央行数字货币与货币政策：文献综述（英文版）.pdf

Munich Personal RePEc Archive Central bank digital currency and monetary policy: a literature review Beniak, Patrycja Narodowy Bank Polski, Western Australian Treasury Corporation 22 October 2019 Online at mpra.ub.uni-muenchen.de/96663/ MPRA Paper No. 96663, posted 24 Oct 2019 13:33 UTC Central bank digital currency and monetary policy: a literature review Patrycja Beniak Narodowy Bank Polski This version: 22 October 2019 Abstract Rapid digitalisation of payments leads to greater cost and time efficiency, yet could also potentially trigger legal and security challenges as well as lead to weakening of finan- cial stability and less effective monetary policy transmission. In order to ensure greater safety, central banks are contemplating and testing solutions thanks to which public using payment innovations could transact in funds that are ultimately backed by central bank. One of these solutions is central bank digital currency, a digital version of cash. The pro- posed versions of central bank digital currency are very diverse. Depending on the version assumed by a particular central bank, central bank digital currency can have an impact on central bank interest rate setting, monetary policy implementation and transmission mechanism. This relates most notably to effective lower bound which could either rise or fall, conditional on design on central bank digital currently. JEL codes: E42, E52, E58, G21, G28. Keywords: virtual currencies, central bank digital currency, monetary pol- icy, effective lower bound 1 Introduction Recent years have seen an acceleration in payment innovations. The innovations have been concentrated in facilities using private money denominated in currencies backed by the central banks, and virtual currencies, or rather cryptoassets, which are not backed by any entities (ECB The presented views and ideas should not be associated with the official position of Narodowy Bank Polski. I would like to thank the participants of 2019 Shanghai Forum as well as High-Level Conference on Successes and Challenges in the CEE Region for all valuable comments. 1 2019). The latter had the advantage of the technilogical innovation (distributed ledger tech- nology, including Blockchain), yet the drawback of heightened volatility. This has prompted development of stablecoins, pegged to one or several global currencies or to a valuable asset (BIS 2019). Stablecoins, however, share the challenges with virtual currencies, such as over- sight, legal and security risks, as well as impairment of monetary policy transmission mechanism which could resemble dollarisation of the economy, nancial stability threats related to currency and duration mismatch in households and corporates balance sheets and disintermediation of banks. On the top of that, stablecoins, particularly of global reach, could lead to heightened volatility in international nancial markets as well as to a fall of supply in high-quality liquid assets globally which could make meeting the Basel III standards harder for commercial banks (BIS 2019). Most importantly, however, growing popularity of the stablecoins could challenge the appropriateness of current payment system solutions. In order to ensure greater payment safety and nancial system stability, some central banks considered providing the individuals with solutions that would allow them to transact with funds that are ultimately backed by the central bank (BIS 2018, Adrian and Mancini-Grioli 2019). Hence, in mid-2010s, they started to contemplate the concept of issuing their own digital currency, a digital version of cash (Barrdear and Kumhof 2016). The concept is quite fresh and unstructured, which is reected in the variety of designs proposed by individual central banks. However, three features of CBDC stand out. It would be a legal tender, available solely in a digital form and issued by the central bank. Most proposals suggest that the amount of CBDC would be very limited, at least in the beginning (Barrdear and Kumhof 2016, BIS 2018, ECB 2019). However, the case when cash is eliminated completely, leaving CBDC the only central bank currency in place, is also subject to research (Borio and Levin 2017, Davoodalhosseini 2018). The paper aims to provide an overview of still quite scarce literature on CBDC, and more specically, on monetary policy-related problems. It discusses hypothetical challenges stem- ming from the appearance of CBDC for monetary policy strategy, implementation as well as the impact on the broader economy. Since, for the time being, the discussion on CBDC is dom- inated by central banks from advanced economies, the paper highlights some emerging-market economy perspectives, currently almost non-existent in literature. This remainder of the paper is organised as follows. Section 2 provides an overview of the current state of work on CBDC in central banks, as well as the design options for the CBDC. Section 3 assesses the impact of the emergence of CBDC on monetary policy strategy and implementation, depending on the design option chosen. Section 4 discusses the inuence of CBDC on monetary policy transmission mechanism as well as the economy. Section 5 concludes. 2 2 State of work on central bank digital currencies and their design In line with a BIS survey (conducted in late 2018 among central banks of countries constituting together 90 per cent of global GDP), around 70 per cent of central banks are currently studying the concept of CBDC, and the share of these banks is increasing. However, only 15 per cent seriously consider issuing any form of CBDC over in the coming years; 75 per cent of central banks might not be in position to issue CBDC due to legal constraints (Barontini and Holden 2019). The scope of studies on CBDC is in most cases purely hypothetical (Mancini-Grioli 2018, BIS 2018), with slightly more than 60 per cent of central banks currently being at the research phase. Around 30 per cent of them are already experimenting, and around 10 per cent have launched pilot projects (Barontini and Holden 2019). The most recent example is the Peoples Bank of China that said it was almost ready to launch its own CBDC. CBDC for wholesale purposes are being developed in Canada or Singapore (Bank of Canada 2018, Monetary Au- thority of Singapore 2017). At the same time, according to the BIS survey, almost 90 per cent of central banks are contemplating oering CBDC to the general public as the ultimate target of their respective projects. Most central banks cite providing safe and/or ecient payment system as their main mo- tivation behind launching CBDC projects. This motivation clearly stands out in individual reports issued by central banks being at an advanced stage of research on CBDC (Sveriges Riksbank 2017, Sveriges Riksbank 2018, Norges Bank 2018). The need for CBDC in advanced economies does not seem particularly urgent. Cash remains popular means of payment, and supervision framework is strong, also with respect to digital payments, facilitated by the central bank (ECB 2019, Danmarks Nationalbank 2017). However, as Lowe (2017) points out, in less developed emerging market countries, due to a lack of sucient infrastructure in the rst place, innovative payment solutions, developed beyond the central banks supervision, can play a predominant role in payment services. In eect, they might cause distortions in nancial system, monetary policy implementation and interest rate transmission mechanism, which is reected in compelling motivations related to these areas in these economies. In addition, emerging market economies mention the reduction in the size of the shadow economy as another important motivation behind the work on CBDC (Barontini and Holden 2019). Judging from the available literature, the idea of what CBDC should be diers very much across central banks. In line with i.a. BIS (2018), Mancini-Grioli (2018), Norges Bank (2018) and Sveriges Riksbank (2018), the division lines concern, but are not constrained to, 3 the following criteria: (i) accessibility (only for nancial institutions or for the general public), (ii) technology (account-based or value/token-based), (iii) remuneration (interest-bearing or not). The stage of development of a version of CBDC that is accessible only to nancial insti- tutions, also called a wholesale CBDC, is quite advanced in a number of economies. Such a version of CBDC is simply a technological advancement. Within such a framework, monetary policy strategy and implementation would be unchanged if central banks stopped at this stage. Yet, as highlighted above, most central banks assume that CBDC will eventually be available to the general public. In such a case, the central bank would be simply another commercial bank (Lowe 2017, Danmarks Nationalbank 2017), with a great market power, which might potentially lead to a conict of interests between monetary policy or nancial stability and commercial activity. If corporations are also allowed to set up an account with the central bank, this idea would resemble an inecient monobank system in communist Central and Eastern European economies before 1990 (Racocha 2004, Rod 2014, Szpunar 2018), even if in a very narrow form, restricted to deposits. Meanwhile, commercial banks have a greater expertise in oering deposits as well as risk management practices related to these services (Norges Bank 2018). Such reservations are particularly valid in case when the CBDCs design were account- or registered-based, with accounts held by individuals at the central bank, as suggested by Sveriges Riksbank (2017, 2018). In line with Norges Bank (2018), Sveriges Riksbank (2017) and BIS (2018), in an account-based system, nancial assets would be kept in a centralised system, preferably at the central bank or within a system controlled directly or indirectly by the central bank. The consequence is that any transactions by individuals cause a change in the central bank balance sheet, and, hence, the need for adjustment, thus complicating the central banks balance sheet management policy. The alternative to the account-based system is a value- or token-based system (Sveriges Riksbank 2017, Lowe 2017). Within this system, the public would have an access to CBDC through payment instruments, not accounts. The most commonly cited solution is that the user might for instance draw on a commercial bank deposit with an app, which resembles the conventional solutions already in place. According to Norges Bank (2018), with such a system, there should be hardly any changes in monetary policy, as the central bank balance sheet would not be directly aected. This system is also labelled as a “two-tier” CBDC because CBDC would still be channelled to the public through the nancial sector. Currently, the central banks that are contemplating the introduction of their own digital currencies are increasingly inclined towards value-based CBDC. However, the most important distinction regarding CBDC with respect to monetary policy- 4 related issues is the question of whether the account-based CBDC would be remunerated or not. The answer to that would determine the impact of its introduction on the eective lower bound, and, indirectly, the level of the ination target, as discussed in the following section. 3 Impact on monetary policy strategy and implementa- tion The sole introduction of CBDC would have an impact on the overall conditions in which monetary policy is carried out, even if, over longer run, central banks should ensure that CBDC, like cash now, is a neutral, autonomous factor for monetary policy. In line with the literature consensus, the largest impact on monetary policy would be for the account-based CBDC available to the general public. This is because, as discussed above, value-based systems and solutions restricting CBDC provision to commercial banks would be rather a technological change within the current payment system. Meanwhile, with an account- based system, CBDC would have to set conditions under which it would run the accounts for the general public, competing with the ones oered by commercial banks. The discussion in literature is centred around the impact of CBDCs introduction on the eective lower bound, which would be very dierent, depending on the remuneration of CBDC accounts. If CBDC accounts were non-interest bearing, then, as long as the policy rates were signif- icantly above zero, CBDC would simply be an autonomous factor for monetary policy. The situation would change once the central bank tried to cut interest rates below zero, and com- mercial banks tried to pass it on to deponents. Since the cost of holding CBDC is close to zero, the deposit holders would most probably switch to CBDC accounts with the central bank. The situation would hence be dierent from now when, due to costs associated with holding cash, slightly negative interest rates are possible. This means that the introduction of non- interest bearing CBDC leads to higher eective lower bound, thereby narrowing the room for manoeuvre for monetary policy. This conclusion is an adaptation of a liquidity trap concept, as drafted by Keynes (1934). A rise of the eective lower bound to zero has further important implications not only for interest rate setting itself, but also for monetary policy communication. As Norges Bank (2018) points out, the ability of central bank to inuence expectations, for instance with forward guidance, would be weaker than it is now. If CBDC accounts were remunerated, the CBDC rate would constitute the oor for the policy rates. With cash still in place, the eective lower bound would be largely unchanged. It would still be the yield on cash, i.e. slightly below zero (Barrdear and Kumhof 2016). As 5 the role of cash diminishes, the eective lower bound might even fall, since it would be easier for the central bank to impact interest rates in the economy with CBDC interest rate setting, thereby increasing the monetary policy space (Sveriges Riksbank 2018). Eective lower bound might be even eliminated along with cash, which, by allowing unrestricted monetary policy space, could prompt central banks to lower their ination targets (Bordo and Levin 2017). Although the view is not consistent with the literature consensus, it seems that for value- based CBDC, the eective lower bound would also be impacted. The scale and direction, however, would vary with time. Initially, due to a technological change in place, CBDC might be costly to hold and transact in, therefore the eective lower bound could even fall. However, as the technology becomes cheaper and more widespread, the eective lower bound could rise. Since it is assumed that value-based CBDC would be non-interest rate bearing, the eective lower bound might even approach zero over longer run, and so be higher than now. As long as cash is in circulation, a rise in popularity of CBDC could actually be conducive narrower policy space (Norges Bank 2018). This is becau