Short version: a stablecoin is a crypto asset pegged 1:1 to the US dollar. One USDT is supposed to equal one dollar. One USDC is supposed to equal one dollar. That sounds simple — but the word "pegged" hides an entire financial machine underneath, and the machine has broken several times in the last decade. Each time it broke, somebody got wiped out. This piece walks through the four mainstream stablecoins, two failures I actually lived through, and how the picture looks specifically if you read this from the US or Europe.
§1 · Why do stablecoins exist at all?
BTC and ETH routinely move 5% a day. Through the 72 hours that LUNA-UST died in May 2022, I was awake at four in the morning watching a stablecoin print zero, which is the kind of thing that makes you reread the word "pegged" very carefully. Nobody is going to pay rent in something that might be 10% cheaper by Friday. Nobody is going to do payroll in an asset that has weekly drawdowns the size of a small recession. So crypto needed a price-stable settlement layer, and the dollar was the obvious choice because the dollar is what people already think in.
Before stablecoins existed, the only way to take risk off the table was to wire money back out to a bank. That process is slow, expensive, and pinched at every customs checkpoint between you and the bank — cross-border wires take days, hit foreign-exchange spread, and incur compliance friction at both ends. When Tether launched USDT in 2014, the breakthrough was simple but powerful: your dollars could stay on-chain, ready to flip back into BTC the moment you wanted to re-enter. That single innovation is what made 24/7 global crypto trading actually fluid.
Today stablecoins are the bloodstream of the entire market. Aggregate stablecoin supply sits roughly between $150B and $180B as of early 2026. More than 70% of spot trading pairs on major exchanges are quoted in stablecoins rather than fiat. If you have spent more than a week buying or selling crypto, you have touched one — even if you did not know it.
§2 · Four stablecoin types, four very different peg mechanisms
Type 1 · Fiat-reserve-backed (USDT, USDC, FDUSD, PYUSD)
The most direct model. The issuer takes in one dollar, mints one token. When you redeem, the token is burned and a dollar is wired back to you. Behind the token there is supposed to be an actual dollar, or a dollar-equivalent asset like short-duration US Treasury bills, sitting in a regulated banking partner. USDT, USDC, FDUSD and PayPal's PYUSD all use this design.
The risks are easy to enumerate but harder to defend against:
- Are the reserves real? Tether was hammered for years on suspicion of under-collateralization, and while no large failure has materialized, the audit standard remains looser than USDC's.
- Is the banking partner safe? This is exactly what blew up USDC in March 2023 — more on that in a moment.
- Regulatory action: a US authority can freeze the issuer's accounts, and the issuer can freeze your token at the smart-contract level. USDT and USDC both have on-chain blacklist functions. Hundreds of millions of dollars worth of tokens have been frozen this way at OFAC's request.
Type 2 · Over-collateralized crypto-backed (DAI / sDAI / USDS)
DAI, issued by MakerDAO (now rebranded to Sky), takes the opposite approach. There is no bank account. Instead, you lock 150% (or more) of dollar value in ETH or other crypto into a smart contract and the protocol mints you 100 dollars of DAI. If your collateral falls below the threshold, the protocol forcibly liquidates it.
Advantages: every dollar of backing is on-chain, instantly verifiable by anyone with a block explorer. No bank, no jurisdictional risk. Disadvantages: capital-inefficient (you tie up $1.50 to use $1.00), and during violent price moves the liquidation mechanism can fail mechanically — as it did during the March 12, 2020 crash, when gas prices spiked so high that liquidation bots could not bid, and some vaults were auctioned at zero.
Type 3 · Algorithmic stablecoins (USTC, dead)
The most dangerous category. There is no external collateral at all. The peg is held entirely by algorithmic mint-and-burn loops with a sister token, and by arbitrageurs being willing to step in. The textbook failure is Terra's UST (now renamed USTC because it collapsed). That model has been tried dozens of times and has failed dozens of times — Iron Finance, Beanstalk, Basis Cash all went to zero in different mechanical ways. UST was just the largest, wiping out roughly $40B of paper market cap.
Type 4 · Yield-bearing / synthetic dollars (USDe, sDAI, sUSDS)
A newer category that emerged in 2023-2024. Ethena's USDe is the headline example: it generates a synthetic dollar by combining long spot ETH with an equal-size short ETH perpetual futures position, hedging the price exposure to zero and capturing perp funding rates as yield. The yield is real when funding is positive. When markets dislocate and funding inverts, the yield collapses and the peg can drift. The mechanism is fundamentally a derivatives trading strategy wrapped in a token. I personally stay away from these as a savings vehicle.
§3 · The four mainstream stablecoins, side by side
| Stablecoin | Issuer | Type | Market cap | Transparency | Notable depegs |
|---|---|---|---|---|---|
| USDT | Tether Ltd. | Fiat reserve | ~$110B | Quarterly attestation; limited full audit | Repeated minor depegs, never a structural failure |
| USDC | Circle Internet Group | Fiat reserve | ~$35B | Monthly attestation by Deloitte; highest compliance grade | March 2023 SVB event — depegged to $0.87 |
| DAI / USDS | MakerDAO (Sky) | Crypto-backed | ~$5B | Fully on-chain | March 12, 2020 — brief depeg during gas crisis |
| FDUSD | First Digital (FD121, HK) | Fiat reserve | ~$1B | Monthly attestation | August 2024 — brief depeg on rumored issues |
The market-cap numbers are early-2026 ballparks. They move daily — pull current data from CoinGecko or DefiLlama if you need it live.
§4 · Lesson 1 · How LUNA-UST actually died
Monday, May 9, 2022. I opened my phone in the morning and saw UST trading at $0.98. That is a stablecoin. It is supposed to print $1.00, mechanically, all day every day. A 2% deviation is already an event. My stomach did the small drop. I pinged a friend on Twitter who said, "Anchor at 20% APY was always going to break this. It is happening now."
The next three days:
May 10 — UST at $0.60, LUNA from $60 down to $30.
May 11 — UST at $0.30, LUNA at $5.
May 12 — UST under $0.10, LUNA at $0.01. Effectively zero.
Roughly $40B of market cap erased in 72 hours. I had a small LUNA bag — about one month of salary at the time — and I did not save any of it. Not because there was no time to exit. There was. The exits were open all the way through Wednesday morning. It was that I could not bring myself to believe a stablecoin would actually go to zero. The whole time my brain kept telling me "this is a peg, it has to mean-revert." It did not.
That week taught me the single most important lesson about stablecoins: a stablecoin with no real-world reserves is not, in any meaningful sense, stable.
The death spiral, mechanically
The Terra mechanism worked like this: $1 of UST could always be minted by burning $1 worth of LUNA, and vice versa. In theory, arbitrageurs would keep UST pinned to a dollar. The hidden assumption was that LUNA had to have enough market depth to absorb the redemptions.
Once a critical mass of UST holders tried to exit at the same time — triggered by a coordinated sell on Anchor Protocol — the system had to mint enormous amounts of LUNA to absorb the redemptions. LUNA supply exploded. LUNA price collapsed. The collapse made breaking the UST peg even cheaper, which sent another wave of redemptions through the system, which minted more LUNA, which collapsed further. Within three days the loop completed. Every algorithmic stablecoin with no external collateral has this exact failure mode latent inside it. No pure algorithmic design has ever survived a real run.
§5 · Lesson 2 · The USDC-SVB depeg
Friday, March 10, 2023. Silicon Valley Bank was placed into FDIC receivership — the second-largest US bank failure since 2008. Circle had roughly $3.3 billion of USDC reserves parked at SVB. That was disclosed Friday night.
Over the weekend, with US banking closed, the on-chain market did the only thing it could: it priced the uncertainty. USDC fell from $1.00 to $0.87 — a 13% depeg, the deepest in USDC's history.
I happened to have a USDC position parked at the time. Sitting there on a Sunday afternoon, doing the math — if SVB went to zero, this stack was 13% smaller. Should I swap into USDT? But USDT had its own transparency questions, and Tether was also a Silvergate-era counterparty. I did nothing. Just watched.
Monday morning, the FDIC announced it would make SVB depositors whole, including uninsured deposits. Circle's $3.3B was fully recoverable. USDC traded back to $1.00 within hours. The sweat had been wasted. But it was useful sweat.
What this episode teaches
There is no such thing as an "absolutely safe" fiat-backed stablecoin. Reserves can be transparent, audited, and properly held — but the moment they are deposited at a specific bank, you inherit that bank's risk. Circle was lucky in March 2023. The US government intervened. Next time the political environment might be different, or the bank involved might be too small to warrant intervention.
My approach since 2023 has been to diversify across stablecoins, not to bet on any single issuer: a slice in USDT (deepest liquidity, fewest gas pain points), a slice in USDC (highest compliance, best US/EU rails), and a small slice in DAI/USDS (on-chain transparency as a hedge against off-chain failure). Concentration in one stablecoin is a category of risk most retail underestimates.
§6 · The US & European reader perspective
The first five sections of this article apply whether you are sitting in Singapore, Buenos Aires, or Riyadh. The mechanics of the peg do not care where you live. But if you are reading this from the United States or the EU, your access, your regulatory environment, and the specific stablecoin you should be holding are different from what a non-US reader would default to. This section is the part that does not appear in the Chinese-language version of this article.
The GENIUS Act and the US federal stablecoin framework
For most of 2024 and 2025, the US Congress has been working through several competing stablecoin bills. The one most likely to land is the GENIUS Act — "Guiding and Establishing National Innovation for US Stablecoins" — which would create a federal regulatory framework for payment stablecoin issuers. The headline provisions are roughly: 1:1 backing by short-duration cash equivalents (US Treasury bills, repos, FDIC-insured deposits), monthly attestation, federal or state chartering options, and clear redemption rights for holders. Pair that with the Clarity Act on broader digital-asset structure, and you have the first US federal regime for stablecoins. If GENIUS passes in roughly its current form, USDC's regulatory position becomes a structural advantage, while Tether — which has explicitly stated it will not seek US issuer status — becomes US-restricted.
NYDFS, Paxos and the Trust Charter route
Before any federal stablecoin law existed, New York was already running its own regime through the New York Department of Financial Services. NYDFS BitLicense plus the Limited Purpose Trust Company charter is the pathway most US-regulated stablecoin issuers have used. Paxos Trust Company issues USDP (Paxos USD) and previously issued BUSD under contract from Binance until NYDFS forced the wind-down in February 2023. Gemini Trust Company issues GUSD on the same charter framework. PayPal's PYUSD, launched in August 2023, is also issued by Paxos under NYDFS supervision — that arrangement is the regulatory reason PayPal could launch a stablecoin at all without seeking federal authority.
Circle, USDC and the 2024-2025 IPO
Circle, USDC's issuer, is headquartered in Boston, regulated as a money transmitter in essentially every US state, and supervised at the federal level through Treasury's FinCEN. Circle filed for its public listing during 2024 and completed its IPO in 2024-2025 (on the NYSE), making it one of the very first publicly traded pure-play stablecoin issuers in the world. Reserves are held in the Circle Reserve Fund, an SEC-registered money market fund managed by BlackRock, with the cash portion at Bank of New York Mellon and other regulated banks. Monthly attestations are produced by Deloitte. As of early 2026, more than 80% of USDC's backing is short-duration US T-bills, with the remainder in regulated bank deposits. For a US reader, USDC is the cleanest stablecoin holding from a transparency-and-reserves perspective. The trade-off is liquidity: USDT still has 3x the on-chain depth and dominates non-US venues.
Tether, USDT and the question of US access
Tether is incorporated in the British Virgin Islands. Its operational center has moved several times — currently it runs banking through Cantor Fitzgerald (Howard Lutnick's firm, which now also handles a substantial portion of Tether's US Treasury holdings on behalf of the issuer). Tether has settled with the CFTC ($41M, 2021), with the New York Attorney General ($18.5M, 2021), and faces ongoing oversight, but no headline reserve failure has happened. The reserve composition is now disclosed quarterly. As of 2025-2026, the bulk is in US T-bills, with smaller portions in secured loans, Bitcoin, and precious metals. For US persons, the practical question is access: Coinbase delisted USDT for retail US accounts in 2023. Kraken did the same for some products. Tether is functionally an offshore stablecoin from the US perspective, even though much of its reserve sits in US instruments.
March 2023 deep dive · what FDIC pass-through insurance actually covers
The widespread misconception that the SVB-USDC episode exposed is the belief that stablecoin balances are themselves FDIC-insured. They are not. What FDIC insurance covers is your deposit at a regulated US bank, up to $250,000 per depositor per institution. A stablecoin issuer's deposit at a bank may qualify for pass-through insurance — meaning if the bank fails, each individual customer's pro-rata share of that deposit, traced back through the issuer's customer records, is insured up to the standard limit. But the stablecoin token itself, sitting in your wallet, has no FDIC coverage. What saved USDC in March 2023 was not pass-through insurance (which would have only protected the first $250K). It was the FDIC's separate decision to invoke a systemic risk exception and guarantee all SVB deposits, insured or not. That decision was political, not mechanical. Do not assume it will be made again the same way next time.
The MiCA effect, USDT delisting in the EU, and Circle's regulatory edge
The EU's Markets in Crypto-Assets Regulation became fully applicable in December 2024, with the stablecoin-specific rules in effect since June 2024. Under MiCA, any "asset-referenced token" or "e-money token" offered in the EU must be issued by a CASP (Crypto Asset Service Provider) holding an EU authorization, with reserves held in EU banks under strict liquidity rules. Circle restructured its European operations to comply and obtained an EMI license in France. Circle's USDC is fully MiCA-compliant. Tether opted out. As a result, USDT was systematically delisted from regulated European venues during the first half of 2025: Coinbase EU delisted USDT in March, OKX Europe and Crypto.com Europe followed. Binance restricted USDT trading pairs for EEA users while keeping the token itself accessible. The practical effect for EU readers: USDC is the default MiCA-compliant stablecoin parking option, while USDT access continues to narrow.
PayPal USD and the US payment stablecoin push
PayPal launched PYUSD in August 2023, issued by Paxos under NYDFS supervision. The integration matters less for the token itself than for what it represents: a US-listed payments company embedding a stablecoin directly into its consumer wallet and into Venmo. PYUSD circulating supply remains relatively small (under $1B as of early 2026) but adoption is structural, not speculative. Other payment-stablecoin moves on the US side worth noting: Stripe's acquisition of Bridge.xyz in 2024, which positions Stripe to do cross-border B2B settlements in stablecoins; Visa's stablecoin settlement pilot on Solana with USDC; and Mastercard's tokenization framework. The thesis these companies are working on is that stablecoins are slowly absorbing cross-border B2B payments, payroll for distributed teams, and on-ramp/off-ramp friction — none of which is speculation, all of which is real volume.
For US readers · what to actually hold
Pulling this together into something practical:
- Primary holding: USDC if you are based in the US. Cleanest reserves, monthly Deloitte attestation, NYDFS-comfortable counterparties, and aligned with where federal regulation is headed.
- Secondary: PYUSD or GUSD if you specifically want a Paxos-issued, NYDFS-Trust-Charter-backed alternative. PYUSD also gives you native PayPal/Venmo integration if you actually use those rails.
- Avoid or minimize: USDT for US persons. Increasingly restricted on US-regulated venues. Not banned, but its operational use case in the US has narrowed.
- Custody: not at the issuer level. Holding USDC at Coinbase is exposing you to Coinbase's solvency, not Circle's. For meaningful balances, self-custody on hardware wallet (Ledger / Trezor) is the right answer.
- Tax: stablecoin-to-stablecoin swaps are taxable events in the US. The IRS treats every disposition as a realized gain or loss, including swapping USDT for USDC. Small fluctuations net out to roughly zero, but the reporting requirement is real. Tools like CoinTracker or Koinly handle this.
For EU readers · what to actually hold
- Primary holding: USDC as the MiCA-compliant default. Circle is now the only fiat-backed major stablecoin with full EU regulatory standing.
- EUR-denominated stablecoins are emerging: Société Générale's EURCV, Circle's EURC. These are smaller in liquidity but eliminate any FX exposure if your spending currency is EUR.
- USDT access at compliant EU venues is now restricted. If you held USDT pre-MiCA, you can still hold or transfer, but trading on EU venues is increasingly limited.
- Tax regimes are not MiCA-harmonized. Germany still gives 12-month tax-free disposal for individual holders. Portugal tightened in 2023 but remains favorable for long holds. France treats crypto-to-crypto as taxable. Italy moved to 26% on gains above €2,000. Check your specific jurisdiction.
§7 · Small amounts, mid-sized, large amounts — what to do
Small daily amounts (a few hundred to a few thousand dollars)
USDT or USDC both work. Pick whichever your exchange supports with the lowest fees and deepest liquidity.
On-chain, USDT-TRC20 (Tron network) has near-zero transfer fees and is the dominant rail for small global transfers. USDC on Solana or Base is similarly cheap and fast, with the bonus of being on a more credible regulatory footing.
Avoid: sending small amounts on Ethereum mainnet. Gas can spike to $10+ per transaction; moving $50 with a $10 fee is throwing away 20% of your principal.
Mid-sized balances (a few thousand to a few tens of thousands)
Split it. Half USDT and half USDC, on different chains. If one issuer hits a regulatory event or one chain has a temporary outage, you still have access to the other. Keep it on tier-one regulated venues — Coinbase, Kraken, or Binance for US/EU contexts — not on small offshore exchanges. The FTX lesson should be permanent: do not park meaningful balances at exchanges you cannot independently verify the solvency of.
Large balances (six figures and above)
Diversify further across USDC, USDT and DAI/USDS so a single-issuer event does not wipe you. Do not park large balances at any exchange. Move to your own hardware wallet — Ledger or Trezor — and treat the seed phrase like the only copy of a will. If you want to earn yield on idle balances, mature DeFi lending markets like Aave or Compound v3 are the options, but every smart contract carries non-zero hack risk, so cap the percentage. My personal rule: never more than 30% of stablecoin holdings in DeFi at any time, ever.
Do not trust any "20%+ APY" stablecoin product. Anchor Protocol offered 20% on UST. Hundreds of thousands of people lost their savings. Real-world short-dollar rates are 4-5%. If a protocol or platform is offering you 20% on a stablecoin in 2026, the spread is being generated by something you cannot see — and historically that "something" has been undisclosed leverage, rehypothecation, or trading risk dressed up as yield. Either it is fraud, or you are unknowingly holding a high-risk derivative wrapped in a stablecoin-shaped wrapper. Both end the same way.
§8 · Where the regulation is going
Stablecoin regulation has gone from non-existent to one of the most actively worked areas of crypto law in less than three years. On the US side, the GENIUS Act plus the Clarity Act would, if passed, create the first federal payment-stablecoin regime. NYDFS continues to operate the state-level model for trust-chartered issuers. On the EU side, MiCA is now fully operative. Stablecoin issuers either comply or lose access to regulated venues. Hong Kong's Stablecoins Ordinance, in force as of 2024-2025, creates an HKMA-licensed regime — HSBC, Standard Chartered and several PRC-linked banks are positioning to participate.
The longer-term trajectory is clear. Compliant, audited, fiat-backed stablecoins — USDC, PYUSD, GUSD, the emerging EUR-denominated set — become the default rail for retail and institutional payment use cases. USDT will continue to dominate non-US trading-pair liquidity for years because the on-chain network effect is hard to dislodge, but its US/EU regulatory footprint will continue to shrink. Pure algorithmic stablecoins have, as far as I can tell, no future in a regulated world — the math has failed every time and the regulators have noticed.
§9 · Closing thoughts
Stablecoins are not "automatically stable." They have a peg mechanism, reserve risk, counterparty bank risk, smart-contract risk, and regulatory risk. Any one of those can turn your "stable" position into a loss.
Three rules I would put in front of any new stablecoin holder:
- Diversify holdings. USDT and USDC both. Maybe a slice in DAI or USDS. Do not concentrate in any single issuer, no matter how good the recent attestation looked.
- Do not park at small exchanges. Large balances go to self-custody. Small operational balances stay at tier-one regulated venues.
- Avoid high-yield stablecoin products. The reasonable yield range is 3-8% in the current rate environment. Anything above that is buying you risk you almost certainly cannot price.
If your next step is going to be DeFi, stablecoins are going to be the tool you reach for every day. Read the DeFi piece next.
