# Hernández de Cos and tokenized money: everything must change so that banking remains

> Hernández de Cos backs tokenized bank money over stablecoins and defends preserving the monetary architecture
> By Observatorio Blockchain · Stablecoins · 2026-08-29

The general manager of the BIS, the Spaniard Pablo Hernández de Cos and former governor of the Bank of Spain, identifies several real problems in current stablecoins. However, the alternative he proposes opens a much deeper question: whether the technological revolution in money should also change the banking architecture that sustains it or be limited to modernizing it. That tension between technological innovation and institutional continuity runs through his entire speech on stablecoins and tokenized deposits.

There is undoubtedly something Lampedusan in his proposal: that the technology of money should change profoundly so that its institutional architecture remains. Money can be tokenized, made programmable and circulate on new infrastructures, but the central bank and commercial banks would continue to occupy the center of the monetary system. It is not a matter of slowing technological transformation, but of incorporating it without substantially altering who sustains, intermediates and guarantees money.

## Hernández de Cos and stablecoins

The current BIS executive does not question the usefulness of tokenization. On the contrary, he expressly acknowledges that it can bring programmability, atomic settlement and uninterrupted operations. What he questions is whether stablecoins, as they currently exist, can by themselves provide the properties that allow money to function with the trust that an economy requires.

His diagnosis is that in their current form, stablecoins do not yet comply with the fundamental principles of money, and he points to problems related to parity, interoperability, liquidity, panic risk and financial integrity. These are real problems, but what is truly interesting about his speech begins when he moves from diagnosis to solution. Because Hernández de Cos does not only explain what is wrong with stablecoins, he also sketches the architecture that, in his view, should sustain the tokenized money of the future.

And that architecture looks a lot like the current one. To explain one of the problems with stablecoins, Hernández de Cos uses a [simple example](https://www.bis.org/speeches/20260828-pushing-monetary-frontier-stablecoins-and-tokenised-deposits). Ben has USDT, the stablecoin issued by Tether, and wants to transfer one dollar to Marie. However, Marie only accepts USDC, issued by Circle. To complete the payment, Ben would have to sell USDT on the secondary market, buy USDC and then send it to Marie.

## Problems with stablecoins

The problem is that stablecoins can deviate from their parity on secondary markets, especially during stress situations. Therefore, the conversion might not take place exactly one-to-one. There is no mechanism that guarantees uniqueness, argues Hernández de Cos. The issue affects the very definition of money used by the BIS. A dollar should be a dollar regardless of the instrument used, and a euro should be a euro. In the traditional financial system, that equivalence does not arise spontaneously either. It is sustained by an institutional architecture built around central banks, commercial banks, supervision, regulation, liquidity and settlement systems.

[Stablecoins work](https://observatorioblockchain.com/en/stablecoins/stablecoins-espana-banca-dinero-tokenizado/) differently. They are private instruments that seek to maintain a certain value relative to a currency and that can circulate on public blockchains. Their stability depends, among other things, on reserves, liquidity, redemption mechanisms and trust in the issuer. Hernández de Cos also notes that blockchains are fragmented across different networks and scaling layers. Even the same stablecoin can exist on different chains without being directly interoperable between them. To this he adds the problems related to self-custody wallets and transfers between individuals when applying anti-money laundering and terrorist financing rules.

## Crisis of confidence and risk of mass withdrawals

Hernández de Cos also warns that stablecoin issuers remain exposed to the risk of mass withdrawals. A crisis of confidence could force the rapid sale of government bonds or withdrawal of deposits and transmit tension to money markets, he indicates. So far, the BIS general manager describes specific vulnerabilities of current stablecoins. The question changes when he says that the alternative he considers most promising is tokenized deposits. In his view, tokenized deposits offer a more direct way to take advantage of tokenization while preserving the foundations of the monetary system, he states.

The important word is perhaps preserving, because a tokenized deposit can use a new technological infrastructure, but it remains a liability of a commercial bank. Money can be made programmable, the infrastructure on which it moves can change, the way certain operations are settled can change, but the bank is still there. And behind the bank, the central bank is still there. In the model described by Hernández de Cos, payments with tokenized deposits debit the payer's balance and credit the beneficiary's, while interbank settlement is carried out through central bank accounts. This, he explains, preserves the uniqueness of money.

## The institutional architecture remains

The technology changes, but the institutional architecture remains. Among the recommendations included by Hernández de Cos is expressly integrating tokenization into the two-tier monetary architecture anchored in central bank money. Central bank, commercial banks and users. It is the current monetary architecture transferred to programmable infrastructures.

Here it is worth separating two questions. The first is whether current stablecoins have problems of parity, interoperability, liquidity, financial integrity or panic risk. Hernández de Cos argues that they do and points to these weaknesses as some of the main obstacles to their being able to function credibly as money at scale. The second question is different: do these deficiencies prove that the tokenized money of the future must be organized fundamentally around central banks and commercial banks? Hernández de Cos clearly opts to preserve that architecture. But then we are no longer dealing solely with a diagnosis of the limitations of stablecoins or of the technology that underpins them. We are dealing with a much deeper proposal about how the money of the future should be organized institutionally. And this is where one of the most interesting questions in the debate arises.

## Limitation of tokenized deposits

Hernández de Cos himself acknowledges an important limitation of tokenized deposits, stating that currently, there are no multi-bank or cross-jurisdictional ecosystems that issue tokenized deposits in an interoperable framework. This is a fundamental clarification. Stablecoins exist, circulate on different blockchains and their problems can be observed in the market. By contrast, the international system of interoperable tokenized deposits that Hernández de Cos considers most promising has yet to be built and also presents significant challenges.

The BIS general manager mentions among them interoperability, governance, legal clarity, cybersecurity, error management and recovery, and the need to make the new infrastructures coexist with existing financial systems. He even acknowledges that some of the bank tokens currently available, although they may provide greater trust and regulatory adaptation, share many of the deficiencies of stablecoins.

There is another interesting paradox in the speech. When Hernández de Cos asks what would be needed for stablecoins to function credibly as a means of payment at scale, he raises questions related to reserves, backing mechanisms, liquidity, clear redemption rights, governance, resolution, supervision and international regulatory coordination.

He also raises whether issuers might need access to central bank accounts or liquidity mechanisms and, if so, under what safeguards. He does not state that all those mechanisms must necessarily be provided to issuers, but rather formulates the questions that should be resolved for stablecoins to be able to play that role. But it is interesting to observe where those requirements lead. The greater the ambition of a stablecoin to become money used on a mass scale, the greater the requirements for stability, reserves, liquidity, transparency and supervision. And the greater those requirements, the more institutional elements it will have to incorporate.

The question is how many institutions of the traditional financial system does a stablecoin need before it starts to resemble the monetary system it initially operated outside of? Technology can decentralize certain functions, but monetary trust is much harder to decentralize.

## Who grants credit

There is another reason why the BIS watches the growth of stablecoins so closely. Deposits are not only used to pay, they are also a source of funding for banks, which use those resources to sustain, among other activities, the granting of credit.

Hernández de Cos warns that widespread adoption of stablecoins could alter that mechanism. If issuers' reserves were held mainly as wholesale bank deposits, he explains, part of retail funding could be replaced by more concentrated wholesale liabilities that are more sensitive to interest rates. This could raise banks' marginal funding costs and tighten lending conditions.

Nor would the effects necessarily be the same for all entities. Hernández de Cos notes that smaller banks could be especially affected, with possible consequences for small business financing. Tokenized deposits retain a different logic. According to Hernández de Cos, they maintain the link between deposit-taking and lending, in addition to keeping funds within the banking system.

This difference helps explain why the debate on stablecoins is much more than a competition between payment technologies. If a significant part of the money deposited in banks were to shift toward new private issuers, the way financing circulates within the economy could also change. What is at stake is not only who moves the money, but also who intermediates it and who grants credit.

## Deposits finance banks

There is another reason why the BIS watches the growth of stablecoins so closely. Deposits are not only used to pay: they also finance banks, and banks use part of that funding to grant credit. Hernández de Cos warns that widespread adoption of stablecoins could alter that mechanism. If issuers' reserves were held mainly as wholesale bank deposits, he explains, retail funding could be replaced by more concentrated wholesale liabilities that are more sensitive to interest rates. This could raise banks' funding costs and tighten lending conditions, with a potentially greater impact on small entities and small business financing.

Tokenized deposits retain another logic. According to Hernández de Cos, they preserve the link between deposit-taking and lending and keep funds within the banking system.

This difference helps explain why the debate on stablecoins is much more than a competition between payment technologies. If a significant part of bank deposits were to shift toward new private issuers, the way financing circulates within the economy could also change. What is at stake is not only who moves the money, but also who intermediates it and who grants credit.

## He does not propose eliminating stablecoins

Hernández de Cos does not propose eliminating stablecoins. His scenario envisages coexistence with tokenized deposits, although assigning different functions to each. Tokenized deposits should take on most everyday payments and wholesale settlement, while stablecoins could retain specialized functions, for example in decentralized lending markets. To be used as means of payment, they should operate under robust and transparent regimes that guarantee redemption at par. Hernández de Cos also raises the possibility of their being treated explicitly as investment products, with the corresponding conduct and disclosure rules.

In the future described by the BIS general manager, the heart of the tokenized monetary system would remain anchored in central bank money and bank deposits. Stablecoins could coexist with it, but performing other functions or subjecting themselves to sufficient guarantees to be used as means of payment.

Hernández de Cos expressly acknowledges their advantages, but also the deficiencies of the current system. His proposal consists of modernizing the system without abandoning its institutional foundations. At the end of his remarks he summarizes this philosophy by noting that the path toward the future monetary system involves improving the old while making the new possible.

It is a modernization proposal, but also a clear institutional bet. Money and deposits can become tokens and central bank reserves can be integrated into programmable platforms. What Hernández de Cos wants to preserve is the system of trust that connects those new forms of money with the central bank. The battle for tokenized money is only just beginning.
