Working Paper Series no. 642: Monetary Policy and Digital Currencies: Much Ado about Nothing?

Non-technical summary

This paper investigates the possible consequences of digital currencies (DCs) for monetary policy, referring to “digital currencies” as “crypto-currencies”. DCs are viewed as a combination of two elements: an asset and an exchange mechanism which allows payment and settlement through the use of distributed ledger technology (DLT). A DL is a bookkeeping system which can be shared (“distributed”) and updated in real time via a network of “nodes” (i.e. computers) in a secured manner. DC schemes operate along two different designs: public systems, which incorporate an embedded currency, like the Bitcoin, and private systems which do not necessarily incorporate an embedded currency but always involve a DL.

Motives for a wider usage of DCs could differ between the public and financial institutions. As far as the public is concerned, there are three main reasons for holding DCs: they allow making transactions can be made under pseudonyms, thus helping preserve privacy, like currency already does; they can be used to make relatively cheap and speedy payments at the global level, although this does not seem to be frequent; above all, they can be acquired for speculative motives or to evade capital controls. However, DCs currently perform very poorly the three functions associated with money (medium of exchange, unit of account and store of value), which limits their greater usage. By comparison, the use of DL technology by financial institutions, within private DC schemes, is usually deemed far more promising to the extent that the cost of validating transactions can be kept sufficiently low.

Three main scenarios in which the usage of DCs could become widespread are described and their respective likelihoods in the medium to long-term qualitatively assessed. In a first scenario, labeled A, financial institutions would use DCs only internally and between them, within private DC schemes. This scenario is seen likely, provided some technical issues are solved, and as having very limited monetary consequences. In a second scenario, labeled B, there would be a convergence of DC schemes and banking activities, with DC schemes collecting deposits and distributing credit and/or banks issuing DCs. In both cases, the monetary consequences would vary according to the services provided by DCs: with minimum payment services provided, DCs could substitute for paper currency and unremunerated sight deposits; if DL technology enabled paying interest on DCs, they could also substitute for remunerated sight deposits and term deposits. Scenario B is seen as rather unlikely or at least remote. In a third scenario, labeled C, the central bank itself would issue central bank DC (CBDC). The monetary consequences of scenario C would at least partly undo and at most reverse those of scenario B. They would depend on who can have access to CBDC but also, as in B, on the services provided by CBDC. In line with scenario A, a CBDC accessible only to financial institutions is seen as rather likely and would also have limited monetary consequences. A CBDC providing minimum payment services and accessible to the public is seen as unlikely, with more serious monetary consequences. A CBDC providing full banking services is seen as extremely unlikely as the central bank would then compete with commercial banks, with potentially also more serious monetary consequences. The three scenarios and their respective likelihoods are summarized in the table below.

In order to assess the consequences for monetary policy, a distinction is drawn between consequences for monetary policy strategy, conduct and instruments. Regarding strategy, the risks that the widespread use of DCs could inflict a deflationary bias in a growing economy, prevent the smoothing of the business cycle or cause monetary policy lose its potency are overdone since they are predicated on the assumption that DCs would be widely adopted by the public although the inelasticity of their supply makes them clearly unfit as monies. Such risks could only materialize if sovereign currencies were hit by a major credibility loss. Regarding the conduct of monetary policy, the informative content of monetary aggregates could be transitorily reduced but the definition of these aggregates could be redefined to encompass DCs, restoring their usefulness as monetary policy indicators. There could also be one-time and limited changes in the monetary policy transmission mechanism. However, CBDC would not enable the central bank to set negative interest rates more easily than today, unless currency is withdrawn or an exchange rate is set for it. Finally, also in scenario C, the role of the central bank as a lender of last resort (LoLR) could be enlarged due to the enhanced substitutability between central bank’s and commercial banks’ liabilities and, in the very unlikely case the central bank would choose to compete with the banking sector and the public would approve, the increased dependence of bank on central bank refinancing. This would call for making access to LoLR operations more rule-based in order to limit moral hazard considerations. Regarding monetary policy instruments, the central bank would have various possibilities to react to a possible reduction of the demand for reserves and could also choose to implement monetary policy by withdrawing rather than supplying liquidity. In scenario C, the central bank would also have to choose a rate of remuneration for CBDC.

Overall, the consequences for monetary policy of the use of DCs should thus be limited, unless the central bank chose to compete with the banks. However, the decision to embark into the latter project would be highly political and would not necessarily be accepted by the public.