 It allows me to pass on to the next paper. So, yep, indeed. So the speaker is gonna be Anatoly Segura from Bankad Italia and we're turning to the cross-border banking and the ring fencing. And it's very nice to see that this is a paper that not for you this time, you have exceeded that age, but there was sponsored by us for under the Ranfalusi Fellowship Program for your young co-author, Ying Zhang. So that's very nice to see the success of this paper. Close yours. 20 minutes, please. Thank you. Thank you very much to the organizers for inviting us to present the paper. We are particularly delighted to present the paper in this institution, in this Euro Tower, as opposed to the other big tower. And as for reasons that will become apparent from the very first slide. So this is joint work with Ying Zhang, who, by the way, is sitting there physically and also joint work with John G. Lawrence, who is sitting virtually somewhere at the Webex platform. So let me stress that the usual disclaimer applies. These are not necessarily the views of the Bank of Italy nor the Euro system. So let me motivate the paper. So as you know, cross-border banks frequently use an organizational structure that is operating through subsidiaries. And that gives them a lot of flexibility when it comes to the allocating internal capital within them. That flexibility in particular means that they can let fail a subsidiary that is in distress. And they can also voluntarily decide to support it from the parent. Now, national authorities don't like so much that flexibility. And sometimes they tend to limit those cross-border capital flows, particularly during crisis times, in order to defend what is called national interest. And people usually refer to those actions by the authorities as ring fencing. Now, during the global financial crisis and also the European debt crisis, there were many, many episodes of bank distress and there were some examples, both of voluntary support and of ring fencing. So this generated a vivid open, and I think to some extent, still open ring fencing debate. And basically the overall idea is that ring fencing somehow constraints the efficient capital allocation within a group and tends to lead to some suboptimal resolution of banking crisis. But at the same time, we've seen that despite the awareness that ring fencing can create problems, that is little progress in supervisory coordination to remove local ring fencing powers. And this applies even to the Euro Area Banking Union as is highlighted by the very frequent speeches by Andrea and Ria and the Elkaconic, in which they kind of complain that ring fencing is still an issue in the Euro Area Banking Union. So given this motivation, what this theory paper tries to answer is the following questions. So we want to understand when that ring fencing arise. And what our answer and something very important, we are going to be emphasizing and it's not very surprising, but we think it has been somewhat overlooked by the policy debate and also the literature is that correlation between the assets of the cross-border bug units is very important. Well, we find this that for large correlation and under national authorities, there is going to be ring fencing when a cross-border bank is in distress. Now, a supranational setup with very strong powers is going to remove ring fencing, but in the end ring fencing is affecting the capability of the cross-border bank to share risks between units. So affecting that those research impossibilities through the supervisory setup is going also to have an effect on the cross-border bank restaking. So our answer, so we ask ourselves what is the effect of different choices of supervisory setup on cross-border bank restaking? And our answer is that it depends and we will emphasize different channels. And finally, all this to get to the final and more important question is when is the supranational authority beneficial from a welfare perspective? And our answer will be that not always, but yes, when there is large correlation, positive correlation across the assets of the cross-border bank and this cross-border bank is also risky. So in a way, we will say maybe the Euro Area Banking Union was the right response through the centralization of decisions to address a ring fencing problem that would have been important in the case of the Euro Area. But then we have the flip side of the answer is that since not always it's beneficial, in particular, there is no need to care about ring fencing or supranational coordination when there is negative correlation between the assets of the cross-border unit. And this is precisely the case when the cross-border bank is going to create more value. So in a way, this is a reassuring message that when cross-border banks create more value, then ring fencing is not a problem. So let me get to the model. There are three dates and there is universal risk neutrality. This is a very stylish model. And there is a cross-border bank and there are some authorities, supervisory resolution authorities. The cross-border bank has two subsidiaries located in two countries and each unit has a risky asset and some deposits that are insured by a national deposit insurance fund. And the structure of the bank is a pure bank-holding structure. So you have a pure bank-holding company and two subsidiaries and the bank-holding company holds the equity of the group. Now, the assets of the bank, they have some payoffs, an interim payoff that is free-slash at equal one, which is small r and a risky payoff at the final date, which is capital r in case of success and zero in case of failure. And now, the successful probability as of T equal one can be high. In that case, we say the unit is healthy and or can be low. In this case, we say that the unit is in part. Now, as I said, correlation across the units is important. It is captured in the model by the correlation parameter rho that captures both the correlation at equal one and the correlation at equal two. And just after some normalization that you have to bear in mind, rho equals zero means independence, rho positive is positive correlation, and as you get to one is maximum positive correlation, rho negative is negative correlation. So now, let's start with the interim date and let us think about what national supervision is. So the way we model national supervision is what we say that is one national authority that curves about each unit and that at the interim date can intervene in order to minimize the on deposit insurance costs. Now, the important interesting case is when that authority sees an impaired unit and she can take some sort of early intervention that we call liquidation, but it's not necessarily that. And it's an action that for the impaired unit is going to produce the deposit insurance costs. So these are the deposit insurance costs in case of early intervention on an impaired unit and this would be the deposit insurance costs in case of no intervention where you see the probability of default and the loss is given default because these losses are costs for the deposit insurance fund. Now, what we assume is that the early liquidation or the intervention reduces the deposit insurance costs. So this authority would like to intervene. Intervention destroys some asset value. So it is something from a welfare perspective and also from the perspective of the bank, the bank is not going to like. So what the bank can do is in order to avoid the liquidation of the impaired unit is to recapitalize it. And there are two ways of recapitalizing the unit, either with some issuance of equity, external equity that is costly, both privately and socially, or with what we call the internal capital market from which is or support in case the other unit is healthy. Basically, the other unit can inject some funds in the form of an intra-group loan. Now, the bank is going to prefer cross-unit support in order to save on the external cost of capital. But that is the other authority, the authority that is responsible for the healthy unit that might not like that transfers of resources from the healthy unit to the impaired unit. And in case she introduces some limits to the support, we will refer to as a reinfencing action. So in terms of modeling, how we are modeling this game between the bank and the two authorities in the interesting case in which one unit B is important and the other is healthy. So the cross-border bank is going to propose a recapitalization plan that consists of how many units X of external capital are going to be raised and injected in the impaired unit. And also the cross-unit support, which is a loan that transfers small s of units from the healthy to the impaired unit against of a junior promise capital s at the final date. Now, each authority says yes, I agree, going forward or not. And if there is unanimous approval, the plan, the recapitalization plan is undertaken, otherwise there is the liquidation of the impaired unit. So now let us look at the approval decision given a recapitalization plan. So now the authority responsible for the impaired unit is going to say, okay, if the plan is not approved, I am going to liquidate. This will be my cost of liquidation. If the plan is approved, this will be the expected cost. There won't be default today, but with some probability there will be default tomorrow. The injection of capital, both external and internal capital is going to reduce those costs for the depositational cost. So if the injection of capital is sufficiently large, no matter whether external or internal, that authority will say yes, I agree. Now let us look at the other authority, the authority of the responsible for the healthy unit. And here and everywhere in the paper, what is blue, in blue we highlight things that are good from the perspective of who is taking the decision and in red things that are not so liked. And what you see again, that authority is going to look at the deposit insurance costs under no recapitalization and deposit insurance costs in its own country under recapitalization. And the authority responsible for the healthy unit, this likes the injection of funds from the healthy to the to the to the per unit. This is because in case of default tomorrow of the healthy unit, which can happen, there will be less resources in order to repay deposits. So that is why this is in red. But the authority realizes that the support is not just a gift, it's against a loan. And the loan whose promise repayment is capital S will be repaid in some contingencies. That is why it enters with the blue. Now correlation matters and it enters with the red because when correlation is high, it is going to be more likely that the two units fail or succeed at the same time. So from the perspective of the authority that cares from the deposit insurance costs, the value of the intra-group loan is going to be very low because the intra-group loan with very high positive correlation, the intra-group loan will only be repaid upon success when the two units are succeeding and then the value of the loan will be appropriated by the shareholders of the bank holding company, not in a way the deposit insurance fund which is the creditor. So this gets us to our first result when we have national supervision, the outcome of this resolution of an input unit when the other is healthy is low correlation. This means independence, some slightly positive correlation and very negative correlation. Low correlation, the recapitalization will be fully via cross-unit support. High and positive correlation, there will be some reinforcing and some limit on how much support the healthy unit can give to the input unit and the bank will be constrained to issue at the parent level or the holding company level. Some step-by-step. Let us now think of what the strong supranational authority that is responsible for the two units would do. Now, the mandate of this supranational authority, we assume is that of minimizing the overall deposit insurance costs. So it cares about the average or the overall deposit insurance costs. It doesn't care about potential redistributions between the two deposit insurance funds. So it is going, the supranational authority is going to approve a recapitalization plan if when adding up the two authorization constraints, we were seeing in the previous slide on average or on aggregate, that constraint is satisfied. And let me highlight here, again, blue, good, red, bad. Let me highlight again how this authority is seeing the support from the healthy to the input unit. So she realizes that support is good in terms of reducing deposit insurance costs in the input unit. But she also realizes that it is bad in terms of increasing the deposit insurance costs in the healthy unit. But the overall effect of this is not a pure redistribution because the cross-unit support reduces overall deposit insurance costs. And this is because in a way, funds are flowing from a part of the bank holding company that is safer to a part that is riskier. So those funds will contribute to pay the deposits with higher likelihood when they are in the input unit. So in a way, by providing voluntary support, the bank holding company is foregoing some of the limited liability protection that it has. The supranational authority is aware of that and it is going always to when there is such authority, we will have that the outcome of the game is that the input unit is always fully regapitalized through cross-unit support. There is never reinfencing. But importantly, this is going to give rise to some redistribution across deposit insurance funds as a solution to the crisis. Now, for local relation, national and supranational supervision is the same. Because there is no reinfencing, but for high and positive correlation, the outcomes are different. And the cross-border bank is going to get more value from the resolution of the input unit when there is supranational supervision. Why? Because there is no need to raise external capital. So this takes us or takes me to the second question, which is okay. We've seen that the different supervisory setups are going to give rise to different ways of solving this stress at the interim date, different resharing possibilities within the cross-border bank. What are the restaking effects of this? And in order to address this question, what we do is to endogenize the risk of the units. And we assume that the state of each unit at equal one depends on something endogenous, that is some costly effort, that the manager of the cross-border bank exerts at equal zero in each of the units. And it depends also on some exogenously on something that we call the strength of the fundamental riskiness of the bank. And we focus on large correlation because it is when there is a discrepancy between the two supervisory setups. And basically, in this expression, what you have is what is the value that the bank-holding company is generating from the possibility of support. And it is this, with some probability there will be cross-unit support and the cross-border bank will appropriate some gains from being able to provide support. And these gains will depend on whether there is a national regime in which those gains will be lower because there is a freelancing or a supranational in which those gains will be maximized because there is no freelancing. And what we find is that, I know I'm running a bit out of time, but there is a buffer, a time buffer from the previous paper, which should be usable because we have a stress that buffers have to be usable when needed. And not just... Okay, so there are two effects. I won't spend too much time, but by removing re-infensing through supranational supervision, the cross-border bank realizes that there are more support gains. And this has two counter-valuant effects. There is a nice and within novel effect that fosters incentivizes effort, which is what we call the Charter Value Effect. If I know I can give support to a distressed unit, I have more incentives to be in a position to provide support. This effect, for this effect to arise, it is necessary that the decisions are taken at the bank holding company level. It's someone that internalizes the value at the entire group of taking good actions at the unit level. And then there is also a more standard, like free-riding effect when you have some, the more resharing possibilities that are, the more incentives that are to take risk. But the two are pushing in opposite directions. What we show is that depending on how risky for fundamental reasons the bank is, one is going to dominate or the other. And this, once we take into account what are the exact risk-taking implications of the different supervisory setups, we can address the last question, which is when is beneficial one setup or the other? And what we find is that for weak banks, these are like banks that are fundamentally risky, the Charter Value Effect dominates. This is the effect that was, when you introduce a supranational supervision, it's going to push towards more effort, so less risk. And so there will be less risk with supranational supervision. So this is good ex ante. And also exposed removing, re-infensing is good. So ex ante good, exposed good. So we will have that welfare is going to increase and deposit insurance costs are going to get reduced when we introduce the supranational authority. Things could be different when the banks are very safe. So somewhat paradoxically, you'd like very safe banks not to be supervised centrally. But I mean, this is a very simple model, very stylized by my, we had to say where the Euro area banking union seats probably we are sitting here or we were sitting here over the past decade. And so supranational supervision is something good. And then to conclude, just let's get back to the issue of the correlation. High and negative correlation that is not going to be re-infensing exposed. And from an ex ante perspective, there will be many situations in which an input unit will be able to be supported by a healthy unit because of negative correlation. So in these cases, the cross-border bank creates a lot of value through that internal capital market and re-infensing does not arise. So national authorities, there is no need of coordination when resharing within the group is more important. And for the higher positive correlation, there is this tension. When re-infensing is exposed very severe, it will be from an ex ante perspective kind of unlikely. Okay, because that is very high correlation. So in a way, the gains from supranational supervision are somewhat limited. Because tensions exposed mean from an ex ante perspective, those tensions will not arise. So it's not so likely that they will arise. And let me conclude. Let me conclude. Just very simple framework of financial restructuring and history of cross-border banks in which we in the generalized cross-unit support, the possibility of re-infensing and restricting how it depends on different supervisory architectures. Contribution to literature, which is contribution to the paper most related to our paper is the paper by Martin M.K. and Patrick Bolton on the resolution of global banks. We stress the important role of asset correlation in financial restructuring. Policy takeaways, the most important one, in the Euro area banking union, supranational supervision is goes in the right direction, in the direction that the model predicts welfare will increase and the deposit insurance costs will decrease. In other, for other cross-border banks that have operations, more global operations so that correlation could be lower or even negative, maybe re-infensing is something we don't have to care so much about. That's it. Thank you, Anatoly. You used your liquidity buffer to the last drop. Discussant in Washington, DC is Angela Madaloni. Angela, can you hear us and see us and speak? 10 minutes is your time budget. Okay, I made it. Sorry, Philip. I couldn't unmute myself. Hello to everybody. It's very nice to see a lot of friendly face and I'm really sorry I cannot be there with you. I hope you see the slides well and I hope you can hear me well. Okay? Yes. Okay, great. So, I mean, of course, the usual disclaimer apply because I mean, we all work in policy situation. Since we don't have enough, a lot of time, let me start. So what does this paper do? Well, they provide a framework to analyze basically different supervisory architecture and it is important that this framework, to emphasize that this framework compares national supervision and super national supervision in a very focused dimension, which is when considering intervention in cross-border banks. I mean, as Anatoly very well explained, I mean, there are cross-border banks that are subsidiary into different countries. And the other main point, which I think, and it was also said by Anatoly, I mean, the main inspiration to this approach is to look at the issue of ring fencing of assets. And a lot of us that have been working in central banking for a long time know exactly what happened during the great financial crisis. I mean, there were a lot of instances and a lot of talks about the consequences of this ring fencing of assets across subsidiary of the same group. I mean, the model is very complex. I think you have seen it and there are a lot of implications. In terms of ring fencing, I think the important point, which was also emphasized is that with sovereign national architecture of supervision, generally there is no ring fencing. And then there is, and here I use maybe a bit of a terminology that I see in the paper, there is this increase in this convergence of bank assets. And then ambiguous impact on this taking. I mean, Anatoly talked about this. I think it didn't stress so much this implication on the convergence in the area that is supervised by the Supernational Authority. And there is this idea instead that the Supernational Architecture would induce this convergence of the default risk among cross-border banks. Okay. Why is this important? I mean, just want to mention it. I think this idea of ring fencing is something, I mean, this was also mentioned also in previous talk. I mean, there is this idea, even now in the Euro area, with a somewhat sovereign architecture, which is the SSM, there is the idea that there are a lot of limitations to how much capital and liquidity can be transferred across a sub-series. And indeed, I took this chart from a speech, a previous speech of Andrea a couple of years ago. And here, I mean, he really talks about non-transferable liquid assets. I mean, in non-domestic subsidiaries in Europe as trapped liquidity. And this is like very important because there is this idea that this trapped liquidity is somewhat inefficient in the system. Okay. So I'm having some issue. Okay. What are the key features of the model? And I hope you realize as I said before that the model is very complex, but I'm gonna talk and focus on two main points. One is the riskiness of the assets. So the riskiness of bank assets was the parameter gamma in the model. And then there is the risk-taking aspect, which of course in the model is linked to the bankers' efforts, which is induced by a certain setup in the supervisory architecture. And then there is the correlation of assets payoff. And then Natalie talked a lot about this, which is generally exogenous in the model. Okay. I want to say that these models provide a framework to analyze the implication of different supervisory architecture in a context, in a very specific context, which is the context of recapitalization and resolution. So resolution is not mentioned so much in the paper, but at the end, I mean, if there is no recapitalization, banks are resolved. And this is why all the implication are actually linked to of course the existence of the national deposit insurance. However, I was thinking that, I mean, supranational supervision has a lot of other dimension. It is not only this dimension that basically is important when there are decision of recapitalization to be taken. And I think this model, and this is not a critique, it's just a fact that does not provide insights about the role of supranational supervision, especially in any, as an institution to increase resilience exactly. And this is, then I went to why the SSM was established a bit in the Euro area. And just to be clear, I mean, the three reasons why it was established, I copied and paste from the website of the SSM is the objective of the SSM are ensure the safety and soundness of the European banking system, increase financial integration stability and ensure consistent supervision. So you can see there is a lot of emphasis on the ex-hunteries taking and resilience of ensuring the resilience of the financial, the banking sector before. And this is something that, I mean, I think the way the setup of this model cannot really tell us much, okay? This is important to take into account. Now, let me talk about the restaking aspect. Now, as Anatoli explained, of course there is this sort of ambiguous effect of supranational supervision on restaking the aspect. Now, the point here is that on this, actually this is an aspect on which we have a lot of empirical evidence right now. We had evidence in other regions, like in particular in the US, but more recently we have a lot of evidence for the Euro area linked of course to the implementation of the SSM. So we know that when the SSM was implemented, banks decreased their exposure, decreased their lending, decreased their asset side, their reliance on also funding, they increased the risk-weighted assets. And also, I mean, during the pre-SSM period, banks that were around the threshold of the assets to be supervised by the SSM or not, I mean, some of them decreased their assets in order not to be supervised by the SSM. So what I'm saying here is that there is actually quite evidence that banks perceive supranational supervision as tougher supervision and therefore somehow they are taking action in order to deal with this, new tougher supervisor. I'm also mentioning here some new work in progress than I'm doing, which is related to the TRIM, which is the internal review of, sorry, the SSM review of the internal model that was done a couple of years ago. And here again, the announcement of the review, I mean, we find that banks increased the risk-weighting of the risk-aid exposure, land to safer borrower and recover more from defaulting entities. So overall, I think there is quite some evidence that supervisory, I mean, supranational architecture is linked to a general decrease in the risk of the asset, which of course is a bit difficult to square with whatever you have on the paper. I mean, this is a point that I wanted to make. The second point that I wanted to make is the implication on payoff correlation because as you said, I mean, this is a key parameter of the model, the correlations of the payoff. And when the national supervisors, sorry, national supervision and supranational supervision is more or less efficient or optimal is linked to the correlation of the assets. And okay, here, you said it only at the end, but I think you mentioned these results, which in a way really struck me. In general, if the correlation is high, I mean, this is my second point, these are the results from the paper. Supranational architecture maximizes aggregate welfare, okay? However, these welfare gains are home shaped in asset correlation and they are limited. And Nanatoly mentioned this at the end of his presentation. And something that really struck me is that in the paper, the authors write, so avoiding the infencing does not per se justify supranational supervision. So the idea is, if we just have in mind to avoid the infencing, maybe we should consider that there are limits to the benefits arising from supranational supervision. Now, I've been thinking, what is affecting pace of correlation? Since correlation seems to be so important, okay, what are the factors that are affecting the correlation of payoffs? And I've been thinking that definitely convergence of supervisory criteria, increasing integration of banking market, increasing possible MAE cross-border, of course, are all factors that are likely to affect the correlation of payoffs of assets of banks in different areas. Now, let me skip here. Europe banking markets remain very segmented. I mean, these are charts that I took from the latest financial integration report. And they're also, as was mentioned also before, I mean, there is this general call from policymakers to try to achieve and to increase integration in banking market. So there is a general call to remove barriers, legal and prudential that creates obstacle to the freedom of movement of capital and liquidity within banking group. And also there are general call to increase standardization and the consistent implementation of regulatory standards. So I've been thinking that these calls are gonna have consequences on the correlation of payoff of bank assets. And therefore, I mean, the way we can look at the distinction between national supervision and financial supervision in the context on the model is gonna be affected, but what is really going on, for example, in Europe. Okay, so I've discussed basically these two aspects, which I think are particularly important, the risk of the assets and the correlation of payoff. Now, let me conclude with a couple of points that I wanted to make, which are not directly related to the paper, but I think they are related to the issue of national supervision and super national supervision, which I think they're important. So the first one, which is that, as I said before, there are a lot of other dimension to super national supervision, which are not taking into account because the model looks at a very specific dimension. And one thing that I was thinking is actually important also in the current environment that we are now living through, I would say, is that super national supervision may make it easier to assess the interaction spillover between regulatory policy and other macroeconomic policies. And this is not something that is just conceptual. This is something that, for example, happened during the COVID-19 pandemic. Okay, that said, it was like crisis management mode, but I mean... We need slowly to conclude because there's not a fortune... I'm concluding, I just have two points. So we have taken unprecedented supervisory decision quickly in close coordination with monetary policy measures. That was said by Andrea in a previous speech. And this is, as I said, seems to me very relevant also in the current financial and economic environment. And the last point that I want to make, and in this I close the circle and I go back to the re-infensing. Of course, we think about re-infensing in the contents of the model and in general, when we look at the assets of banks and transfer of liquidity, for example, from banks to another bank subsidiary in the same banking group. However, I think maybe we should start thinking more in the future what we think about re-infensing of assets when we consider possibility of transfer to other non-bank intermediary in the same group. Because there could be instances in which, it may be optimal from a financial stability point of view the transfer of certain assets of liquidity to support other non-bank intermediaries, but these may, of course, be not something that is not allowed by supervisors in general. Okay, so very interesting paper, a lot of food for thought. And with that, I conclude. Thank you very much. Thank you, the floor is open again. David, please use the previous practice to introduce yourself. Yes, Davy Żachowski from the European Central Bank. Thank you very much, very interesting paper and the discussion. I have one comment and a question. So the comment would be on the interpretation of the pay of rates and the increase correlation. When you presented the paper, I was thinking that it might be during the systemic crisis that pay of rates tend to correlate more. And I think this is exactly the periods when re-infensing arrangements are more likely to appear. Right, and now I think about Vienna initiatives when you wanted to break those re-infensing arrangements. So that's, I think, a nice interpretation to this extended correlations during systemic crisis. Now the question is also Andrea showed that, Angela, sorry, showed that one of the reasons to set up the SSM was a possibility of further financial integration. But still in spite of the common supervision framework, in spite of the common resolution framework and common deposit insurance framework, we have barely seen over the last couple of years cross-border mergers or acquisitions. So how does it square? Can you shed some light on this through the lenses of your model? And if not, maybe you can try to ask those questions via your model, thank you. Yana, and then there was a third question in the back. So, and Natalia, so you, so in the paper, you have either a national supervisory authority that minimizes the national cost of deposit insurance or a super national one that minimizes the overall cost of deposit insurance. But in your area, we still do not have the complete banking union, right? With common deposit insurance. So what is the objective function of the super national authority in that setting? I jumped the question from the right there. So maybe we go zigzag to the left afterwards. Daniel Groh, European Central Bank, supervisory arms. So a very practical question. Do you feel linking to also to the previous paper on cyber risk, do you think incidents such as those that we have been seeing in the cyber risk paper. So sudden liquidity shock due to infrastructure, unavailability or any other such shocks that we see now. For example, in the Silicon Valley bank, the Twitter bank runs as they call them. So depositors withdrawing deposits much more quickly due to panic involving social media. Are those elements something that somehow makes the discussion about the removal of ring fencing a bit more relevant and a bit more calling for action? Thanks a lot. That's the last question in the back there. Hi, Jeremy Bassani from the assistant. You seem to pay a lot of attention. I mean, you seem to pay. You pay surely a lot of attention to this change in supranational supervision, but the issue here I think unfortunately in parallel regulation has not changed. So even all this aspect about ring fencing that again in Andrea's pitch has been highlighted several times are not due to the fact whether there is national supranational supervision, about the fact that we have a regulatory system which is pretty much a national based and that's why we don't basically have cross border integration. There's been some benefits from supranational supervision, but without an attendant change in regulation, we have already an embedded ring fencing in the system for which supervision can do nothing. So thank you very much for the many questions. And I will answer selectively. So I will answer what I think is more important than I can give an answer to. So in the model, the supranational authority has a lot of powers and has all the powers or many of the powers Andrea and react to complaints about the SSM or the Euro area banking union framework not having. The supranational authority in a way can induce redistribution of costs between national deposit insurance funds. This is something I think the single resolution board or cooperating with the SSM in solving a cross border bank will have problems with because of the reasons that you were saying in the current Euro area banking union is as opposed to that very stylist set up in the model is incomplete. So the supranational authority in the model is what the Euro area banking union may achieve at some point when the three pillars are completed. And this when it comes to the deposit insurance or the lack of completion of the banking union. And then when it comes to some people that were saying we don't see that cross border integration. So maybe the model would tell you because that potential value that cross border banks would create if the supranational set up the institutional set up were truly supranational with all the three pillars completed are not still there. So many of the banks anticipate that there will be frictions that in the end we are still in some somehow national set up and part of the of those potential gains from cross border banking cannot be materialized because in other things ring fencing. Then when it comes to some of the comments by Angela and I think something that is interesting and it is and you need a model and you need a model. You were saying the SSM was going to be was created to be tough and being tough with banks is something good. Our model says something different. Our model says the SSM will be more lenient with banks when it comes to providing voluntary support with across units and being lenient, more lenient than a national set up. Banks anticipate that allows them to preserve value and then they might be willing for this charter value effect. They might be willing when fencing with this more lenient supervisory authority in this particular aspect too, they might be willing to take less risk, not more risk. And then regarding the mandate of the way we are modeling our mandate of our authorities is minimizing expected deposit insurance costs. So in a way these authorities dislike financial instability, dislike bank risk because when there are bank problems, this turns out showing up as costs for the deposit insurance. So in a way I think we are capturing some in a stylized manner and very concrete and focused manner some of the part of the mandate of the SSM. Thank you very much. Obviously the big elephant in the room is the fiscal arrangement in the union. So maybe I wonder whether there's another game going on that deserves its own model where that is actually fully taken into account where there are spreads that can go up and down and so on and change the fiscal position or their starting positions are different. But I'm looking forward to the next step of this research where that pattern which we all learned over the years is so important is also addressed.