CBDC and Banks: Disintermediating fast and slow

Title: CBDC and Banks: Disintermediating fast and slow  (with Timothy Jackson and Matthias Rottner)

Abstract: We examine the impact of introducing a central bank digital currency (CBDC) in

a medium scale macroeconomic model with a banking system. As such we are able to

analyze in a unified framework several of the most pressing debates over the implications

of CBDC for financial intermediation, financial stability, and the broader economy. In

particular, our model is rich enough to allow for the possibility of endogenous runs

from bank deposits to CBDCs in times of crisis (‘fast’ disintermediation) and also a

lower frequency impact on bank funding through households’ reduced willingness to

hold deposits with banks in normal times (‘slow’ disintermediation). We are able to

quantify the welfare implications of CBDC under various assumptions on its design

(such as its degree of remuneration), the comparative advantage of banks in funding

projects (relative to households or the central bank), and on how a central bank may

use the proceeds from CBDC sales, as well as additional seigniorage revenue. Our

headline results show that, for a given degree of leverage, the introduction of CBDC

makes runs less costly for depositors and therefore a ‘fast disintermediation’ run-event

is more likely. However, the general equilibrium effect is for CBDC to produce a degree

of ‘slow disintermediation’, reducing bank leverage and the likelihood of a destabilizing

bank run. Thus, slow and fast disintermediation are inextricably linked and must be

considered jointly for a comprehensive view of CBDC’s impact upon financial stability

and welfare more broadly.