Stablecoins are no longer a crypto-trading instrument. They are payment infrastructure. By May 2026 the total stablecoin market capitalization had reached roughly $320 billion, up from levels that would have looked like a complete market only a few years earlier. Tether's USDT accounts for around $189 billion of that total and Circle's USDC for around $77 billion, together representing the large majority of supply.
The more telling number is throughput. By some industry estimates, stablecoins settled on the order of $46 trillion in transfer volume during 2025, a figure that, on those estimates, exceeds PayPal's annual volume many times over and approaches the scale of major card networks. In the first quarter of 2026, stablecoins accounted for roughly three quarters of all crypto trading volume. Whatever the precise figures, the direction is unambiguous: stablecoins now move serious money.
That scale changes the nature of the risk. When stablecoins were a niche trading tool, a failure was a trading loss. As they become rails for remittances, merchant settlement, treasury operations, and institutional collateral, a failure becomes a payment failure. And the weakest part of the stablecoin stack is not the token or its reserves. It is the layer the token moves across.
The Asset Is Getting Safer. The Rails Are Not.
The reserve side of stablecoins has improved markedly. Regulation has forced it. In the United States, the GENIUS Act requires payment stablecoins to be backed one-to-one by high-quality liquid assets, limited to US dollar cash, short-dated Treasury bills, overnight repos, and central bank balances. In the European Union, MiCA requires authorized issuers to hold one-to-one reserves, to keep a substantial portion in bank deposits, and to honor redemption at par at any time.
These regimes are not identical, and that is a problem in itself, but they share a goal: make sure that one unit of a stablecoin is genuinely backed by one unit of value, redeemable on demand. On the issuance side, the question of "is this token actually backed" is being answered.
The movement side is a different story. A stablecoin issued and fully reserved on one chain does not stay on one chain. Users and applications want it everywhere, so it gets moved across chains. The dominant mechanism for that movement, the cross-chain bridge, is where the losses have concentrated.
How Bridges Concentrate Failure
Most general-purpose bridges work by locking a token on the source chain and minting a representation of it on the destination chain. The original is held in a contract. A new token, a synthetic copy, circulates on the other side. As long as the lock holds, the copy is backed. If the lock is broken, the copy is not.
This lock-and-mint model creates synthetic supply that is only as sound as the bridge securing it. When a bridge is exploited, the representation on the destination chain can become unbacked instantly, and it can depeg regardless of how sound the issuer's actual reserves are. The reserves are fine. The copy is the problem.
The numbers bear this out. Cumulative losses from bridge exploits have exceeded $2.8 billion, and bridges have accounted for a large share of total value stolen across the sector. These are not edge cases. They are a recurring structural failure of the same design.
Recent incidents show the pattern reaching stablecoins specifically. The Infini stablecoin platform lost roughly $49 million in early 2025 after an actor retained privileged access and drained funds. In a separate 2026 incident, the Saga protocol paused one of its chains after a multi-million-dollar exploit that triggered a stablecoin depeg through cross-chain fund movement. In each case the failure was not in the idea of a dollar token. It was in the infrastructure carrying that token between environments.
Why This Matters More Now
When stablecoins were used mainly to move between trading venues, a depegged wrapper was a contained loss absorbed by sophisticated participants. That is no longer the context.
Stablecoins are entering regulated payments and treasury use. An importer settling an invoice, a payroll provider distributing wages, a fund posting tokenized dollars as collateral: these users are relying on the token being worth exactly one dollar at the moment they need it. They are not in a position to absorb the failure of a bridge they may not even know is in the path. For them, "the issuer's reserves were fine, but the bridged version on this chain was compromised" is not an acceptable outcome. It is a settlement failure.
The industry has started to respond at the application level. Circle's Cross-Chain Transfer Protocol, for example, avoids the synthetic-wrapper problem by burning canonical USDC on the source chain and minting canonical USDC on the destination, so there is no wrapped copy that can desync from reserves. That is a meaningful improvement. But it also concedes the underlying point: the lock-and-mint model that still carries large volumes of stablecoins is a structural liability, and the integrity of stablecoin supply across environments is now a first-order question.
The Question Worth Asking
The stablecoin debate has spent years on reserves: what backs the token, who audits it, how often. That debate is being resolved by regulation. The next debate is about settlement: when a stablecoin moves, what guarantees that the thing arriving is the real, fully backed asset and not a representation that can come unbacked.
For payment use at scale, that is the question that matters. The reserves can be perfect and the user can still be exposed if the rails are not. The rest of this series looks at what settlement finality actually means for regulated stablecoins, and what it takes to issue and move them on infrastructure where supply integrity is built in rather than bridged on.
This article is for informational purposes only and does not constitute investment advice.
Mintlayer Web Services helps institutions issue and settle assets on Bitcoin-native infrastructure. Learn more →
