This article proposes a novel central bank digital currency (CBDC) implementation scheme from a bottom-up “exchange” perspective, achieving the dual objectives of “centralized control and distributed operations.”
Author: Yao Qian, Director of the Science and Technology Supervision Bureau, China Securities Regulatory Commission (CSRC)
Blockchain is a trusted technology originating from Bitcoin but transcending it. Blockchain innovation has not only catalyzed various private cryptocurrencies but also attracted widespread interest and exploration from central banks worldwide. Indeed, most current national CBDC experiments are built upon blockchain technology. Yet, whether CBDC should adopt blockchain remains controversial. A typical view holds that blockchain’s decentralization conflicts with central banks’ centralized management, thus advising against its adoption for CBDC.
The author contends that blockchain technology is advancing at an unprecedented pace and deeply integrating with mainstream technologies. Consequently, real-world blockchain applications—both technically and operationally—differ significantly from “fundamentalist” interpretations. How to leverage blockchain to better serve distributed operations under centralized management may be the key direction CBDC must explore today.
This article discusses three representative scenarios to illustrate potential blockchain applications and solutions in CBDC. It argues that although blockchain’s technical hallmark is independence from central authorities, this does not preclude its integration into existing centralized institutional frameworks. With thoughtful design, central banks can precisely use blockchain to effectively unify distributed operations and strengthen centralized CBDC oversight—no inherent conflict exists between the two.
Scenario One: CBDC Authentication
The author previously proposed a “One Coin, Two Repositories, Three Centers” CBDC architecture. “One Coin” refers to the CBDC itself—a centrally bank-guaranteed and digitally signed cryptographic string representing a specific monetary value. “Two Repositories” denote the Digital Currency Issuance Repository and the Commercial Bank Digital Currency Repository. The former is a database hosted on the CBDC private cloud where the central bank stores CBDC issuance funds, managed per its cash operations framework; the latter is a database where commercial banks hold CBDCs—located either in their own data centers or on the CBDC private cloud—and governed by commercial banks’ cash operations standards. “Three Centers” comprise the Certification Center, the Registration Center, and the Big Data Analytics Center.
Among these, the Registration Center records CBDCs and corresponding user identities, performs title registration, logs transaction histories, and tracks the full lifecycle—from creation and circulation to reconciliation and retirement. Its core functional components include issuance registration, title confirmation publishing, title confirmation query web application, and distributed ledger services. Issuance registration logs CBDC issuance, circulation, and redemption processes along with title ownership; title confirmation publishing asynchronously releases anonymized title information from issuance registration to the CBDC title confirmation distributed ledger; the title confirmation query website provides public online title verification services based on the distributed ledger; and distributed ledger services ensure consistency of CBDC title information between the central bank and commercial banks.
In plain terms, we leverage the immutability and forgery resistance of distributed ledgers within the Registration Center to build an “online currency authenticator”—the CBDC title confirmation ledger—offering public verification services via the internet. For current distributed ledger technology, this design offers an ingenious application approach under the central bank–commercial bank binary model—simultaneously isolating and protecting the core issuance registration ledger from external exposure while harnessing distributed ledger advantages to enhance security and trustworthiness of title verification data and systems. Moreover, since the distributed ledger serves only public query access, transaction processing remains handled entirely by the issuance registration system. By refining atomic transaction granularity, distributed computing handles transactions—thus effectively circumventing existing distributed ledger performance bottlenecks in transaction processing through business-level design. Clearly, this design fully exploits blockchain’s technical strengths to ensure credible CBDC authentication without compromising the central bank’s holistic CBDC oversight.
Notably, this dual-ledger inclusive design preserves the maturity and stability of traditional technologies while reserving space for emerging distributed ledger technologies—enabling mutual compatibility, parallel coexistence, and complementary advantages, with evolutionary competition selecting optimal solutions.
Scenario Two: Wholesale Payment and Settlement
Most ongoing CBDC experiments globally target wholesale scenarios and predominantly employ blockchain technology. Examples include Canada’s Jasper project, testing a large-value payment system built on blockchain; Singapore’s Ubin project, evaluating tokenized digital Singapore dollars for payment settlement on distributed ledgers; the European Central Bank (ECB) and Bank of Japan’s Stella project, researching distributed ledger technology (DLT) applications in financial market infrastructures and assessing whether specific functions of existing payment systems operate securely and efficiently in DLT environments. Similarly, Hong Kong’s LionRock project and Thailand’s Inthanon project both experiment with blockchain-based CBDCs. All such blockchain applications unfold under strict centralized management and supervision by central banks.
Take Singapore’s Ubin project as an example: it adopts the same Digital Deposit Receipt (DDR) model used in Canada’s Jasper project. To support DDR issuance on the distributed ledger, Singapore’s existing electronic payment system (MEPS+), i.e., its Real-Time Gross Settlement (RTGS) system, established a dedicated DDR collateral account. At the start of each day, participating banks request the central bank to transfer funds from their RTGS accounts into the DDR collateral account. These funds serve as collateral, enabling the distributed ledger to mint equivalent DDRs and distribute them to each bank’s DDR wallet—allowing interbank transfers and payments on the distributed ledger. At day-end, the distributed ledger system submits a network settlement file to MEPS+, which adjusts the DDR collateral account balance to match participants’ DDR balances on the DLT network.
Clearly, decentralized distributed ledgers do not conflict with mature, centrally governed financial infrastructure—they can seamlessly integrate and complement each other. On one hand, the blockchain-based DDR payment system offers a novel, account-independent payment method supplementing existing payment and clearing systems. On the other, DDRs represent a digital extension of electronically settled legal tender within RTGS; they can ultimately be converted back into RTGS account value and settled externally via the RTGS system—meaning the RTGS system resolves finality for blockchain-based DDR settlements into traditional account funds. This further demonstrates how blockchain settlement finality can organically integrate into existing clearing and settlement frameworks. Additionally, since DDRs are generated via 100% fund collateralization, they do not affect money supply, ensuring distributed ledgers pose no threat to central banks’ aggregate monetary control.
Clearly, from a technical logic standpoint, a central bank–led, blockchain-based new payment system is entirely feasible. In a sense, referencing Ubin’s DDR model, payment institutions and commercial banks could directly interconnect via a unified blockchain network over a dedicated financial intranet—eliminating the need for intermediaries like the NetUnion payment platform. Given that blockchain transaction performance remains under active evolution, such clearing operations are best suited for wholesale-level deployment.
It should be noted that blockchain’s “decentralization” refers to removing intermediaries—not regulators. Within consortium chains, central banks and other regulatory bodies can not only exercise centralized oversight over blockchain-hosted businesses and associated risks but also implement transparent, off-site supervision.
Scenario Three: Cash Digitization
Cash digitization appears fundamentally similar to reserve digitization (i.e., the aforementioned digital deposit receipts), differing only in that the former targets the general public while the latter operates solely among banks. However, opening accounts for the public at central banks raises a critical challenge: overwhelming service pressure on central banks and potential deposit migration leading to narrow banking.
One proposed solution is the 100% reserve-backed model. Agent operating institutions deposit 100% reserves with the central bank, after which issuing corresponding digital currency on their ledgers qualifies as CBDC. IMF economists term this synthetic CBDC (sCBDC). Under this logic, once Chinese third-party payment institutions fully deposit reserves with the central bank, funds in their virtual accounts become CBDC—making China the world’s first major economy to achieve fiat currency digitization.
Yet, closer examination reveals flaws in this approach: First, technologically, 100% reserve backing means the entire CBDC lifecycle—issuance, circulation, and redemption—must rely on traditional account systems. Cross-institutional CBDC circulation requires not only updating CBDC ledgers but also settling corresponding reserve accounts, forcing rigid system flexibility constraints (e.g., quota controls) and necessitating specialized clearing institutions for interoperability. This increases central bank system complexity and pressure—failing to alleviate central bank service burdens—and struggles to meet the “loosely coupled accounts” requirement. Second, managerially, this model tightly couples central banks and operating institutions throughout issuance and circulation, leaving central banks bearing centralized operational pressure. Ensuring agent operators do not over-issue currency post-100% reserve deposit—especially when their payment networks evade centralized oversight—makes it harder for central banks to monitor monetary issuance at the agent layer. This constitutes, to some extent, a key argument against applying blockchain to CBDC.
Perspective shapes thinking: adopting a different viewpoint yields a superior, alternative solution. Currently, many conceptualize CBDC top-down—from central bank issuance to commercial banks, then to individuals—leading to perpetual concerns about uncontrolled issuance. Physical currency issuance is constrained by printing/minting processes, but digital currency “printing/minting” occurs instantly, eliminating such constraints—the very advantage of digital currency. Viewing CBDC bottom-up reveals that end-users lack an “issuance” concept; instead, they engage in “exchange”—exchanging physical cash or deposits for CBDC. Thus, the concern over uncontrolled issuance diminishes: CBDC exchanged by agent operators represents users’ actual cash or deposits exchanged at par, not central bank-allocated issuance quotas. The central bank merely aggregates and supervises relevant data holistically. Indeed, both private stablecoins and national CBDCs currently follow demand-driven exchange—not balance-sheet expansion—a critically important distinction. For monetary policy, this implies no fundamental change; for technical architecture, it allows freedom from physical currency issuance workflows, enabling simpler system designs and dramatically improved outcomes.
Based on this bottom-up exchange perspective, a simplified CBDC implementation scheme can be proposed. The specific approach is: business originates from end customers, who apply to exchange CBDC and entrust it to agent operating institutions. Agent operating institutions maintain detailed ledgers recording entrusted CBDC for each customer. Upon receiving a customer’s CBDC exchange and entrustment request, the agent institution simultaneously collects cash or deducts customer deposits and credits equivalent CBDC to the customer’s detailed ledger. It then remits cash or reduces reserve deposits to the central bank, consolidating entrusted CBDC in bulk to the central bank. The central bank maintains a master ledger for each agent operating institution—a total-value concept—forming a two-tier dual-ledger structure with the agent’s detailed ledger. When CBDC payments occur between customers of the same agent institution, only the institution’s detailed ledger needs updating—no changes to the central bank’s master ledger. For cross-agent CBDC payments, related agents first process interactions and update CBDC ownership on their respective detailed ledgers; subsequently, the central bank periodically updates each agent’s master ledger in bulk. To improve efficiency and reduce risk, mechanisms such as continuous net position adjustment and liquidity-saving mechanisms (LSM) may be introduced.
This proposal offers the following advantages: First, it clearly establishes that holders retain full control over their CBDC. No other party may access or utilize CBDC without the explicit signature or consent of the holder—thereby endowing CBDC with genuine cash-like attributes, fundamentally distinguishing it from deposit-based money. Second, the central bank does not maintain individual records for end users; in other words, the general public does not open accounts directly at the central bank. This reduces operational pressure on the central bank while genuinely fulfilling the requirement of “account loose coupling.” Since reserve accounts are adjusted in batches, the CBDC system operates relatively independently from the RTGS system. Third, each authorized operating institution may, based on its own interpretation and within the framework of unified standards, leverage its unique strengths to build its own digital currency operating system—fostering healthy competition and enabling users to choose freely. As CBDC is issued via on-demand exchange rather than balance-sheet expansion, concerns about over-issuance by operating agents are eliminated. Furthermore, although end-user transaction data is stored only at the intermediate layer—not on the central bank’s ledger—the central bank retains the authority to request detailed information from lower-tier authorized operating institutions when required for policy or regulatory purposes, thereby achieving centralized oversight under a distributed operational model.
Conclusion
Blockchain, as an emerging technology potentially capable of serving as future financial infrastructure, supports distributed operations for the central bank–commercial bank dual-system architecture without compromising centralized management. This paper further demonstrates—through three representative use cases—that blockchain’s decentralized characteristics can be effectively integrated into both the distributed operation of CBDC and the central bank’s centralized governance framework. Blockchain technology can be applied to CBDC’s registry ledger to authenticate notes and ensure trustworthiness. In wholesale scenarios, experiments conducted globally have confirmed the feasibility of blockchain-based CBDC and payment systems. Regarding retail scenarios involving cash digitization, this paper argues that current CBDC development approaches have yet to fully realize the advantages inherent in a centrally governed yet distributed operational model—largely due to an entrenched top-down “issuance” perspective. To address this, the paper proposes a novel CBDC implementation framework grounded in a bottom-up “exchange” paradigm, successfully achieving the dual objectives of “centralized control” and “distributed operation.”
“Govern things without being governed by them”; “What transcends form is the Dao (the Way); what takes physical form is the Qi (the vessel); the Dao governs the Qi”—these are insights from ancient Chinese philosophers. Centralized governance and distributed processing have always required dialectical unity; it is inappropriate to hastily oppose institutional-level centralization with technical-level distribution. Today, central bank digital currency (CBDC) experiments leveraging blockchain technology are advancing rapidly across countries, covering broad topics including privacy protection, data security, transaction performance, identity verification, delivery-versus-payment (DvP), and payment-versus-payment (PvP). As a nascent technology, blockchain inevitably has various shortcomings—but these are no justification for abandoning it prematurely. Facebook’s Libra project is already developing a next-generation financial infrastructure built upon a secure, scalable, and reliable blockchain—a completely new frontier where opportunities and challenges coexist. As the ancient saying goes: “Sailing upstream—if you do not advance, you retreat.”