle publishes a monthly attestation. The most recent file (April 2026, signed by Deloitte) breaks out:

  • Roughly 80% in the USDXX Circle Reserve Fund — a BlackRock-managed government money-market fund holding US Treasury bills and overnight repurchase agreements with the Federal Reserve. CUSIP-level holdings are published.
  • Roughly 20% in cash deposits at US-chartered banks. Bank names are disclosed in the SEC quarterly filings.

The composition is narrow and verifiable. It is also tightly coupled to two specific risks: a failure of one of the disclosed custodian banks, and any disruption to the overnight repo market or the BlackRock fund itself. The first risk realised in March 2023; the second has never realised in modern history but is not zero.

USDT reserves in 2026

Tether publishes a quarterly attestation. The Q1 2026 file (signed by BDO, dated 2026-04-30) shows:

  • Cash and cash equivalents at 84.1% of reserves. The bulk is short-dated US Treasury bills held directly or through repo arrangements. Tether disclosed roughly 100 billion of T-bill exposure on its 2025 year-end balance sheet, making it one of the larger non-sovereign holders of US Treasuries globally.
  • Corporate bonds and precious metals at roughly 8.5%.
  • Digital assets, secured loans and other investments at the remaining 7.4%.

The mix is more diversified than Circle's and harder to verify line by line. The "other investments" bucket has shrunk from a peak above 15% in 2021 but still contains instruments that are not redeemable for cash on demand. The trade-off is real: less single-counterparty risk in exchange for less transparent composition.

The SVB weekend, in plain English

The single piece of evidence that disproves "USDC is safer because it is regulated" is the March 2023 weekend. The short version, for readers who want it without the hour-by-hour reconstruction (which is in our standalone SVB piece):

  1. Silicon Valley Bank failed on Friday, March 10, 2023.
  2. Circle disclosed 3.3 billion of USDC reserves at SVB on Friday evening.
  3. USDC fell to 0.8774 on Coinbase on Saturday morning. In specific Curve pools the price briefly traded below 0.82.
  4. Circle paused minting and redemption over the weekend because US banking rails were closed.
  5. The Treasury, Fed and FDIC announced on Sunday evening that SVB depositors would be made whole, including the uninsured portion.
  6. USDC returned to peg by Monday's banking open.

Two lessons that the desk continues to bring up two years later. First: regulatory tightness pulls you closer to the regulated system. When that system has a wobble, you have a wobble. Second: the resolution required a discretionary government action. The Treasury chose to invoke the Bank Term Funding Programme. That choice was political. A different bank failure, with a different depositor base, could produce a different outcome.

Did USDT benefit?

It did. During the same weekend, USDT briefly traded slightly above peg on global venues as a flight asset. This is one of the better-documented inverse-correlation events in stablecoin history. Holders on Binance, OKX and other non-US venues who held both rebalanced toward USDT during the depeg and back during the recovery. The point is not that USDT is the safer choice; the point is that they fail in different ways and at different times, which is the textbook case for diversification.

The NYAG case, in plain English

The corresponding piece of evidence that disproves "USDT is structurally sound because nothing has ever happened" is the 2021 New York Attorney General settlement. The desk goes through it in more detail in the cornerstone; the headlines:

  • The NYAG concluded that Bitfinex (Tether's affiliated exchange) had a 850 million dollar shortfall at a payment processor in 2018-2019. Bitfinex used Tether funds to cover the shortfall. The disclosure to customers was inadequate.
  • Tether had at certain points overstated the relationship between USDT supply and dollar reserves.
  • The settlement imposed an 18.5 million dollar penalty and required quarterly reserve reports for at least two years.
  • The reports continued voluntarily after the requirement lapsed.

The settlement is the foundation of the public reserve transparency Tether has today. Before 2021, reserve disclosure was sporadic. The quarterly attestation people read in 2026 exists because of this enforcement action. It is also the answer to "has Tether ever been investigated?". Yes. By the New York AG. The result was the 2021 order and the disclosure regime that followed.

Comparing freezes

Both issuers freeze addresses. The numbers as of early 2026:

  • Tether: cumulative frozen address count above 1,400 since 2017. Process typically triggered by OFAC updates, US DoJ requests, or specific exchange security incidents.
  • Circle: cumulative frozen address count in the high hundreds since 2018. Process is faster, sometimes within minutes of an OFAC sanctions update.

For a normal retail holder this risk is theoretical. The exception worth knowing: if you receive funds from a counterparty whose source of funds was traceable to a later-sanctioned address, your address can be frozen along the trail. The standard guidance — withdraw to a fresh wallet that has not interacted with mixers or sanctioned protocols — applies to both tokens equally.

What the desk holds

For working balances on Binance, OKX and other global venues, the desk holds USDT. The order-book depth, the cross-chain coverage and the fee schedule all favour USDT for the operational use case. For DeFi positions on Ethereum and Base, the desk holds USDC. Pool depth and protocol defaults favour USDC for the yield use case. The split is roughly 60/40 USDT/USDC, weighted to where the working balance actually moves.

The cold-storage portion is split closer to 50/50. The reasoning is exactly the "they fail differently" point. A position concentrated entirely in either token is exposed to whichever risk that token carries; holding both reduces correlation without requiring a forecast of which crisis comes next.

Three myths the desk hears most

Myth 1 · "USDC is backed by US Treasuries, so it cannot break peg"

USDC is backed by Treasuries through the BlackRock money-market fund, plus cash deposits at US banks. The Treasuries themselves do not break. The cash deposits at the bank holding part of the reserves did break in March 2023. Backing by safe instruments does not make the token safe if the path between the instrument and the holder includes a fragile link.

Myth 2 · "USDT is unaudited, so the reserves do not exist"

Neither USDT nor USDC has a full annual financial audit at the standard a listed bank would file. Both publish attestations from accounting firms. Tether's quarterly BDO attestation has shown reserves above circulating supply every quarter since 2021 and the company has processed every redemption it has received over the same period. The absence of a full audit is a fair criticism. The claim that the reserves do not exist is not consistent with the redemption track record.

Myth 3 · "MiCA will kill USDT in Europe"

MiCA has constrained USDT trading on EU-licensed exchanges, with several venues phasing out USDT spot pairs from mid-2024 through 2026. The token still trades widely on non-EU venues accessible to EU residents, and the OTC market for USDT remains liquid in Europe. The regulation has changed where the volume sits, not whether it exists. For an EU resident whose primary venue is Binance, the practical effect has been mild; for someone using only EU-licensed venues, USDC and EURC have become the practical choice.

The actionable summary

  • If your primary venue is Binance, OKX, Bybit or another global exchange, your working balance is going to be USDT. That is fine. The token has the longest track record and the deepest liquidity for the spot use case.
  • If your primary venue is Coinbase, Kraken, Gemini or another US-regulated exchange, your working balance is going to be USDC. That is also fine. USDC is regulator-tight on those venues and the zero-fee USD conversion on Coinbase removes friction.
  • For DeFi positions, default to USDC unless the specific protocol is USDT-deeper (rare on Ethereum, common on BNB Chain and Tron).
  • For cold storage above a few thousand dollars, hold both. The split ratio is less important than the diversification.
  • For positions in any single stablecoin above the rough threshold where you would feel the loss meaningfully (whatever number that is for you), the desk's working answer is the same: split.

If you want to act on this

The desk uses Binance as its primary global venue for the operational reasons described. The referral link opens a Binance registration page pre-filled with the StableDesk referral code BN16188. Registering through that link does not change your fees; Binance pays the referral service fee from its own marketing budget. The full disclosure is on the disclaimer page. US residents whose primary venue is Coinbase or Kraken should ignore the referral; the safety analysis above does not change.

Issuer attestations and case files

  • Circle's monthly attestation, current and prior twelve months, on Circle's transparency page.
  • Tether's quarterly attestation, signed by BDO, on Tether's transparency page.
  • NY DFS Order Imposing Civil Money Penalty (Bitfinex / Tether, 2021), available on the OAG site.
  • FDIC announcement on Silicon Valley Bank, March 10-12 2023.
  • Circle Internet Group, Form S-1 and subsequent 10-Q filings (2024-2026), available on SEC EDGAR.
  • DeFiLlama stablecoin dashboard for circulating-supply history and chain breakdown.

Anything you can verify yourself, you should. Corrections to this article live on the corrections page.