Consider that the world's central bank for central banks just declared war on a $100 billion monthly market. On August 28, at the Jackson Hole Economic Policy Symposium, BIS General Manager Agustín Carstens formally rejected stablecoins as viable payment instruments. He used a three-test framework—singleness, interoperability, finality—and concluded that stablecoins fail every single criterion for sound money. The market's response? Monthly stablecoin transaction volume hit $100 billion, up 300% year-over-year. The disconnect is not an anomaly. It is a structural signal.
This is not a debate about technology. It is a battle for the architectural soul of the global payment system. On one side, the BIS and its Project Agorá initiative, pushing tokenized deposits as the institutional evolution of commercial bank money. On the other, a consortium of twelve global banks—including Bank of America, Wells Fargo, and Santander—building stablecoin joint ventures on public chains. The outcome will determine the next decade of financial infrastructure.
I have spent the last nineteen years dissecting protocol architectures, from auditing Uniswap V1's core contracts in 2017 to reverse-engineering Groth16 proof generation circuits in zkSync Era. This conflict between tokenized deposits and stablecoins is not merely a regulatory squabble. It is a fundamental fork in the road for how value moves across the internet. Let me break down what the BIS is actually saying, what the banks are actually building, and why both sides are simultaneously right and dangerously wrong.
The Context: Two Competing Visions for Institutional Money
The BIS position is not new. Carstens has been a vocal critic of private digital currencies for years. What changed at Jackson Hole is the specificity of the critique. He did not dismiss stablecoins as risky or immature. He argued they are structurally incapable of functioning as sound money. The three tests he deployed are worth examining with forensic precision.
Singleness refers to the property of a uniform unit of account. In a well-functioning monetary system, one dollar is always one dollar, regardless of who holds it or where it is held. Stablecoins fail this test because they operate on fragmented rails. A USDT transaction on Tron does not directly interoperate with a USDC transaction on Ethereum. They require conversion, which introduces friction and breaks the uniformity of the currency. This is not a bug that can be patched. It is an architectural property of permissionless public chains.
Interoperability is the second test. Different payment systems must seamlessly interact. The current stablecoin ecosystem is a patchwork of isolated networks. Cross-chain bridges exist, but they introduce new security risks. I have audited enough bridge contracts to know that every bridge is a honeypot waiting for a sufficiently motivated attacker. The BIS argues that tokenized deposits, built on shared institutional infrastructure, eliminate this problem by design. All participating banks would operate on the same ledger, with the same settlement rules.
Finality is the third and most critical test. Central bank money has an implicit guarantee of finality backed by sovereign credit. When a central bank settles a payment, the transaction is final. There is no counterparty risk. Stablecoins, by contrast, carry issuer counterparty risk, reserve composition risk, and an evolving regulatory framework risk. The BIS argues that this makes stablecoins fundamentally inferior to central bank money for settlement purposes.
These are not trivial arguments. They are technically sound. But they are also self-serving. The BIS is an institution whose raison d'être is the stability of the existing banking system. Tokenized deposits preserve the two-tier banking structure, keeping commercial banks as intermediaries. Stablecoins threaten to disintermediate banks entirely. The BIS is not defending sound money. It is defending its member banks' monopoly on the creation of money.
The Core: A Code-Level Analysis of the Two Architectures
Let me be precise about what each architecture actually is, because the marketing obscures the technical reality.
Stablecoins are, at their core, a simple smart contract. A central issuer holds reserves and issues tokens on a public chain. The token is a claim on the issuer, not on any specific asset. The reserve composition varies. Tether holds a mix of commercial paper, treasuries, and other instruments. Circle holds mostly treasuries. The token's value derives from the issuer's ability to maintain the peg through redemption. This is a centralized system with a decentralized settlement layer. The trust model is bifurcated: you trust the issuer for solvency, and you trust the chain for settlement.
Tokenized deposits are fundamentally different. They are not tokens in the crypto sense. They are digital representations of commercial bank liabilities, recorded on a shared ledger. The ledger is permissioned. Nodes are operated by regulated banks. Settlement occurs through the central bank's real-time gross settlement system. The trust model is unified: you trust the bank for solvency, and you trust the central bank for finality. The blockchain is merely a coordination mechanism, not a trust anchor.
This distinction matters because it determines the security assumptions. Stablecoins inherit the security of the underlying public chain. Ethereum's proof-of-stake consensus, for example, provides settlement finality through economic slashing. But the issuer is a single point of failure. If Tether's reserves are mismanaged, the entire system collapses. Tokenized deposits inherit the security of the banking system. The bank is the point of failure, but the central bank acts as the lender of last resort. The systemic risk is different.
Now, let me address the fragmentation problem that Carstens highlighted. He is correct that stablecoins are fragmented. But this is not an inherent flaw. It is a market structure problem. The reason USDT on Tron does not interoperate with USDC on Ethereum is not a technical limitation. It is a business decision. Tether and Circle are competitors. They have no incentive to build shared infrastructure. The fragmentation is a feature, not a bug, from the issuers' perspective. It locks in network effects and makes switching costs prohibitive.
Tokenized deposits, by contrast, are designed for interoperability from the ground up. Project Agorá, the BIS initiative, brings together seven central banks and major commercial banks to prototype cross-border tokenized deposit settlement. The shared infrastructure is the point. All participants operate on the same ledger, with the same rules. This eliminates the fragmentation problem by fiat. But it also eliminates the permissionless innovation that makes public chains so powerful.
Here is the trade-off that neither side acknowledges. Stablecoins sacrifice singleness for openness. Tokenized deposits sacrifice openness for singleness. There is no free lunch. The question is which trade-off is more valuable for the specific use case.
For cross-border wholesale payments, tokenized deposits are arguably superior. The participants are known, the volumes are high, and the need for finality is absolute. A permissioned ledger with central bank settlement is the optimal architecture. This is why the BIS is pushing Project Agorá. It is not a technology experiment. It is a strategic move to keep settlement within the banking system.
For retail payments and DeFi, stablecoins are superior. The openness of public chains enables composability, which is the killer feature of crypto. A stablecoin on Ethereum can be used as collateral in Aave, traded on Uniswap, and settled in a zk-rollup. This is impossible with tokenized deposits, which are walled gardens. The composability of stablecoins is what drives the $100 billion monthly volume. It is not a bug. It is the entire point.
The Contrarian Angle: The BIS Is Solving a Problem That Does Not Exist
Here is where I diverge from the consensus. The BIS's critique of stablecoins is technically correct but strategically misguided. The fragmentation problem is real, but it is not the problem that needs solving. The actual problem is the lack of a universal settlement layer for the internet. Stablecoins are a stopgap solution. Tokenized deposits are a different stopgap. Neither is the final answer.
The BIS is betting that tokenized deposits will become the institutional standard. The twelve-bank consortium is betting that stablecoins on public chains will achieve institutional grade. Both bets are based on a flawed assumption: that there is a single winning architecture. The reality is that the future will be multi-architectural. Stablecoins will dominate the crypto-native economy. Tokenized deposits will dominate the traditional financial economy. The two will coexist, connected by bridges and settlement layers that do not yet exist.
This is where the real opportunity lies. The infrastructure that connects these two worlds—the settlement layer that enables a tokenized deposit to be swapped for a stablecoin without friction—is worth more than either architecture individually. This is the layer that no one is building. The BIS is building tokenized deposits. The banks are building stablecoins. Neither is building the bridge.
Let me also address the elephant in the room: the GENIUS Act. The US Payment Stablecoin Act was enacted on July 18, 2025, with enforcement beginning January 18, 2027. Seven agencies have already missed the one-year rulemaking deadline. The regulatory landscape remains fragmented and ad hoc. This is not a bug. It is a feature of the political system. The delay creates a window of opportunity for market participants to position themselves before the rules are finalized.
The GENIUS Act is a bet that stablecoins can achieve institutional standards. It imposes reserve requirements, transparency obligations, and redemption guarantees. This will drive consolidation. Small issuers will be unable to comply. Large issuers like Tether and Circle will benefit from the compliance moat. The market will bifurcate into regulated stablecoins and unregulated ones. The regulated ones will gain institutional adoption. The unregulated ones will remain in the crypto-native economy.
This is not a prediction. It is a logical consequence of the regulatory framework. The question is whether the compliance moat will be worth the cost. The GENIUS Act requires 1:1 reserves, monthly attestations, and bankruptcy remoteness. These are expensive requirements. They will increase the cost of issuance, which will be passed on to users. The question is whether the market will accept higher fees for regulatory clarity.
The Takeaway: The Future Is Not a Choice, It Is a Bridge
The BIS has drawn a line in the sand. The banks have crossed it. The market is voting with its feet. Monthly stablecoin volume of $100 billion is not a rounding error. It is a signal that the demand for programmable, internet-native money is real and growing. The BIS can reject stablecoins, but it cannot reject the demand that drives them.
The next five years will be defined by the battle between these two architectures. But the real winners will be the builders who recognize that this is not a zero-sum game. The future is a bridge between the permissioned world of tokenized deposits and the permissionless world of stablecoins. The infrastructure that connects these two worlds will be the foundation of the next generation of financial services.
Trust is math, not magic. The math of tokenized deposits is central bank finality. The math of stablecoins is cryptographic consensus. Both are valid. Both are incomplete. The system that emerges will be a hybrid, and the builders who understand this will be the ones who capture the value.
Composability is a double-edged sword. It is what makes stablecoins powerful. It is also what makes them fragile. The same composability that enables a stablecoin to be used as collateral in a DeFi protocol also enables a vulnerability in one protocol to cascade into another. I have seen this firsthand in my analysis of the Aave-Compound interaction during DeFi Summer 2020. The systemic risk is real. But it is manageable with rigorous auditing and careful design.
Speculation audits the soul of value. The stablecoin market is driven by speculation, but it is also driven by real demand for dollar-denominated value on the internet. The BIS can dismiss this as speculation, but it cannot dismiss the underlying need. The question is not whether stablecoins will survive. It is whether they will evolve into something that meets the standards of sound money.
Zero knowledge speaks louder than proof. The future of this battle will be determined not by marketing but by technical rigor. The side that builds the most secure, most efficient, most interoperable infrastructure will win. The BIS has the institutional weight. The banks have the capital. The crypto-native builders have the innovation. The outcome is uncertain. But the direction is clear.
Architects build, auditors break. The BIS is an architect. The banks are architects. The auditors are the ones who will determine whether their creations survive contact with reality. I have spent my career breaking things. I know that every system has vulnerabilities. The question is whether the architects are willing to listen to the auditors before the market forces them to.
Silence is the ultimate verification. The Fed chair, Kevin Warsh, spoke at Jackson Hole hours before Carstens and did not mention digital assets at all. That silence is more telling than any statement. It suggests that the Fed is not ready to take a position. It is waiting to see which architecture wins. That is the smart play. The rest of us should do the same.
Innovation decays without rigorous scrutiny. The stablecoin market is growing at 300% annually. That growth is unsustainable without corresponding investment in security and compliance. The BIS is right to demand rigor. The banks are right to demand institutional standards. The market is right to demand innovation. The challenge is to achieve all three simultaneously.
Patterns emerge from chaos, not noise. The chaos of the current regulatory landscape is not a bug. It is the process by which the market discovers the optimal architecture. The BIS is pushing one direction. The banks are pushing another. The market is voting with its feet. The pattern that emerges will be the one that survives. My job is to analyze the pattern, not to predict it.
The BIS has made its position clear. The banks have made their bet. The market has made its choice. The only question that remains is whether the infrastructure will be built to connect these competing visions. That is the opportunity. That is the challenge. That is the future.
Trust is math, not magic. The math is still being written.