 Dear friends, colleagues, get back to your chair. It's 4.30. We should start on time because I was just telling Tony that we're doing extremely well on timing. So let's keep with that. I'm Arnaud-Nommel from the ECB and NCPR. And we are going to have another session this time on the side effects of CBDCs. That's a very broad topic. We can't cover everything. So we're going to zoom in on two aspects. One is on financial stability. And the other are the external financial speed lovers of CBDCs. And we have two great papers with Tony Anert from the ECB and Alessandro Morrow from Macca d'Italia, which are going to be presented today in that session. And we are going to start now with Tony's paper, which in fact is actually been produced by a quadruple of ECB economists. So let's see what the music is like. OK. Thanks, everybody, for coming back. And thanks to the other co-organizers, including this paper in the program. So it's a joint work with my three colleagues from the research department, Peter and Jeser and David. And it's trying to understand a bit better the implications of CBDC for financial stability and kind of banking stability in particular. So let me not motivate CBDC for this audience. So we think about CBDC as a spiritual cash in this paper. And in particular, for several years now, there's been these concerns by various policymakers to study the financial stability implications of CBDC. And in particular, it was voiced, for example, in some BIS documents that the existence of CBDC could increase the risk of a bank run, especially during crisis times or in terms of stress and turmoil. And the reason why that might be is that CBDC is a very safe store of value. So basically, as long as the kind of political order of a country or currency area is still in place, the central bank will honor its obligations. So this will be one of the last claims to be defaulted upon in the economy. And apart from this inherent safety as a safe store of value, the CBDC might also be remunerated, adding to the potential incentive of an investor to withdraw from a troubled bank and re-deposit with the central bank. So in the policy debate, several proposals have been put forward to adjust the design of CBDC in order to mitigate that bank run concern. So in particular, people think about whether and how CBDC should be remunerated. We've already today discussed quite a bit the issue of holding limits. And we haven't discussed it much today, but there's also this notion of contingent remuneration. So making the remuneration of CBDC contingent on the financial system, kind of how much stability or turmoil there is in the system. So the objective of this paper is to study the financial stability implications and to address these policy questions by dividing the consequences of various CBDC designs in the context of a parsimonious model that helps us guide and understand these various proposals. And any further, I should say that these are our own views and not necessarily the views of the ECB. And we are not part of the digital euro team, so these views need not be the views of the digital euro team. So the paper in the slide, it's a parsimonious model of bank runs because the policy concern is really about runs on banks, use global games methods to derive a unique equilibrium. And the main positive result of the paper is that remunerating CBDC, the main way of modeling CBDC is that CBDC can be remunerated in kind of a narrow sense. This would be a pecuniary remuneration, but in a broader sense, it could also be non-pecuniary benefits of CBDC. So it's a digital means of payment, so it can be used for making purchases online while cash cannot be used to buy things on Amazon, for example. So we could think about this remuneration in a slightly broader sense. So the main positive finding of the paper is that increasing the remuneration of CBDC has two opposing effects on bank fragility. So first, and in line with the policy debate for the last years, higher CBDC increases the withdrawal incentives of depositors. And therefore, bank fragility goes up. So that's very intuitive. You're worried about the bank, and now you can withdraw from the bank, and suddenly, instead of just putting it under the mattress and getting 0% normal interest rate by holding it in cash, you can now deposit it with the central bank, and you might get 2%, 4%. So that on the margin makes you more likely to withdraw from the bank, especially if you're worried about the health of that bank. That's a very robust channel and in line with the policy debate. However, and that's kind of the contribution of this paper, there's a second channel. So CBDC will be a competition for banks. So in order to retain funding, banks need to offer better deposit rates. So we look at the endogenous response of the banking system, and these better deposit rates will make the bank more stable. So if the bank offers me a good rate, so if I stay, so I'm more willing to keep my funds in the bank. So in this second indirect effect, we'll reduce bank fragility. So we have a force, a tradeoff between this direct and indirect effect, and we show that taking both of them into account, so if you want kind of moving a bit more towards kind of a general equilibrium analysis by endogenizing deposit rates, taking both of these effects into account, we get a U-shaped relationship between CBDC remuneration and bank fragility. So the immediate first policy implication is that if the relationship is U-shaped with a positive remuneration as the point where bank stability is maximized and fragility minimized, well, then policy makers who want to introduce CBDC should remunerate it. So this would be good for financial stability. Then we use this parsimonious model to evaluate several policy proposals. The first one is holding limits, and we show that holding limits have an ambiguous effect. So under some circumstances, they can be good, but in some circumstances, they can be bad as well. So the policy implication would be that we have to be a bit cautious. So it's not panacea, so holding limits may have unintended consequences. And the third policy tool that we evaluate is contingent remuneration. So the idea is that in normal times you get 4% on your CBDC, but when they are withdrawal from the banking system, so when we are starting to enter a crisis period, then the central bank will lower the return on CBDC, say, to 1%. And that tool turns out to be quite effective, and we show that this can improve financial stability. And in fact, one of the bottom lines will be that contingent remuneration, at least in our model, could be more effective than holding limits, just to put the interesting result out there, by the way. So just placing it a little bit in the literature, so this idea that CBDC competes with bank deposits and gives depositors an interesting outside option has already been studied in various papers, including by my discussant and Janet's paper in the JPE. But most of that existing literature looks at how this increased competition affects credit supply, so how it affects the bank lending decisions. In contrast, our focus is on financial stability. So there's been some work about the financial stability implications of CBDC already, usually in the context of a diamond dip big model. Our contribution is to use global games methods, pioneered by Carlson and Van Damme, and then popularized by Morse and Chin, among others. And the benefit of this approach is that it allows us to give us a unique equilibrium so we can precisely state how changes in the bank deposit rates affect financial stability, and we can precisely calculate how changes in CBDC design affects financial stability. So that's the way we hope to contribute to the literature. So the model will have three parts. It looks at the withdrawal behavior of bank deposits and the global games approach gives us an endogenous probability of a bank one, which will be our measure of financial stability or financial fragility, as it's often called in this particular literature. So this will be at the withdrawal sub game. We then move to the funding stage that banks try to attract funding by offering deposit rates when they compete with remunerated CBDC. And we look at the response of banks to the existence of CBDC. Well, when you say the response of banks, I'm really the only dimension we consider in this paper is the deposit rate. And then finally, we turn to the design stage where a social planner considers various design features, hoarding limits, contingent remuneration, regular remuneration, and so on. And we see how that affects welfare and financial stability. So we're going to use a standard global games models of bank runs. So very much in the spirit of the seminal Goldstein and Paulsner Journal of Finance 2005 paper. And the model, the particular model we're using is based on a paper by Elena Caletti and Yesa Lonello, who's a co-author on this paper and Robert Marquez in the paper they published this year in the JVE. So it's a real model. It's one good, three dates. Everybody's with neutral. There's a single commercial bank and a continuum of investors with endowment. The bank has access to a profitable investment project. So it offers a demand deposit contract to depositors to get their funding. The fact that deposits are demandable is not micro-founded. We take it as given. There are many famous theories out there why that might be the case. Also empirically, a lot of funding, both in Europe and in the US, is demand deposits or kind of has a similar structure. So we take that as a given here. So the project is costly to liquidate and yields a return r times theta in the final date. I guess I have to say here. So theta is a random variable that captures the fundamentals of the economy and r is just a parvameter that captures the profitability of lending or of financial intermediation more broadly. And then we have this deposit contract where kind of a standard diamond-dip-vick type deposit contract where you get a low deposit rate if you stay with the bank for one period but you get a high deposit rate if you stay with the bank for two periods. And then we have this exogenous lower bound on the deposit rate to capture the fact that these are demand deposits. At the initial date at the funding stage, the investor decides whether to keep their funds in cash yielding a return normalized to zero or hold bank deposits. And now with CBDC, there's this outside option of depositing with the central bank as well. So there are many features CBDCs might have. We zoom in on one feature, which is its remuneration. So CBDC can offer a positive remuneration. And in the narrow sense, this would be the pecunary remuneration that we have already discussed today, but there could also be non-pecunary benefits associated with it. As it's standard in these bank run models, at the interim date, so all of us are depositors, we will all have deposited with the one bank, say the Arno Bank. We've all deposited with the Arno Bank. And in most states of the world, the Arno Bank is very healthy, has a high realization of theta. Unfortunately, probably my time allocation is moving down to two minutes very shortly. But unfortunately, in some states of the world, the Arno Bank has a low realization of the fundamentals. So maybe it's better to get our money out and withdraw while we can. So each of us, that's the global games approach, each of us gets a private signal about the health of the bank. So that's the signal SI, conditionally independent, fundamental plus noise. And then based on the signal, each of us decides whether or not to withdraw from the Arno Bank. Let N be the number of withdrawals at the interim date. So there will be a cutoff. If enough of us withdraw, then the Arno Bank will be insolvent at the final date. You know, the withdrawals lead to costly liquidation of investment that will eventually lead to insolvency. Even worse, if more of us, if almost all of us withdraw, then Arno is liquidating the project, but it will not be enough. If enough of us withdraw, then the bank will be illiquid and cannot even meet the redemptions at the interim date. So that's the model. It might seem innocent, but given that we kind of have analysis at three dates, right, date one, date zero, and then if you want date minus one, when we think about policy design, it's not that innocent. So we would like, actually, we need to, we think we need to make simplifying assumptions. The first one is vanishing private noise. So the epsilon term is kind of the noise associated with that term is going to zero. And then there will be bankruptcy costs. So whenever the bank fails, very little is recovered. So we actually assume nothing is recovered. So the idea is you all pay it to lawyers and the process of bankruptcy is usually very costly. So empirically in the US, the evidence is that about 30% of asset value is destroyed in bankruptcy. Okay, and that's a simple two period model. So we work backwards. We start with the withdrawal behavior at date one, at the interim date. Then we go to the funding stage and determine bank deposits, bank deposit rates. And then we consider the impact of CBTC remuneration and then eventually the other design options as well. So this is just the equilibrium for different realizations of FETA. So if you are no bank has made amazing investments, we are above FETA upper bar and even if everybody withdraws, there will be enough funds for all of us. Also there's a lot of dominance regions when the investments turn out to be bad then even with nobody withdraws, there will be insolvency. And in the interim, in the intermediate region, I mean this is a version of Diamond Dipwick if you want, you have multiple equilibria where both survival and failure of the bank is possible. So it's kind of a 1980s and 90s coordination game. What the global games gives us is it removes the multiplicity in this intermediate region to a single point. So in equilibrium there will be unique FETA star such that the bank fails if and only if they realize fundamental is below FETA star. And I'm going to spare you the math but just to get the intuition of how we determine FETA star from the final equation on the slide. So these are the incentives of the marginal depositor. Remember each of us deposited with unknown and each of us get a private signal. For those of you who got a very favorable signal, you decide not to withdraw. Or some of you who got an unfavorable signal, you choose to withdraw. And there will be one person in the room who just got the marginal signal, kind of the threshold signal. This person is exactly indifferent between withdrawing and not withdrawing. So this person considers if I stay, I get the high return R2 at the final date. But I'm not always getting it. I'm only getting it when the bank survives which happens when less than and had people withdraw that was the insolvency threshold. If I withdraw, I get a low deposit return R1, but I get it more often. I get it whenever the bank is liquid at the interim date. So that's the fundamental trade-off and that we would already have in the earlier bank run work. What we are adding here are the omegas. That's the CBDC or CBDC remuneration more precisely. So we see the direct effect in blue. If I withdraw from the bank at date one and then I read the deposit with the central bank, I get the cross remuneration omega for one period. So there's a blue omega. This is the direct effect. This will make me withdraw more often, makes the expected payoff from withdrawing will go up. However, and that's what the policy makers have focused on in most of the debate, but the paper contributes is the wet effect, the indirect effect. So banks will respond to the existence of CBDC and adjust deposit rates. So the deposit rates are, will adjust. So then we just solve for the unique failure threshold. We show that indeed the direct effect is that higher remuneration makes the bank more fragile. Then also show that in the relevant part of the equilibrium, higher deposit rates will make the bank more stable. And to study the total effect of CBDC remuneration on fragility, we need to combine the direct and indirect effect, right? So that is we need to pin down deposit rates. So banks maximize expected profits subject to the participation constraint of depositors. And if I go to the bank, I get R2 whenever the bank survives. But if I go to the central bank and deposit in CBDC, I get CBDC remuneration, but now I get it for two periods. I get it between date zero and date two. So that's how CBDC remuneration affects the participation constraint of depositors and does the deposit rates offered by the bank. We make some parameter assumptions and the key result here is that the higher CBDC remuneration increases the long-term deposit rate. Remember, the long-term deposit rate was what in the relevant part was the one that makes the bank more stable. If the bank offers me more attractive long-term rate, then I'm more willing to stay with the bank. So we put the direct and indirect effect together and we get the main result, the main positive result of the paper, the U-shape. So we see financial fragility, Vita star as a function of CBDC remuneration taking all effects into account. So in here, we see that financial fragility is minimized at an interior level of remuneration. So that suggests that policy proposal to have zero remuneration on CBDC is undesirable from a financial stability perspective. So moving to the normative part of the paper and the design of CBDC, we view the central bank as a constraint planner. So it takes us given the informational friction, so the dispersed private information, the signals that we all get and the privately optimal deposit rate setting behavior of the bank. And it makes, in order to maximize utilitarian welfare. I mean, there's always the questions, what's your welfare criterion, but utilitarian welfare seems kind of the least objectionable criterion. And in this simple model, it turns out that welfare is maximized, so welfare is just expected payoffs to everybody and welfare is maximized whenever fragility is minimized. So if the central bank can pick remuneration directly, it should pick the omega min, the remuneration that minimizes fragility. But if CBDC is widely adopted, maybe CBDC remuneration will then be used, I mean, in 10 years from now, for monetary policy perspectives as well, because it will affect almost everybody in the economy. So it could be that this remuneration is not available for financial stability perspectives, for financial stability objectives. So instead, we can think about holding limits and the way we think about holding limits is that the proportion of wealth can be put in CBDC and the rest has to be put in cash, which in our model is without loss of generality because in equilibrium everybody's homogenous. So what we see basically, introducing a holding limit shifts out to you. It shifts out to you to divide and depending on where you are, this is good or bad for financial stability. So that's the sense in which the effect of holding limits is ambiguous. It will also have redistributive effects. We have several statements on redistribution in the paper. I won't have time today to talk about this. So finally, there's been a proposal to consider contingent remuneration of CBDC and in particular to pay a lower rate once they are withdrawals in the system. And this will be very effective. And the reason why this is very effective is that our indirect effect, which works via the participation constraint of depositors and the bank deposit rates is very little affected by contingent remuneration. So details on in the paper, but basically this would be very promising and perhaps a more promising approach than holding limits. There are various extensions. And to wrap up, so we offered the parsimonious model of the financial stability implications of CBDC. CBDC is modeled as an improved outside option consistent with the literature. And we consider both the endogenous withdrawal incentives and endogenous deposit rates and evaluate the efficacy of various CBDC design options. The main positive result is the U-shape, where financial fragility first decreases and then increases in CBDC remuneration because of the straight-off between the direct and the indirect effect. And we use the model to evaluate these various policy proposals. And I very much look forward to the discussion by my former colleague, Yusuf. Thanks, Yusuf. We have indeed live from China, from Renmin Moon University, Yusuf. I'm not exactly what time it is for you, but it's probably late in the night. No, if I get the time difference right. Naked, it's great to have you here and to have your discussion. Hello. Hello, can I hear me? We hear you loud and clear. Okay, that's great. So thank you very much for giving me the chance to discuss with this very interesting paper. Ideally, I hope I could be there in person and hang out with my older and new friends, but since at least I didn't work all that way, so I have to stay here. But yeah, this is a very interesting paper. I pretty much enjoy reading it. I learned a lot. So does the slide move? The slide does move. Okay, great. So let me first summarize this paper. It looks at the following question. How does the CBDC affect financial stability measured by the probability of bank failures? I feel this is a very important question related to CBDC, but in my mind, it's understudied in the literature. There are a few papers, but I think there are still a lot of things we need to know. And this paper provides a new angle basically using a diamond divix dial model to think about the probability of bank failures that hasn't been discussed before. And the model has a bank with a monopoly power and they use this global game technique to get a bank failure probability. And the conclusion is the CBDC can increase financial stability. Of course, there are all other parts of the paper, but I think that's a very important punchline. You see this argument is similar to our previous paper. It's like a discipline argument, but we were looking at the lending quantity, but here it's looking at financial stability. And in this type of argument, bank market power is crucial. Okay, so it's a great paper, very elegant. I'm gonna just talk a little bit about some simplifications in the paper that makes the model very elegant and easy to solve. There is a question, whether that would have implications on the robustness of the conclusion. Okay, so let me first reproduce the model setup. So it's a three period model in times zero, the central bank says CBDC rate or enumeration, and then bank use a deposit contract to specify two interest rate R1 and R2, that's the interest you would get in T equals one and T equals two to obtain deposits and invest in risky project, okay? Depositors then choose whether to hold CBDC or hold back deposits. And T equals one, there is a fundamental shock to the investment that's realized and depositors all observes a private signal of the shock and decide whether to withdraw or not. And then if they withdraw, they would hold CBDC instead. And then if there are people withdraws, bank have to liquidate investment to satisfy these withdrawals, and T equals two, the investment matures and depositors withdraws remaining deposits and consume. And you can see all consumptions hand present T equals two. So there are a few simplifications and deviations from Diamond Divic to make the model very tractable. And there is a question, what's the implication are these deviations? First, all consumers are assumed to be patient, they only consume and they end. There is no early consumers as in Diamond Divic model, but they are always neutral. That means essentially there is no need to insure liquidity. Okay. And implication of that, I guess that is true is that if there is no restriction, it's optimal for the bank to set R1 equals zero just to rule out any early withdrawals. But here there is an assumption R1 is at least one just at the end of the day, R1 is equilibrium equals one. Okay. A third simplification is that there is no recovery value if the bank fails, everybody gets zero. Okay. This simplify depositors value. These are all nice for making the model tractable, but what are the implications of these deviations for the result? Which ones are crucial for the results? I don't know. I would like to see more discussion on that. And of course, if there are crucial deviations there is a question, are these deviations realistic? Okay. Also, there is also another thing that's related to deposit insurance. For example, in Ski, David Ski's paper that's cited in this paper, he basically says if you have a deposit insurance like policy, there will be no wrong. I'm interested to see if there is a deposit insurance here what's gonna happen. Okay. Second comment is more about writing. I would like to see more intuition of the main result. So the main result is this U-shaped relationship, but somehow it looks like it depends on this elasticity of the background color with respect to this R2. And this elasticity is first larger than one, and it's then less than one. There is a question. Is that, what's the intuition behind that? I didn't quite get it. Is that due to parametric consumption or is that robust? I would like to know more about that. That's a second comment. And there are a few other comments, mostly about writing. For example, I think the model description can be improved. There are certain parts, wasn't clear from the text. For example, the depositors preference is not clear from the text with that. There are early depositors who want to consume early or not, that wasn't clear. I get that from there is a timeline table, but it's not clear in the text. There are other assumptions. For example, there is another one that says the liquidation value equals the R, that's the return of the acidine, and the end, if it materials, if theta is high enough, that's the fundamental. So that assumption I think needs a little bit more explanation. And I also want to see more intuition that's relating different model assumptions to the results. One thing is what I have already mentioned that R equals one is that because of there is no liquidity insurance need or not. And last thing is like how can we indoctrinate the market power in this setup? There is an extension where there is a bargaining between the bank and the depositors basically to give the depositors a little bit bargaining power, but is there a way to indoctrinate that bargaining power? That's another open question. And to conclude, I find this is a very interesting paper addressing an important question. It brings new approach to the literature. I think that's very valuable. We can talk about probabilities for bank failures. And I think it's good if the authors could clarify the effects of the simplifying assumptions. And of course, I look forward to raise the next version of the paper. Thank you. Okay. Thanks for the discussion which is not only excellent on content but it's also perfect in terms of timing because we're even ahead of schedule. So that's great. So do you want to say something already on the discussion or we can move with the questions? I mean, I would... Okay, so we have even more time for the Q&A then. So here are some questions. Timothy, I think you have a question. Do I see your right? Yes, I don't think. Hi, how should I think about the net worth channel in this case? Because normally if you think of introduction to CBDC more competition for the bank, that would mean reduced profits for the bank in this case because of higher deposit rates, reduced profits, lower net worth, more likely to fail rather than the opposite which you're saying. I see one on my right which would be your left, I think. Yeah, have you looked at the case where R2 equals R1 squared which would be kind of mirroring this omega squared kind of returning? And Derek, yep, there we go. I have problems thinking about this run out of deposits when I don't know what the central bank is going to do with those funds. So how should I think about the asset side of the central bank? So what is this omega in that context? If there's a run and people move out of deposits into CBDC, central bank presumably invests those funds somewhere wouldn't that not mean that also this omega is sort of low or maybe high if they buy this stuff in a crisis, something like this? And more generally if the omega is a parameter what I didn't get is why from a normative point of view you wouldn't set this omega as high as possible, move everything into CBDC and make welfare unboundedly high, this I didn't get. So I have Malte. Katrin now, so the question. Yeah, so I was wondering whether your simulations have some quantitative implications. So there's 105 means that is 5% remuneration of CBDC is optimal and could you think about having some quantitative interpretation maybe by looking at theta as the probability to fail or something like that? Thank you. Okay, Tony's very successful. There is a question on the fourth row in the middle. That's right. Yeah. Okay, maybe it's more a comment than a question but I think it is time to do it also because it is for all the papers that we have been seeing today. So I have the feeling like those papers are not about CBDC but they are about some imperfection in the deposit market or some fail of competition and they feel like all the paper has these interest rates on deposit and competition with deposits as the solution for something that is not working properly that could be financial stability or the transmission or monetary policy. But so if at the end what we listen at the beginning of today that the CBDC will not be remunerated. So what I have to take from all these papers and in particular for this one of competition in deposits to increase the say financial stability. It's a very deep question. Massimo, do I dare you write that there is no questions in the chat? Okay, then I don't see. And I, no. It's a bit like in an auction noise. People mostly say, okay, so I misinterpreted some of the body movements. So Tony, go ahead. Okay, awesome. These are great comments. Thank you so much. Let me try to answer at least some of them. So first of all you thank you for the excellent discussion. Please send me the slides. We'll definitely work on these kind of clarifications. I think we can do more. In terms of the relevance of the various assumptions. So I agree that market power is key and I didn't have time to talk about it in the presentation but once there's a high degree of competition the U shape breaks down. I mean in particular under perfect competition we can even show analytically that it no longer is there, right? So there's sufficient amount of market power in the deposit market is a critical ingredient for our result. Let me use this to answer or to respond to the very last comment. Hopefully it's a bit clearer in the paper but the friction, the fundamental friction is imperfect competition in the deposit market, right? I agree that we also have a survey paper where we discussed this a bit that could be direct regulation addressing that but in the absence of such regulation CBDC is one indirect measure of trying to get at that friction. The second point you that you made about the wall of the assumptions is that indeed we have we kind of have liquidity preference in implicitly in a way that consumers value the ability to withdraw but once they withdraw and they get one they're happy and then beyond that there is no liquidity insurance motive. So we received that comment at a conference last week from a discussion as well and we are currently working on another extension so it looks like we can actually do that. So we do a more like traditional diamond dip style risk-averse depositors so they value a higher short-term rate and then on top of everything I've shown you today we can then also study how a change in CBDC innovation affects the short-term deposit rate which then indirectly again will affect financial stability. So that's I think we have a set up of working this out we don't have results yet but in a new version there will be a result on this. So hopefully that will address that comment. Thirdly you discussed bankruptcy costs. The honest answer is we tried without this assumption and we couldn't get to results more than a year ago so that's why it was a simplifying assumption but I have a paper on asset encumbrance in the RFS published several years ago and there we had studied both the model with and without bankruptcy costs and then turned out that economic channels were the same. It's just without bankruptcy costs there's much more math because there are more terms to play around with but that didn't affect the results. Here we even make the stronger assumption that the bankruptcy costs even at the interim date but we have a new project with my long-term co-author Christoph Berge Agnese-Leonello who's on this paper and Robert Marquez from Davis where we look at Silicon Valley Bank and failures in West Coast banks this year in the US and trying to study that in the context of a global games bank run model and how there were kind of failures in bank risk management and there we can show that if we don't make this bankruptcy cost assumption at date one the results go through. So that gives me comfort that the results are unsensitive to the bankruptcy cost assumption. So this liquidation value improving is kind of a standard global games assumption. The global games people know that it's ugly but it has to be assumed. The rest of the profession tends to be less forgiving. So we maybe just put in the appendix. Let me just say something about deposit insurance. So implicitly, I mean these are uninsured deposits, right? So the short answer could be that some deposits are insured and some are uninsured and then the fertility comes from the uninsured part to the extent that these are not just reducing credit failures but there's kind of systemic banking crises involved. So these are kind of macro shocks consistent with the theme of the conference. Then it's likely that deposit insurance funds which in most jurisdictions are actually very poorly capitalized will not be able to cover all deposits. So even with deposit insurance there's this issue of deposit insurance being not particularly credible once you look at its funding. So in that sense, our focus is on uninsured deposits. And I think we looked VV side in AR paper that looks at US data in the 2010s and they find that roughly half of the deposits are uninsured. I'll get back to some of the comments bilaterally. Let me just respond to Katrin that we say we only have numerical examples. So this is a two-peer model we didn't do. I mean, I think we cannot do easily a serious calibration that would be stretching the model. So I think we should be humble and just say that we cannot do it. The final point to Timothy about the net worth channel. So it is a static model and they're dynamic banking models, Keeley, AR, 1990, for example, that emphasize the network channel. We do, however, so we cannot fully address this point. We do, however, have an extension with risk-taking on the asset side. And that gets at this flavor. So the bank chooses how much risk management to do and then the CBC remuneration will affect the deposit rates, will affect the skin in the game and the skin in the game will affect the extent of the risk-taking on the asset side. So we have a little bit of this on the extension but for the most part, we're dodging this. But we have a little bit. We show that we can still get our results. Yeah. Thank you so much for the comments and I'll try to respond to the remaining ones in the next break. Thank you. Thank you, Tony. Okay. Second paper. So after the ECB quartet, we have the Banca d'Italia duo. Do it. Alessandro. So good evening. Thank you very much. The organizing committee, it's really a great pleasure to be here to present this paper entitled The External Financial Spillovers of CBC written together with Valerio Nispilandi, a colleague of mine at the Bank of Italy. And the usual disclaimer applies. So many central banks around the world are working on CBC projects. Some small economies have already issued some CBC, such as the Bahamas, the Eastern Caribbean currency union, the Nigeria. Regarding the major economies, the digital reminbi is already in a pilot stage, while the ECB has started the preparation phase for the digital euro. The Fed is also studying a CBC project, a digital dollar. As we have learned also from previous presentation, the issuance of CBC might have important consequences on the issue in countries, but it might also generate important international spillovers, in particular on emerging market economies, which might experience an increase in remittances if CBC makes transaction costs lower. They might lose monetary policy independence in the presence of currency substitution, or they might face a banking disintermediation. In this paper, we focus on this third issue, and we analyze the macro financial implications for an emerging market economies of a CBC issue by a foreign systemic economy. And to do that, we set up a DSGE model for an emerging market economy that features the original scene, so basically the external depth of the emerging market economies is denominated in foreign currencies. There is a banking sector, a la Gertler Karadi, and then there are three monetary assets, cash, domestic cash, domestic deposits, and the foreign CBC that yield liquidity services but entail a cost in term of anonymity or security loss. So the foreign CBC can be more similar to cash or to deposits, so under these alternative CBC designs, we simulate the transition to a new steady state characterized by a stronger preference toward the foreign CBC. We will see which policy instrument can be used to address the negative effects of this transition, and then we'll also analyze how different levels of investment in the foreign CBC affect the transmission mechanism of a standard shock, a shock on the foreign interest rate. So regarding the results, we show that the implication of the foreign CBC for an emerging market really depends on the design of the CBC. If the CBC is close to cash, it's designed to be similar to cash, household, domestic household substitute, domestic cash with foreign deposits, sorry, domestic cash with foreign CBC with limited macro-financial implications. While if the CBC is designed to be more similar to deposit, domestic households substitute domestic deposits with the foreign CBC, and this leads to a credit crunch in the small open economy and a strong reduction in output. And we will show that there are some useful policies that can address the negative effects of the preference shocks, in particular macro-prudential policies, capital control on outflows, which in our framework are capital control on foreign CBC holdings, FX interventions can help to smooth the transition, while capital controls on inflows are far less effective. Moreover, targeting PPI inflation is better than targeting CPI inflation or exchange rate pegging in dealing with such shocks. And finally, if the remuneration of the CBC is constant, a higher level of investment in the foreign CBC can shield the economy from an increase in the foreign interest rate. So our paper is related to the literature trying to address the macro-economic consequences of CBC. Many papers are papers about closed economy models. Most of them find that a CBC competing with a monopolistic banking system can be welfare improving. Some others find that CBC could be detrimental for welfare depending on their designs. So for example, Agur et al. Find that if the CBC is close to deposit, it might lead to banking disintermediation while if it is too close to cash, it may generate the disappearance of these means of payment. And if households have a preference for a variety of means of payment, this could be detrimental for welfare. Focusing on the problem of banking disintermediation, Brunner Meyer and Nippelt find an equivalence result if deposits in CBC are a perfect substitute because in this case, the loss of the positive banking system can be compensated through central bank injections without altering the equilibrium of the model. And Burlone et al. And as a macro et al. Find that the risk of banking disintermediation can be minimized by choosing properly the remuneration of CBC or imposing quantity restrictions. Regarding the literature on open economy models with a CBC, the literature has focused on the problem of currency substitution, in particular Iqueda. Veradimineso, Melle and Straca also showed that CBC may increase international linkages, so amplifying foreign shocks. Some colleagues at the Bank of Italy, Kova, Notarpietro, Pagano and Pisani find that an economy issuing its own central bank digital currency can restore its monetary policy independence in the presence of global, privately issued stable coins. And finally, there is this paper written by Popescu in which, using a bank-run model, it shows that a foreign CBC could increase the risk of bank disintermediation in emerging market economies. So our paper is closer to this Popescu, the last one, but we address this issue considering a general equilibrium model. So this slide shows the basic architecture of our DSG. The white blocks represent domestic agents, so agents rising in the small open economy while the yellow rectangles are foreign agents. So we have households that can invest in three types of assets, of liquid assets, cash issued by a local central bank, the foreign CBC issued by the foreign central bank and domestic deposits issued by the domestic banking sectors. Houses also provide labor to domestic firms and they can also purchase bonds issued by the local government. Banks collect deposits from both households, domestic households and foreign households and they use these loans, they use these deposits to provide loans to domestic firms. Domestic firms use these loans to buy capital from capital producers. Domestic firm producers are good that is sold both to domestic households and it is also exported to the rest of the world. So now we enter into the more original aspect of our model starting from domestic households. Domestic households derive utility from consumption and this utility from labor as it is standard. But then we assume that they also derive utility from holding liquid assets, this healthy but these liquid assets entail a cost in terms of security or anonymity loss. In the next slide I will describe in detail these three new features and the maximization of household utility is subject to a budget constraint in which on the right hand side you find a labor income plus profit minus taxation plus the return on previous period position on bond, domestic bond, domestic deposit, cash holdings, the value of the position on the foreign CBDC which has and since this is an asset, a foreign asset that dominated the foreign unit of account, its value is adjusted by S which is the real exchange rate and then we have also attacks on outflows and it is important to see that here we specify a return for the CBDC but in most model simulation we set the remuneration of CBDC to zero. And on the left hand side you have how you can spend your income so consumption, new position on domestic bond, on domestic deposit, cash and foreign CBDC. Doesn't, sorry. Okay, so we model the extra utility resident households derive from holding liquid asset as a CES bundle of the three liquid assets. So cash, domestic cash, domestic deposit and the foreign CBDC. Then following a gur et al we assume that holding cash is a disutility in terms of security loss because cash can be easily stolen or lost. Well, deposit yield a cost in terms of loss of anonymity because deposit are fully traceable. So the CBDC can be more similar to cash or to deposit and so denoting with psi the degree of similarity between CBDC and cash we can write the security loss as the value of cash plus C times the value of the CBDC while the anonymity loss can be written as the value of deposit plus one minus psi times the value of the CBDC. So the banking sector is modeled following Gertrude and Karate so its bank invests in corporate loans using domestic deposit, foreign deposit and its net worth and uses these sources to finance corporate bonds to domestic firms and of course these loans are equal to capital times its value. Theta n can be interpreted as a macroprudential measure because it is a subsidy to bank's net worth. Bankers, as in Gertrude and Karate can divert a fraction theta of assets so depositors in order to trust the banking system require that the value of the bank should be greater or at least equal to the value of the invertible asset and this financial friction is important because it induces in equilibrium a spread between the lending rate and the deposit rate. So Gertrude and Karate show that in equilibrium banks choose the same leverage and this leverage is an increasing function of the spread between the lending and the deposit rate. We also assume that the foreign deposit rate depends on three components. The foreign interest rates are star which is exogenous, attacks on foreign deposit and an endogenous risk premium which depends positively on the aggregate level of foreign deposits. The idea is that the overall in-depth vis-à-vis the rest of the world the higher the return that banks have to pay to foreign households in order to have deposits from them. And the solution of the banker's problems gives an uncovered party condition between domestic and foreign deposits in which the currency premium depends on the stock of foreign deposit and on the capital flow management measure on inflows which is the tax on foreign deposits. So we calibrate the model to a typical emerging market economies time t refers to quarters and the banking parameters are taken from Makinchi and Keralto while the other macroeconomic parameters follow the integrated policy framework of the IMF. The liquidity parameters are taken from Cova et al and from Burlon et al. So we simulate the transition of the economy to a new steady state characterized by stronger preferences toward the foreign CBDC and this transition is modeled as an increase in the weight of the foreign CBDC in the liquidity bundle from 0 to 10% in 20 periods that correspond to 5 years. And this transition happens in three different, under three different scenario. A cash-like scenario in which psi is equal to 1 and in which the increase in the weight of the foreign CBDC is compensated by a decrease of the weight of cash, of domestic cash and this is the red line in the figure. Then we have a deposit-like scenario in which psi is equal to 0 and the increase in the weight of the foreign CBDC is matched by a decrease in the weight of deposits in the liquidity bundle. And finally we have a liquidity expansion shock in which psi is equal to 1.5 so CBDC is intermediate between cash and deposits and in which the weight of the deposit and cash are kept constant. So this liquidity expansion shock can be interpreted as a technology shock brought by the CBDC that increases the liquidity of households without affecting the preferences for cash or deposits. This figure illustrates the transition under the three scenario. So when the shock, the preference shock starts to affect the economy, in the cash-like scenario, the red line households substitute domestic cash with the foreign CBDC. So this leads to a depreciation of the domestic currency and also to an higher CPI inflation rate. So the central bank reacts by increasing the policy rate but also the bond rate is equal to the policy rate so these triggers an increase in the bond rate so now households increase the investment in bonds in local bonds and reduce their domestic deposits. So the deposit rate goes up. Domestic banks face an increase in their financing costs and moreover, given the devaluation of the domestic currency, the value of the foreign liability increases in terms of domestic currency and so the banks network go down and in order to preserve their profitability banks have to increase the lending rate more than one to one with respect to the deposit rate given our financial friction and this induces a fall in capital demand and in economic activity. In the deposit-like scenario, which is the blue line households substitute directly domestic cash domestic deposit with the foreign CBDC and this leads to a permanent fall in domestic deposits so domestic banks face even higher financing costs because the deposit rate increases to a greater extent and this increase is more persistent so the lending rate should go up and this depresses even more the capital demand and production. The liquidity expansion shock is in between these two cases because domestic households substitute the foreign CBDC for both cash and deposit. So now we can see which are the policy instruments that can mitigate the negative effect of the preference shock. So we focus on the deposit-like scenario which is the most interesting given its macroeconomic consequences and this scenario will be always depicted with a blue line and we will analyze different policy tools starting with a macroprudential tool so the subsidy to banks network which is the black line and this subsidy has a triangular shape because it increases, tracking the increase in the preference towards the foreign CBDC and then when the preference reaches the new state it decreases. And you can see that this policy tool is effective in reducing the negative effect of the preference shock because it limits the fall in banks to increase to a lesser extent the lending rate in order to preserve their profitability. At the same time also a sale of effects reserve is effective, the red line because it reduces the depreciation of the domestic currency. Similarly attacks on outflows which in our framework is attacks on the holdings of the foreign CBDC is useful to reduce the negative impact of the preference shocks on the banking and the real sector because this tax on outflows of course discourages the household investment in a foreign asset and so it limits the depreciation of the currency. Conversely attacks on inflows which in our framework is attacks on the foreign deposit is counterproductive which increases the financing cost for banks and in fact in the black line you see that the bank's net worth go down to a greater extent. So far we've seen what happens if the central bank reacts to changes in CPI inflation which is the blue line here. So if the central bank responds to PPI inflation it needs to increase the policy rate to a lower extent because PPI in order to stabilize PPI inflation because PPI inflation is not directly affected by the depreciation of the currency. On the other hand if the economy has an exchange rate peg which is the black line the central bank has to increase even more the policy rate in order to avoid a nominal depreciation of the currency and this explains why the PPI inflation case the red line is better in dealing with the preference shocks than CPI inflation and a peg economy. And finally the last result we see now what are the effects of a 1% increase in the foreign interest rate on the economy according to different levels of investment in the foreign CBDC. So you can see that looking at the blue line that an increase in the foreign interest rate has a recessionary effect on the economy if the small open economy has no foreign CBDC holding the blue line because an increase in the foreign interest rate leads to depreciation then to higher inflation the central bank has to increase the policy rate so higher deposit rate reduction on net or higher lending rate lower capital demand and lower production. So while if the small open economy has a medium or small value of CBDC holdings close to 2% of GDP or even better if it has an holding of foreign CBDC close to 10% of GDP what happens is that after the foreign interest rate shock household, domestic household find it convenient to sell the foreign the foreign CBDC. Why? Because the foreign interest rate has increased also the remuneration on domestic bonds so domestic household find it convenient to invest in domestic bonds and to sell the foreign CBDC. The sale of the foreign CBDC reduces the depreciation of the exchange rate and this limits its negative effect on the banking sector however this results holds only if the remuneration of the CBDC remains constant after the increase in the foreign interest rate in fact if the remuneration of the CBDC increases following the foreign interest rate increase what happens is that domestic household demand more foreign CBDC not less it contributes to further depreciation of the exchange rate amplifying the recessionary effect of the foreign interest rate. It is the green line so to conclude we set up the SG model to study what are the effect of a foreign CBDC on an emerging market economy and we have seen that if the CBDC is perceived or is designed to be more similar to deposit there is there could be a banking disintermediation problem in the small open economy a credit crunch so a lower production. We have seen that taxing CBDC holdings selling foreign state reserves and subsidizing banks can help to smooth the transition and we have also seen that foreign targeting is preferable to CPI targeting or exchange rate pagging in dealing with an higher preference toward this foreign CBDC and if the remuneration is constant if the remuneration of the CBDC is constant high stock of this foreign CBDC can help the small open economy to smooth the effect of an increase in the exchange rate. Thank you very much for your attention. Livio, it's going to be the Livio Stacca from the ACB sorry it will be the Livio was my boss that's why I said Livio, but better ok. So Livio Stacca from the ACB is going to be discussed. Ok so thanks a lot for inviting to this very nice conference. I was reminded of this conference a few days ago when I was so after playing tennis I wanted a beer so I went to the bar that only cash is accepted. I didn't have any physical cash and you know therefore before the arrival digital euro I had to give up on my beer so hopefully in a few years time the beer will be ensured. So this is a very I would say not only nice paper but also very mature paper so it's very it's full of robustness checks I will say in a moment so it was a pleasure to discuss this particular paper. How do I go down now with the like this? Ok Ok so in terms of what the paper does I mean you have seen it presented by Alessandro very very effectively so it's essentially a DSG model of a floating small economy you know sold on their kind of preferred foresight so the main exercise of the authors is to introduce a foreign CBDC so say you are in a in a small country in Latin America and people start using a dollar the dollar CBDC so whenever this happens or a digital euro maybe no so and so they find that you know this the foreign CDC in particular if it is deposit like so more similar to deposits so without privacy but with more security crowds out domestic deposits so the banks are model like Ingalls and Karate so you are probably very familiar with that model but they add some EME features so foreign currency deposits so this is what they call the original sin and also the interest rates on those deposits is tied to the economy's external debt and they also have which I don't think was mentioned by Alessandro in the presentation the dominant currency pricing which is a kind of very EME feature and there are two types of foreign CBDC which are financial like and deposit like as was explained so in terms of key results so because in this model banks are special for lending I mean this is a theme that we have seen in other papers there is a loss of output so banks cannot perform this this intermediation role as effectively as before and plus there is also a financial accelerator mechanism through currency depreciation EME features there is a depreciation side that also tightens financing conditions overall what is interesting also the output loss disappears if the EME banks can borrow from the central bank but there are some short-term costs I still am not fully understood what drives this short-term cost so plus this is a question for the authors and then they show but again this is related to the EME features I was talking about so there are some policies that can be implemented to smooth the transition in particular they find that capital controls on outflows are also selling effects reserves and targeting PPI inflation PPI inflation as this exchange rate component to it reduces the loss of output along the transition so in terms of the position of the paper in the leisure first of all is a well-executed paper policy relevant and it was quite a pleasure to read and the paper fits very much in the leisure on digital dollarization and the closest paper to it are Ikea and Popesco so I think Alessandro was very kind to mention our paper but I think it's a different issue so this is about a small open economy importing the CBDC of a large economy and so the key for the CBDC is the displacement of domestic deposits and domestic deposits are a cheap source of financing for banks so this is the idea of the deposit franchise so the banks have the privilege to issue very low interest rate liabilities because the deposits are special and so the introduction of this I mean similar to other papers in this conference the introduction of CBDC threatens that monopoly or that trend so as I mentioned at the beginning the paper is a rather advanced stage so there is not so much unfortunately they can say in terms of commenting the authors have already carried out several robust exercises so here I just kind of give you some open questions which partly are a little paper itself and partly to the implications of the paper so one first question I have is that what is exactly specific to a foreign CBDC in this paper I mean my hunch is that the same results can be derived from a stable coin so if PayPal issues a stable coin in dollars and this is using say Argentina is it different from what you have probably not I would say that it's probably not so different I mean not fundamentally different from a domestic CBDC so what if the EME so if Argentina issues its own CBDC you know how different would that be I guess the overall effects would be quite similar second question is how important are the EME characteristics so I think it's a nice feature of the paper or the model that the authors really made an effort to introduce these EME features but I was also wondering how important are they and I mean one exercise which I don't think they do but they could be done is to simulate the model without them so you could have say an advanced economy version of the model so it's more open advanced more open economy vis-a-vis EMEs more open economy and with these two partitions of the model it would be interesting to compare to compare the developed economy EME version of the models so one point that again is common to other papers in this conference is that I feel that the month for foreign CBDC comes a bit or nowhere so we know the monetary regimes are persistent so we know that even with bad monetary policy and bad central banks which of course is not the case here in Europe but in other countries where you have bad central banks the you know still monetary standard stays on monetary so it's very hard for consumers to switch monetary units so you only do it when situations are really desperate so you have to go to Zimbabwe to really where now the dollar is used as paper toilet to really change the monetary unit and so this EME economy they have in the paper is quite well behaved the central bank has inflation targeting it's well behaved so why on earth should out of the sudden that people want the foreign CBDC it's not clear so it's possible that of course the foreign CBDC has some liquidity services but there I would rather associate those liquidity services to something more like a stable coin maybe it's used in some social media but I mean foreign CBDC is not think also of well behaved small open economies in Europe so why should they out of the sudden adopt the digital euro it's not clear so why country like Denmark so why should the digital euro and where the Denmark dollar rise suddenly just when the digital euro is introduced I don't think it's so clear so this is probably it's a question which goes a bit beyond the model but I think it's a general question for the conference I mean some of these shocks to introduce is CBDC are a bit artificial so what's really happening is not clear so another question I had reading the paper is what is the effect of the transition on household utility so I don't know I mean perhaps I'm reading the paper wrongly but you know if the households want the foreign CBDC you know that's my definition I guess must be welfare enhancing so if you have an output loss you know the output loss is a price of that preference so in a sense you know the household utility is not so the output loss is just one side of this preference shift and the side that maybe has to be accepted so that's the question so it doesn't seem to be a result of an externality it's just you know given the prices and the parameters of the model you know this preference shift leads to less output and finally I mean the paper has a part I mean I mentioned this exercise that you know if the banks can borrow from the central bank you know then this descent borrow from the central bank it restores you know the bank intermediation and reduces the output loss or eliminates the output loss but I guess the broader question for the whole conference is you know when we do these shifts so banks rather than borrowing from depositors they borrow from the central bank you know why is this a problem why is this welfare reducing so one cost maybe is that you need collateral to borrow from the central bank but is it the relevant cost and it would be nice to tailor this cost to EME fissure so is it I mean I'm not familiar enough maybe with EME central bank operating procedures but you know is it difficult for EME banks to borrow from the EME central banks so this is this this part could be kind of enriched with some more kind of real world color but overall as I mentioned this is really nice paper I encourage everybody to read and I think it is a useful addition to the literature thank you Any questions from the from the floor or online from that matter? Massimo, expectedly No, sorry, I have one question about the banks in this model no so one reasonable thing to do for the banks will be to pay more deposits they can give up a bit of the profits and remunerate a little more the deposit to prevent the outflow when the CBDC is introduced it doesn't seem necessarily the case or at least they don't increase the deposit rate as much as necessary to reduce the outflow of funds I'm wondering is this let's say characteristics of how the model is solved or is something that emerged from the dynamic if I remember correctly the Geltre Karadi formalism you have these parameters that tells you how likely is that the bankers run away with the money and interest rates are kind of pinned down by this parameter friction which I guess is not changed when CBDCs are introduced so it stays there is this correct that let's say is a characteristic of the model this latency in the deposit rate or is something that emerged and have not understood correctly thank you, thank you, but it's a very nice paper Ok, another question Yeah, just I guess a policy question which is when you have the foreign CBDC in the kind of liquidity function to me that suggests that foreign households are able to use domestic households are able to use the foreign CBDC for purchasing goods in the domestic country so it would be worth thinking about exactly how easy it would be for the, say the ECB to restrict that Martin Mantler Bundesbank so if I remember correctly you don't have foreign banks in the model but I was thinking well wouldn't it be more likely that for example domestic households would hold deposits with foreign banks but this might be more costly or let's easily doable and then they switch to foreign CBDC or that domestic households could hold foreign currency denominated deposits with their domestic banks and that this would be a country way or CBDC would be a close substitute for this But I have a question from Carlos Arango from the Colombian central bank who is asking how would issuance of the EME help smooth the transition I understand issuance, how would issuance of a CBDC in EMEs I suppose smooth the transition Yeah, I mean you can interpret it That's the cost of the hybrid format I suppose Sorry, I didn't see you Sorry, I didn't see you OK, OK A very quick question You put CBDC into liquidity I just wonder in general if for example EME has better access to foreign capital market would it be easier for me to buy the bonds or stocks in U.N. U.S. or more advanced countries would that be different I just wonder whether the effect is going to be similar to CBDC or what is so special about CBDC relative to other investment opportunities I'm actually impressed how energy the room still is Just a very quick thought and shooting from the hip and I don't know this class of models very well, but it seems that at the core of the banking part there's kind of a more hazard friction like Homestorm T-roll so that you need to ensure no absconding and then the CBDC might actually directly affect that friction as Governor Panetta said earlier there's no access, no access, no access this is really strong privacy protection then it might actually allow you to abscond with funds more easily my larger point is there might be interaction between CBDC and the friction that is at the core of the model Anyone? OK, burden Thank you very much for your questions I start from the last one also because it's the one that I remember Yes, there is there is an agency problem in the banking system so basically the bankers have to increase their profitability in order to convince the depositor to trust and actually there is there is not an explicit interaction with the CBDC in that specific in that specific sector because for banks CBDC is a dominated asset because its remuneration is lower than corporate loans so they invest in corporate loans but we can once can think a way to incorporate also an investment in CBDC by banks but there is an implicit interaction because the higher preference towards the foreign CBDC induces a depreciation of the currencies which makes for banks harder to repay the foreign debt so it's an interaction through general equilibrium Exactly, it's a general equilibrium effect then there is the issue of bonds we have an extension in which we allow domestic households also to invest directly in foreign bonds or if you can think general foreign assets the model results do not change if these assets do not enter in the liquidity bundle otherwise but of course the idea of the liquidity bundle is that this is a reduced form to capture that these assets are immediately available for payment so the alternative will be to have a cash in advance constraint but for tractability purposes we use this reduced form approach and we include that only very liquid assets then the issue of deposits it's true how you describe the Karate model but the idea here is that the stronger preference for the foreign CBDC reduces directly the supply of savings from domestic banks from domestic households to domestic banks in the case of deposit like scenario and indirectly in a cash like scenario because there the causal chain works in a different way but there is substantially a drop of the supply of savings from domestic households to domestic banks then if you don't have a Karate so the banking system is just a deal you have a one-to-one increase of the lending rate with respect to the deposit rate with a financial friction you have a more than one-to-one reaction of the lending rate with respect to the deposit rate so because domestic firm can convince the deposit to trust the systems then I go to leave your comment thank you very much again yes I think you are right it's not so it's not to clear how to distinguish between foreign CBDC vis-a-vis a foreign stable con maybe one can play with the security weight because one can think that a foreign CBDC is more secure than a foreign stable coin because it is privately issued versus a public form of payment it will be interesting to see a difference between a small open economy model for an advanced economy and a small open economy for an emerging market economy we focus on the emerging market economy because it is easier to justify the presence of a foreign CBDC and the absence of a domestic CBDC the relevant aspect of our model more than the dominant currency pricing which we have introduced just in the last version of our paper because the referee asked to do that is the original scene this is a feature that can distinguish a small open economy model for an emerging market than a small open economy model for an advanced economy so here is where I expect to see different dynamics but in particular my guess is that the macro financial implication on an advanced economy will be smaller given that the problem of the foreign denomination of that is less crucial then of course I do not have the time to show but we have also an extension on what I will conclude on what can happen if the domestic central bank provides some liquidity injection to the domestic banks and this is another policy tool that will effectively address the negative effect of the shocks in line with Bruno Meyer and Nippel paper but we are not able to find an equivalence result in our framework because there is also these negative externalities linked to the external that denominating foreign currency basically as it is standard is a small open economy model issue foreign debt and they ignore that they have a negative effect on the other banks because they contribute to increase the overall in depth with the rest of the world and this leads to an increase in the rate on the foreign deposit thank you very much