The ECB’s Governing Council moved to the preparation phase of the digital euro in October 2023. This column studies the implications of the digital euro for financial stability. New survey data indicate that German households are open to the digital euro. However, the demand for digital euro also raises financial stability concerns as households seem more likely to withdraw funds from banks during times of stress if a digital euro is available. A theoretical model matched to empirical data suggests that the digital euro – if complemented with a suitable holding limit – can increase financial stability and improve welfare.
In October 2023, the ECB Governing Council moved to the preparation phase of the digital euro project, a significant step towards launching a European central bank digital currency (CBDC). In recent research (Bidder et al. 2024), we examine the potential impact of CBDC on banks and on the broader economy – key sub-debate within the broader CBDC discussion (e.g. Niepelt 2021, Bindseil et al. 2024). We approach this debate both from an empirical and a theoretical perspective.
Using survey data to explore the public’s attitude towards the digital euro, we find that a substantial fraction of German households would opt to include the digital euro in a hypothetical portfolio if given the option. Part of this demand reflects a switch out of bank accounts – the main existing form of digital money currently used by the public. In normal times, this may make it somewhat more difficult for banks to raise funds, but presumably is unlikely to cause them existential damage. In times of banking stress, however, banks may be vulnerable to runs, and the presence of a digital euro may exacerbate that risk. Why? Because it would be a riskless asset that can be easily held in large amounts – a haven in a bank run. Indeed, in our survey, households predict that they would withdraw more money from banks in the presence of the digital euro than they otherwise would have.
To explore these issues, we develop a macroeconomic model featuring CBDC and endogenous bank runs. We find that a digital euro introduced with a suitable holding limit (the model suggests something between €1,500 and €2,500) increases financial stability and welfare. The limit allows some of the benefits of digital euro as money, while choking off its effect on run risk.
No large developed economy has yet issued a central bank digital currency. As such, our survey evidence is especially useful for benchmarking household interest, in the absence of historical data on actual usage (see Smets et al. 2022 for an alternative approach to calibrating demand). Such information is vital for planning the introduction of the digital euro in pilot form. The survey is especially useful as it contains many questions on the characteristics and finances of households, not explicitly related to the digital euro. This allows us to establish what factors affect household willingness to hold digital euro.
Based on answers from approximately 6,000 respondents, we gain insight into how people might allocate funds in a hypothetical portfolio. We ask how they would allocate funds in ‘normal times’, both with and without the option to hold digital euro. We also ask how they might withdraw funds initially held in a commercial bank account in times of ‘banking stress’, again both with and without a digital euro option.
A key finding from the survey is that Germans appear open to digital euro, as shown in the portfolio allocations in Figure 1 (left panel). A substantial share of German respondents would include digital euro in their portfolio in normal times – replacing, in part, commercial bank deposits. The reallocation is hypothetical evidence for ‘slow’ disintermediation of the banking system. Since banks traditionally connect savers and those who need funding for their businesses or other purposes, the reduction in funding for banks is associated with a reduction in this intermediation role (see Ari et al. 2020 for further discussion on bank disintermediation).
During periods of banking distress, households’ willingness to shift to digital euro appears even larger, implying a risk of ‘fast’ disintermediation, as shown in the right panel of Figure 1. The dominant asset withdrawn to is cash, though the presence of the digital euro again is influential. Were a digital euro available, almost a fifth of deposits are projected to be withdrawn to it, conditional on our distressed banks scenario. This suggests that concerns about digital euro increasing run risk should perhaps be taken seriously.
Figure 1 Portfolio decisions of the households
We build a model of the economy featuring CBDC and endogenous bank runs to explore how to design a digital euro. We tune our model to match key European aggregate moments, where possible. Given the absence of data on CBDC usage, we also set up our model to match the average ratio of cash to CBDC in normal times based on our survey.
The model reveals two contrasting effects of CBDC on financial stability. First, it competes with bank deposits in its usability as money, leading to a reallocation of households’ money holdings from bank deposits to CBDC, shrinking the banking sector. Since banks in our model are prone to damaging runs, this channel improves financial stability by shrinking and deleveraging the banking system. However, CBDC may also spark destabilising ‘fast’ disintermediation, or runs. Unlike cash, a digital euro offers ‘storage at scale’ – it can be stored easily and in large amounts, meaning the banking sector loses even more funding than it might otherwise, leading to dramatic fire sales of securities. This channel leaves the economy more fragile – reducing financial stability overall. For reasonable calibrations, the second channel dominates, leading to a net deterioration in financial stability.
Holding limits are the natural policy design to counter the destabilising effect of CBDC. Correctly calibrated, they can prevent the runs to CBDC, while being sufficiently loose for households to hold adequate amounts of CBDC in normal times.
Our paper contributes to the active debate over the level at which holding limits should be set (e.g. Angeloni 2023, Panetta 2023, Deutsche Bundesbank 2023). Our euro-area-calibrated model combined with our novel survey evidence on CBDC demand suggests these holding limits should be set around €1,500, as shown in Figure 2. Under a more optimistic scenario for household demand for CBDC, the optimal limit would be closer to €2,500. Naturally, these results should be treated with all the usual caveats associated with economic modelling – especially given the novelty of CBDC. That said, our work makes a substantial contribution to realism in modelling CBDC and the digital euro specifically in a context that allows explicitly for bank instability.
Figure 2 The impact of the holding limit on financial stability and welfare
Our survey results indicate that German households are open to using the digital euro as a novel means of payment, but that there may also be risks for financial stability. Our model explores the implications of these two insights and identifies two offsetting channels through which CBDC might affect financial stability – one through reducing the size of the banking system and the other in providing an asset that is especially suited to holding in a run. If CBDC is introduced in isolation, the second (destabilising) effect may dominate. However, our model predicts that a correctly calibrated holding limit can ameliorate the run risk, while preserving the benefits of shrinking the banking system, to the extent that a CBDC, paired with a limit, may improve financial stability.
source : voxeu