Countries representing 98% of global GDP are exploring central bank digital currencies. They must devise digital infrastructure to maximize competition, writes Jeff Alvares.
Central bank digital currencies (CBDCs) are poised to become the next frontier of digital protectionism. As central banks from the Bahamas to China replace paper money with code, the global monetary system faces a profound transformation. Until recently, 137 economies accounting for 98% of global GDP were exploring central bank digital currencies. While the United States has paused its retail CBDC efforts in favor of private stablecoins, the rest of the world is moving ahead, motivated by financial inclusion, regulatory oversight, and a more multipolar monetary system.
The technology behind CBDCs introduces new possibilities by integrating layers that traditional banking keeps separate: currency issuance, settlement infrastructure, payment schemes, and applications (like digital wallets). As nations rush from pilot to launch, they risk using this integration to blur the line between legitimate monetary sovereignty and protectionist monopolization of digital services. A critical question emerges: Where does legitimate monetary sovereignty end, and where do trade‑restrictive digital market policies begin? The answer hinges on the crucial distinction between controlling money itself and controlling the competitive markets that form around it.
This question is not abstract. In July 2025, the Office of the U.S. Trade Representative launched a Section 301 investigation into Brazil’s Pix instant payment system over alleged restrictive policies advantaging the payment system under public control. The investigation echoed earlier World Trade Organization disputes such as the 2012 case of China’s Electronic Payment Services, which likewise addressed state control of payment markets. While Pix is a payment system rather than a currency, the investigation signals a new battleground in global commerce over government‑provided digital infrastructures. As currencies that require dedicated digital payment infrastructures, CBDCs share many architectural elements with payment systems while displaying features that render them even more vulnerable to trade scrutiny.
In earlier analyses of Pix (Part I and Part II), I examined how the General Agreement on Trade in Services (GATS) balances WTO members’ rights to provide public services against commitments to allow foreign firms into the electronic payment services market. The framework turns on the governmental‑authority exception to market access: the principle that the government may monopolize essential infrastructure that naturally precludes multiple competitors while user‑facing services should remain competitive. The exceptions also test to determine whether monopolizing markets adjacent to the essential infrastructure is necessary to achieve public‑policy goals or if alternatives that are less trade‑restrictive are reasonably available.
Understanding how CBDCs fit into the GATS framework requires recognizing the industrial organization of payment systems and what makes CBDCs distinct.
The architecture of payment systems
Traditional payment systems operate on a four‑layer stack that enables competition across multiple levels. First is money: the underlying medium of exchange existing as central bank reserves, cash, and bank deposits. Second is settlement infrastructure: systems enabling transfer of money between financial institutions, such as Brazil’s Instant Payment System (SPI), India’s Immediate Payment Services (IMPS), or the U.S.’ FedNow. Third are payment schemes: rules, standards, and brands governing how end users access the infrastructure, such as Pix, India’s Unified Payments Interface (UPI), Visa, and Mastercard. Fourth are applications: software interfaces and added services implementing these schemes, such as banking apps, fintech platforms, and digital wallets.
This layered framework allows competition where it is feasible, usually at the third and fourth layers closer to ordinary consumers, while preserving central‑bank control where it is necessary (the first layer and, usually, the second). Pix is instructive in its expansiveness. As a construct for transferring Brazilian reais, it comprises the SPI settlement infrastructure, the Pix scheme, and bank and fintech applications for making and receiving money transfers. It monopolizes de facto the infrastructure and the scheme.
Brazil made the controversial choice to legally fuse the infrastructure and scheme layers. First, it mandated that the SPI infrastructure be built exclusively for Pix, barring alternative payment schemes from using these public rails. Second, it required large banks and payment service providers to connect to this closed infrastructure. Third, it compelled those institutions to offer the Pix scheme to customers at zero pricing. While these choices created a unified national network almost instantly, they also foreclosed competition. The monopoly on the rails became a monopoly on the rules, with limited rivalry for applications and clear implications for trade commitments.
India’s UPI demonstrates an alternative model. UPI achieves comparable dominance, powering 85% of India’s digital transactions and serving nearly 500 million users, through openness on three of the four layers. IMPS, the underlying settlement system, is a public utility that any licensed entity can access and that supports alternative schemes for large transfers and business payments. Participation in IMPS and UPI is entirely voluntary, driven by consumer demand rather than regulatory fiat. At the application layer, banks, domestic fintechs, and global platforms all offer UPI‑based apps, competing vigorously on features and services.
UPI proves that achieving universal financial inclusion does not require legislating monopoly at all four levels. Superior user experience, strong network effects, and zero pricing can drive voluntary, market‑led adoption. This openness establishes a critical precedent for trade law: when a state intervenes in digital finance, it must justify any restriction beyond the minimal necessary infrastructure. Brazil’s legal restrictions on Pix were deliberate policy choices, not technological inevitabilities.
How CBDCs differ from traditional payment systems
CBDCs disrupt the traditional payment‑system stack because they merge the first two layers. A CBDC is not infrastructure for moving pre‑existing money; it is the money itself.
There are two types of CBDCs. Account‑based CBDCs, which most major economies are exploring, including China and the Eurosystem, exist as balances in a central bank ledger. These digital balances are created by verifying user identity—the principle of “I am, therefore I own.” The system for recording and transferring ownership is not a service for the money but constitutive of it. The currency cannot be separated from the account system that verifies its existence.
For token‑based CBDCs—portable digital units with cryptographic properties potentially based on distributed ledger technology, such as the Bahamas’ sand dollar—the fusion is similar. These tokens are created by authenticating the digital object itself—the principle of “I know, therefore I own.” The protocols governing creation, validation, and transfer are constitutive of the money itself. Tokens may allow greater wallet‑level separability, but the underlying protocol defining what they are remains inseparable from the money.
The key practical difference: account-based CBDCs prioritize identity verification and regulatory oversight—the system always knows who holds what—making them well-suited for anti-money laundering compliance but requiring robust privacy protections. Token-based CBDCs prioritize portability and potential anonymity—tokens can transfer peer-to-peer like physical cash—but make regulatory oversight more challenging.
In both models, the infrastructure defines the money’s existence and legitimacy. This collapse of the money‑infrastructure divide matters profoundly for trade law. Under international law, states have an unquestioned exclusive right to issue legal tender (money, at layer one). A monopoly on the foundational infrastructure of a CBDC—the core ledger and settlement engine (layer two)—is thus an extension of monetary sovereignty, residing outside the reach of trade agreements.
The danger lies in how countries extend this natural monopoly up the stack to layers three and four. While central‑bank control of wholesale CBDC infrastructure is inherent in monetary sovereignty, it does not require controlling the scheme (layer three) or application (layer four). While some early designs keep these layers separate, many are conflating them, using “sovereignty” as a shield for market foreclosure.
Distribution models and the three policy questions
There are two CBDC systems available to national governments. Intermediated systems can keep layers three and four open markets to competition, while direct systems inherently foreclose them.
Intermediated retail CBDCs use a two-tier structure: the central bank issues CBDC to financial or payment institutions, which then provide accounts or tokens to the public. China’s e-CNY and the European Central Bank’s (ECB) digital euro follow this architecture. Countries choose intermediated models not only to avoid overwhelming the central bank with retail operations, but also to leverage existing financial institutions, preserve their role in credit intermediation, and—crucially—enable the possibility of competition at the scheme and application layers.
In intermediated systems, the three-layer separation becomes architecturally feasible. The wholesale CBDC infrastructure could, in principle, support multiple retail schemes, much as India’s IMPS supports UPI and alternative frameworks for business payments. This means that even if the infrastructure is a sovereign monopoly, extending that monopoly upward into the scheme layer is a policy choice, not a technological inevitability.
Direct retail CBDCs collapse infrastructure, scheme, and application into one layer, with the central bank itself providing accounts and tokens to the public. Because the central bank itself operates the ledger, designs the rules, and delivers the user interface, application-layer competition is not architecturally feasible. Small jurisdictions such as the Bahamas and Jamaica have adopted this model.
These small island economies lack both the domestic financial institutions to serve as intermediaries and the market size to attract competing foreign payment providers. With populations under 400,000 (Bahamas) and 3 million (Jamaica), the economics of operating competing payment applications don’t work: fixed costs of compliance, technology, and support can’t be spread across enough users. The central bank stepping in directly solves a genuine market failure rather than foreclosing a competitive market that would otherwise exist.
For direct CBDCs, the trade-law question is whether this full integration is genuinely necessary for the country’s circumstances, or whether an intermediated design could have achieved the same policy goals—financial inclusion, monetary control, anti-money laundering/combating the financing of terrorism (AML/CFT) robustness—while preserving competitive space. Direct models are inherently more exposed to application-layer scrutiny because they perform functions that in conventional payment markets are commercial services subject to GATS disciplines.
Against this backdrop, countries face three distinct policy decisions, each with its own trade-law implications.
1. Infrastructure closure. Operating the wholesale CBDC ledger is a sovereign act beyond trade commitments. But if that infrastructure could support multiple schemes, restricting it to one official scheme demands evidence of genuine necessity and lack of viable alternatives, not mere convenience or industrial policy.
2. Mandatory scheme participation. Even with open infrastructure, forcing intermediaries to offer the official scheme restricts competition. UPI’s success illustrates that universal adoption can emerge from voluntary participation when the scheme offers superior functionality.
3. Application-layer restrictions. Even where scheme rivalry is limited, wallet-level competition remains possible. Limiting which apps may implement the CBDC scheme, excluding third-party providers, or imposing discriminatory licensing rules would foreclose competition that could otherwise occur. Such restrictions require evidence of genuine necessity.
The analytical test is simple: Does a restriction pertain to the sovereign monetary infrastructure, the rules of a scheme where alternatives are feasible, or the applications where competition is presumptively viable? Each layer demands a distinct and proportionate justification.
When restrictions might be justified
Countries defending more integrated CBDC models typically advance three lines of justification rooted in national context. Evaluating these claims is essential to distinguishing genuine public policy from disguised protectionism.
Technical necessity may justify single‑scheme architectures if managing monetary policy transmission, financial stability, and AML/CFT compliance across multiple schemes creates genuine systemic risks. But claims of necessity must show why regulatory oversight is insufficient. India’s IMPS demonstrates that strong supervision can preserve stability without requiring scheme monopoly.
China’s e-CNY illustrates how these arguments play out in practice. The People’s Bank of China has cited technical necessity for centralized scheme control, arguing that monetary policy transmission and financial stability monitoring require unified oversight of the CBDC scheme. Whether these justifications meet the necessity test remains an open question. But notably, even with centralized scheme control, China has integrated e-CNY with applications like Alipay and WeChat Pay. This demonstrates that scheme-level restrictions, whatever their merit, need not extend to application-layer foreclosure.
Institutional context can also matter. Countries with concentrated banking markets, limited supervisory capacity, or weaker rule of law may argue that they lack the regulatory infrastructure to oversee scheme or application‑level competition. But institutional weakness cannot become a permanent industrial policy by another name. Temporary flexibility should include sunset clauses, require capacity-building programs, and establish clear evolution benchmarks.
Nigeria’s eNaira provides a cautionary example of institutional constraints. Launched in 2021, it struggled with adoption partly because limited regulatory capacity made it difficult to supervise competing private-sector providers effectively, leading authorities to favor a highly centralized direct model. However, three years later, adoption remains under 0.5% of the population, suggesting that institutional weakness used to justify restrictions may create other problems if those restrictions prevent the innovation needed for user acceptance.
Concerns about data governance and financial stability may justify targeted restrictions. But these objectives can often be met through regulation—privacy rules, flow controls, holding limits—rather than foreclosure. Where restrictions are truly necessary, they should be proportionate and time‑limited.
The ECB’s digital euro design process illustrates these trade-offs in practice. On data governance, the ECB has debated whether to restrict transaction data access to regulated intermediaries (more privacy-protective) or permit broader access to competing service providers (more competitive but raising surveillance concerns). On financial stability, proposals include holding limits to prevent excessive flows from bank deposits to digital euros, which could destabilize credit intermediation. In both cases, the challenge is achieving policy objectives through targeted regulation rather than using these concerns to foreclose private-sector participation entirely.
Geopolitical stakes and the path forward
The geopolitical stakes are significant. A global contest is unfolding over the template for sovereign digital money. At 14.2 trillion yuan ($2 trillion) China’s e‑CNY has reached operational scale quickly, while the ECB’s digital euro debates privacy, intermediaries, and data governance with implications that will reverberate globally. By retreating from retail CBDCs in favor of private stablecoins, the U.S. has left a standard‑setting vacuum, one that China and Europe are moving quickly to fill.
For developing countries, the dilemma is acute. The need for digital financial transformation is urgent, but restrictive architectures risk becoming “too‑big‑to‑reform” as millions of users and billions in transactions lock in users and the economy to CBDC systems. Legal clarity must come before launch. Waiting for a WTO dispute may make reform economically or politically impossible.
A new international consensus is urgently needed. The Bank for International Settlements, International Monetary Fund, and WTO should jointly issue guidance establishing clear principles: that wholesale CBDC infrastructure is a matter of monetary sovereignty; that for intermediated models, retail infrastructure should presumptively be open to multiple schemes; that application‑layer competition should be protected; and that deviations must be temporary, transparent, and grounded in evidence‑based necessity.
Together with technical assistance, such clarity would offer a roadmap for central banks, legal certainty for private innovators, and a guardrail against digital protectionism.
The path forward is not state money versus markets, but a synthesis: sovereign digital currency as a secure foundation, with open competition driving innovation above it. UPI shows that public infrastructure and private innovation can be powerful complements. CBDCs present the same opportunity on a broader and more consequential scale. We must seize it by crafting rules that respect sovereignty while championing openness.
Author Disclosure: Jeff Alvares is senior counsel at the Central Bank of Brazil. The views expressed are his own and do not represent those of the Central Bank of Brazil or the Brazilian government. You can read our disclosure policy here.
Articles represent the opinions of their writers, not necessarily those of the University of Chicago, the Booth School of Business, or its faculty.
Subscribe here for ProMarket’s weekly newsletter, Special Interest, to stay up to date on ProMarket’s coverage of the political economy and other content from the Stigler Center.
