On Libra and DC/EP£¨Ï£©

¹Ø×¢ÎÒÃÇ£¬¹Ø×¢ÄϺþ»ù½ðСÕò




¢ó. DC/EP


In this part, I will first summarize public information on DC/EP and infer its design. Then I will analyze its impacts on payment and China¡¯s monetary system.


¢ó.1. Public Information on DC/EP


On Oct 24th, 2019, Chinese President XI Jinping pointed out that "the application of blockchain technology has been extended to digital finance" in the 18th collective study of the Political Bureau of the CPC Central Committee. The key to understand digital finance is to understand DC/EP of the PBoC.


So far, the PBoC has no official document to elaborate systematically on DC/EP design. Several high-level officials, including former governor ZHOU Xiaochuan and current governor YI Gang, have discussed DC/EP in their public speeches and writings. After combing and comparing information from those public channels, I believe DC/EP should have the following key characteristics:

• The aim of DC/EP is to replace M0. DC/EP is backed 1:1 by deposit reserves, pays no interest, and undertakes no social or administrative functions other than the four functions of money, i.e. unit of value, medium of exchange, method of payment and store of value. To ensure that the issuance and redemption of DC/EP don¡¯t impact the aggregate amount of central bank money, there is an equivalent exchange mechanism between deposit reserves and DC/EP. In issuance operations, the PBoC reduces deposit reserves and issues the same amount of DC/EP. In redemption operations, the PBoC increases deposit reserves and destroys the same amount of DC/EP.
  
  
  
• The circulation of DC/EP follows the traditional 2-tier system of the central bank and commercial banks. The PBoC issues DC/EP to commercial banks in a wholesale approach. Commercial banks then distribute DC/EP to the public for retail use. The PBoC works with commercial banks to ensure the normal operation of DC/EP.
  
  
  
• DC/EP takes the form of an encrypted numeric string representing a specific amount of money and guaranteed by the PBoC. It contains a serial number, amount, owner and issuing bank signatures etc. The serial number is a unique identification and unrepeatable and can be used as an index of DC/EP. DC/EP is programmable and can incorporate any user-defined executable script.
  
  
  
• The registration center and certification center of DC/EP. The registration center records all DC/EPs and corresponding owners. It also records all DC/EP transactions, including the whole life cycle of issuance, circulation, and redemption. It is basically a digital coinage center and built in a traditional centralized approach. The certification center plays an important role in the ¡°controlled anonymity¡± of DC/EP. For example, PKI (Public Key Infrastructure) can be used for authentication of financial institutions or high-end users, while IBC (Identity Based Cryptography) can be used for authentication of low-end users.
  
  
  
• ¡°Loose account coupling and centralized management model¡±. DC/EP transactions almost do not rely on accounts so DC/EP can circulate as easily as cash and can achieve  ¡°controlled anonymity¡±. ¡°Controlled anonymity¡± means that DC/EP only discloses transaction information to the PBoC. Without the permission of DC/EP users, even if commercial banks and merchants work together, they still can¡¯t discover the true identity of DC/EP users or trace their transaction history. DC/EP can expand the use of RMB abroad and promote RMB internationalization. In contrast, bank cards and Internet payment follow the ¡°tight account coupling¡± model.
  
  
  
  
• Distributed ledger technology (DLT) is not used in the registration center. In DC/EP, DLT may be used to verify the authenticity of DC/EP, providing an Internet website for external inquiries. It thus functions like an online DC/EP verification machine. This design brings two benefits.  On the one hand, the registration center is securely isolated from external attacks, while DLT improves verification efficiency. On the other hand, DC/EP transactions are not processed by DLT, thus avoiding the efficiency bottleneck of DLT.
  
  
  
• System independence. DC/EP is universally flexible. It can be transacted in all kinds of payment methods, channels and platforms and is compatible with existing financial infrastructures. In theory, whatever payment network boundary can bank deposits and electronic money reach, DC/EP can reach too.
  

The following figure is adopted from the DC/EP prototype system discussed by YAO Qian, the former head of the Digital Currency Institute of the PBoC (Figure 4).














Figure 4: The DC/EP Prototype System


¢ó.2. The Design of DC/EP


Judging from the above information, I think DC/EP does not use blockchain in the same way as the Libra. Although DC/EP is not token inside blockchain, it shares the following features with token: No ¡°double spending¡±, anonymity, unforgeability, security, transferability, divisibility and programmability etc. Therefore, DC/EP still belongs to the token model rather than the account model. At the core of DC/EP is its registration center. This center is managed by the PBoC in a centralized approach and does not need to run any consensus algorithm. So, it can be highly efficient. In DC/EP, blockchain may be used in the online verification system, therefore taking a supporting role.


Based on the DC/EP prototype system, I make the following inference on the design of DC/EP:

• DC/EP uses a centralized ledger based on the UTXO (unspent transaction output) model. This centralized ledger is embodied in the DC/EP registration center. Of course, like DLT, this centralized ledger can be structured via Hash functions and Merkel trees. But that will not make much difference because it is managed by the PBoC. More importantly, the credibility of the PBoC is significantly higher than that of commercial banks and other private institutions, so it is unnecessary to introduce DLT as a decentralized trust mechanism. I consider the centralized ledger based on the UTXO model as a ¡°degraded¡± blockchain with only one node.
  
  
  
• DC/EP wallets. User need to have DC/EP wallets to own and use DC/EP. At the core of DC/EP wallets is a pair of public and private keys. The public key is also the address where the digital representation of RMB is stored. Users can see the addresses of other users, but don¡¯t necessarily know the true identity of address owners. The corresponding relationship between addresses and user identity is known only by the PBoC through a KYC process. But KYC can be far from a strong real-name system.
  
  
  
• DC/EP transactions. Users can initiate address-to-address transactions with their private keys. Unlike on-chain token transactions which are first broadcast to a peer-to-peer network, then packaged by miners into blocks and become blockchain consensus, DC/EP transactions are processed directly by the PBoC in the centralized ledger.
  

¢ó.3. DC/EP¡¯s Impacts on Payment


Figure 5 shows flow of funds in DC/EP:




Figure 5 shows flow of funds in DC/EP


Figure 6 shows DC/EP¡¯s impacts on the balance sheets of the central bank, commercial banks and users. I study the scenarios of DC/EP issuance and redemption on the wholesale side, and DC/EP deposit and withdrawal on the retail side. In Figure 6, drr stands for deposit reserve ratio. The middle of Figure 6 shows the scenario when the PBoC issues A units of DC/EP to a commercial bank, the deposit reserve falls by A (A