Central banks already issue digital money, but only to banks. Reserves are electronic claims on the central bank held by a restricted set of institutions, and the public holds central bank money only as notes and coin. A retail central bank digital currency (CBDC) would change this by making an electronic central bank liability available to everyone. The classification proposed by Bech and Garratt (2017) placed such an instrument alongside cash and reserves according to who may hold it, who issues it and whether it is electronic, and the joint statement of principles by a group of central banks later set out the conditions under which issuance would be consistent with monetary and financial stability (Bank for International Settlements, 2020).

Motives for issuance vary: payment efficiency, financial inclusion, resilience and the declining use of cash all feature in official accounts (Auer et al., 2022). Monetary policy is seldom the leading motive, yet it is where the consequences may be largest, because a widely held CBDC is a new asset on household balance sheets and a new liability on the central bank's. An early report by central bank committees concluded that the implications for transmission would depend on design and could be substantial (Committee on Payments and Market Infrastructures and Markets Committee, 2018).

The question takes a particular form in economies where banks dominate financial intermediation. Where firms borrow mainly from banks and banks fund themselves mainly with retail deposits, anything that changes the price or stability of deposits changes the supply of credit. This review examines what the literature says about that link. It covers many emerging economies and also the bank-based systems of continental Europe and East Asia.

This is a narrative review of published theoretical, quantitative and policy work on retail CBDC and monetary transmission. It concentrates on contributions that state their mechanisms explicitly and that have shaped the debate among central banks, and it does not attempt an exhaustive or systematic survey. Wholesale CBDC, restricted to financial institutions, is left aside, as are the technical questions of ledger design reviewed by Auer and Böhme (2020). The review first recalls how transmission works in bank-dominated systems, then takes the proposed mechanisms in turn, and then considers the design choices that govern their strength.

The standard account of monetary transmission through banks distinguishes several channels. In the model of Bernanke and Blinder (1988), bank loans and bonds are imperfect substitutes, so that policy affects spending through the supply of loans as well as through the bond rate. The broader credit channel described by Bernanke and Gertler (1995) adds the effect of policy on borrowers' balance sheets. Evidence that the lending channel operates through bank liquidity was provided by Kashyap and Stein (2000), who showed that lending by smaller and less liquid banks responds more strongly to policy.

More recent work places deposits at the centre. Drechsler et al. (2017) argued that banks have market power over depositors, that they widen the spread between the policy rate and deposit rates when policy tightens, and that the resulting outflow of deposits contracts lending. On this view the stickiness of deposit rates is a feature of transmission and not merely a friction that weakens it. In lower-income economies with shallow interbank and bond markets the bank channel is often the only one of consequence, and the survey by Mishra and Montiel (2013) found the evidence for strong transmission in such economies to be weak. A CBDC would enter exactly at the point these accounts identify as decisive: the margin between holding a deposit and holding something else.

Four mechanisms recur in the literature. They are summarised in Table 1 and discussed below.

Nearly every result above is conditional on design. Two parameters carry most of the weight: whether CBDC bears interest and whether holdings are limited. Figure 1 arranges the four combinations. An unremunerated CBDC with tight limits is close to digital cash and leaves transmission largely as it is. A remunerated CBDC without limits is a full monetary instrument, with the strongest pass-through and the strongest risk to bank funding.

Intermediate designs attempt to keep the benefits of one while containing the other. Bindseil (2020) proposed tiered remuneration, in which balances up to a threshold earn a rate at or near the policy rate and balances above it earn a lower, possibly penal, rate, so that CBDC serves as a means of payment without becoming a large-scale store of value. Kumhof and Noone (2018) set out principles intended to prevent runs into CBDC, including an adjustable CBDC rate, no guaranteed convertibility of deposits into CBDC on demand, and issuance only against eligible securities. Distribution through banks and payment firms, with the central bank keeping the ledger, is the model most authorities favour (Auer and Böhme, 2020), and it keeps customer relationships with intermediaries.

A further consideration specific to many bank-dominated emerging economies is the presence of competing private digital money. Mobile money and large payment platforms already hold balances that would otherwise sit in bank accounts, and Adrian and Mancini-Griffoli (2019) argued that the relevant comparison for a CBDC is often with these instruments and not with cash. Where such money is widespread, a CBDC may shift balances among non-bank holdings more than it draws them from banks. The framework of Mancini-Griffoli et al. (2018) makes the related point that demand for CBDC will depend on how well existing instruments already serve users.

The literature is long on theory and short on evidence. The models differ in their assumptions about bank competition, the central bank's balance sheet and the substitutability of CBDC for deposits, and these are empirical magnitudes about which little is known. The elasticity of deposit demand with respect to a safe digital alternative has not been measured in any economy at scale. Pilot programmes and the few retail CBDCs already issued offer the first opportunity to do so, and adoption so far appears to have been modest, which limits what can be learned about large shifts.

Three questions seem most pressing for bank-dominated systems. The first is how deposit rates respond when a CBDC is introduced with and without remuneration, which would discriminate between the competition and crowding-out views. The second is whether central banks in such economies can and will recycle funds to banks on the scale the equivalence result requires, given collateral constraints. The third is how CBDC interacts with deposit insurance during stress. None can be answered from first principles, and each calls for data that central banks conducting pilots are in a position to collect and publish.

A retail CBDC bears on monetary transmission in bank-dominated economies mainly through deposits. Theory shows that the effect can run in either direction. A CBDC may drain bank funding and weaken the supply of credit, or it may discipline deposit pricing, strengthen pass-through and leave lending unchanged or higher. The balance depends on bank market power, on how the central bank uses the funds it receives, and on design choices over remuneration and limits that remain open to every issuer. Cautious designs, with no interest and modest caps, would protect the existing transmission mechanism at the price of forgoing most of the potential gains. The case for anything more ambitious rests on parameters that only experience can supply.

To be declared by the author(s) before publication.

To be declared by the author(s) before publication.