 So, hello and good morning everybody. Welcome back in the room for a rather long but very stimulating session. Thanks a lot to begin with to the organizer to have given me the privilege of sharing that session. It's clearly a very relevant one. It has to do with climate and I think what we have there is as a follow-up in a way to climate risk type of assessment. Our supervisors, by the way I'm from financial stability, separation of bridge and not the supervisors, I still vaguely know what's going on in this house. So there's a lot of discussion and talks about how to produce the right incentive plus the right, I would say, framework for risk management with respect to climate in the supervisory world. We would have two papers and a speech raising, as I see it, a number of issues just to give the introduction from my own perspective but of course speakers will elaborate much more than I would do. We would have two papers with two discussants and afterwards a speech. I think the issues as I see it is really requirement from supervisory perspective from green capital. Is that really always good? Can it be debated? Another issue is how stress test can contribute to that? Can stress test fare better? We are doing on the macro side some exercise on the macro potential side in financial stability and there's been, as you know, an SSM supervisory exercise on the macro side. At the end of the day the aim should be to provide some incentive to act on balance sheet composition, on pricing, but at the end of the day is it really going to help with the environment or will there be other channels? And finally, last but not least, I think we can talk also about possibly greenwashing by banks, regardless of the regulatory framework. They could be some attempts there to put things on the side or disguise them so I think that's an interesting issue as well. So I've spoken enough and I think I will give the floor to Martin to begin with from the LSE. For each of the presentation we would have something around 20, 20 minutes plus and then there will be a discussion afterwards about 10 minutes plus the room and the chat understood from Andreas will be open. So Martin, go ahead. Okay. Thank you very much for the invitation to speak about this paper joint work with Microsoft on green capital requirements. For climate change obviously has become a major topic for financial regulators. ECB Bank of England have both conducted climate stress tests. The Fed has announced their climate pilot climate scenario analysis exercise. But in terms of how ultimately the information that we're gathering that way is going to affect financial regulation remains somewhat controversial both I think in the policymaking sphere but also more broadly among economists and the objective of this paper is to look at one particular part of that debate which is capital requirements and what I want to talk about in the next 20 minutes is how or to what extent can capital requirements be used as a tool to address two distinct potential objectives. One is climate related financial risks and the other one is more broadly emissions or carbon externalities caused by firms. In 20 minutes, I always like to get the high level takeaways across first just in case in the end will run out. The first takeaway will be that climate related financial risks or just a risk part can actually dealt with quite well bought by a capital requirements. And the reason in a way is intuitive because if you think conceptually doing that is actually not that different from dealing with other risks. Although there is a key challenge that maybe is present less so in when dealing with more traditional risks, which is estimation of those risks. And addressing financial risks although and that I think is one of the things that comes out of the model I'll show you is is not the same as lowering emissions and in fact one of the results I'll show you in our conceptual framework is that actually sometimes when you increase capital requirements for say dirty or polluting loans carbon intensive loans because you're worried say about transition risks. What may happen is that you will crowd out clean lending and in fact sometimes that will happen even as part of an optimal policy. I hope that I'll get there put in place by a prudential regulator in our framework and externalities on the other hand. So if we're thinking about reducing emissions themselves and using capital requirements as a tool to do that. There the framework I'll show you that actually is clearly inferior to say using something like carbon taxes you know the first tool an economist will usually think about and the model will clarify why that is. And you know just to preview a little bit one of the issues is that capital requirements are ineffective in reducing emissions either if bank capital is very ample in the system or if firms have access to markets that they can substitute into for example public debt markets. And even if kind of second part of that result even if you can affect carbon emissions then sometimes if you want to do that as a capital regulator it will come at the expense of sacrificing your other goals the prudential stability goals. But capital requirements and that's the last point of perhaps not completely ineffective they can have an indirect and that'll be hopefully the last thing I'll get to in my talk an indirect effect in helping with transition to a lower carbon economy. And that role is essentially a loss absorption role where if we worry that putting in place tighter financial regulation sorry environmental regulation leads to stranded acid risk in the banking sector then capital requirements can alleviate that and make governments or environmental regulator more willing to to make these required policy changes. Okay, let me quickly talk about the model. So this is a theory paper. It's a typical banking model single period everybody's risk neutral. There's a continuum of firms in this model or at least in the baseline model they only have access to bank funding for simplicity. There is a section in the paper also about substituting say to bond markets. There are two types of firms. The firm types are given so they're either clean or dirty carbon intensive or not carbon intensive say they can invest at date zero and then they receive a risky cash flow if they invest at date one. And the baseline model features a setting which which is meant to capture kind of the trade off in potentially some of the transition that we're looking ahead to which is that this carbon intensive technology from a financial perspective is more profitable higher expected cash flow than the clean technology. Maybe the clean technology is more expensive. You have to put in carbon filters, carbon capture and so on. But clean technology is more expensive but it reduces emissions. So there's baked in here this trade off between profits if you want and and reducing emissions. Those firms can be funded by a continuum of banks. They're competitive. They start for simplicity with a fixed value of equity. One can relax that they could raise equity at some cost and they raise insured deposits and the friction here standard friction in these types of banking models is that deposit insurance is not perfectly priced. It's going to involve a transfer to the banking sector when the deposit insurance fund pays out. And that means a potential regulator might want to put in place capital requirements to lean against this cost of the deposit insurance fund. Right. And so here's the regulator. And in this model than ultimately if if she wants to set to capital requirements conditional on whether a particular loan goes to a clean or a dirty firm. Right. And when you raise for example this capital requirement then you're going to lower this deposit insurance transfer deposit insurance put you know as it's called. And you're also going to affect who will be funded in in equilibrium. So I want to show you first what in the simple model the equilibrium in the banking market looks like. And then I'll try to use that to go to policy analysis and I'll split that into two parts. The first part is if you want ad hoc thinking about in this equilibrium what happens if we put in something like a green supporting or brown penalizing factor higher capital requirements for polluting loans lower for clean. And then I'll build on that and say well OK but what if you add an objective function and you think about what would optimal capital requirements look like. And there I'm going to look at two mandates. One is kind of the classic prudential mandate that just cares about the financial risks that arise here. And then our contrast that to coming back to the opening question of reducing emissions. What would what we call an impact mandate look like where in addition to the financial goals you also care about emissions themselves and banking sector equilibrium I'll be given the 20 minute time brief here and demand for bank equity each funded loan given capital requirements uses up some equity has to equal supply in equilibrium. Very simple because I'm just fixing aggregate bank equity at an exogenous number capital E. Now the demand you can think of as the demand curve similar to in a you know consumers problem where the consumer has some willingness to pay. Here the equivalent thing is what is the maximum return on equity that a particular loan of clean or dirty type can offer for a unit of bank equity. And that has two components in the numerator here. And the loan generates some NPV that can be shared among the bank and the firm. So that's the financial return. But then it also generates a deposit insurance put that can be shared between the firm and the bank. And so the ROE is the sum of those two components divided by how much equity this loan takes up on the bank balance sheet. If you plot that in the simple model with two firms right we just have the dirty firm the clean firm you're going to get an equilibrium that looks like that. The red line is the ROE on the dirty firm and that is higher than the clean firm because we assumed I assumed at the beginning that the financial profitability which is what this bank cares about is higher for for the dirty firm. And so you rank all of those in terms of the maximum ROE higher than the clean firms and then the equilibrium will just be where this demand curve so the step function is the demand curve intersects the supply curve. And so in this example, you would be able to fund up to here and then return on equity for the bank is is our star right here and these firms in particular on the right hand side of this blue line those clean firms would not be funded. So you can see you can play around with this I'll play around in a second with the capital requirements which will move around these red and green lines. You could also think about what happens when bank equity capital goes up or down by moving this blue line to the left or to the right. Just to show you quickly this is a very kind of reduced model just two types of firms you could say well what would happen if say we have some dirty firms that on average are more productive but there's some distribution of the productivity and the same for the green firms and then you can get something more continuous but a lot of the results that I'll show you will carry over into a setting like this one. So in terms of the positive analysis that we can do with this equilibrium I want to show you what happens if you start say from equal capital requirements for the clean and the dirty. A loan if you start playing around with these capital requirements and I'm going to be interested in studying in isolation some of the things that get proposed quite frequently which is well perhaps we should raise capital requirements for polluting loans or we should lower them the green supporting factor for less carbon intensive loans. And so the first first objective here is just to figure out in that equilibrium what will those interventions do right for now. I'm not maximizing anything from a regular perspective I just want to show you what the effects are of these interventions. I'll show you both the brown penalizing green supporting factor I'll spend more time on the brown penalizing factor because once you've seen that I think you will understand I hope. How this model works. And so I'm plotting here in these in dashed lines the original equal capital requirement equilibrium and under equal capital requirements because the dirty loans are more profitable their entire and so. That is first higher ranked segment right here now two things are going to happen if you say let us increase the capital requirement for those loans. One thing is the return on equity is now going to be lower because you have to put up more capital for those loans and also because there's more capital. There's a less of a deposit insurance put so this line is going to move down. Second thing that's going to happen is every one of those loans is now going to use up more equity on the bank balance sheet so if you want to fund all of them. You have to use up more of your equity which is on the X axis here so this line is going to get longer. And so you can see this is a small brown penalizing factor the first effect you're going to get in the simple two type model is where the line moves down and the line extends and that's the key thing because now. The green line which doesn't move up or down we haven't changed the capital requirement here is pushed to the right right and so effect here is something we didn't necessarily anticipate starting this project is that. Well you may actually be crowding out when you increase the capital requirement for the dirty loan clean loans at the margin. Right so one message here is you always have to think about you know what's funded at the margin and in this case in this equilibrium it's the clean loan. Of course if you keep going in this two type model at some point the are we on the dirty loan if you make a put in a sufficiently large capital requirements is going to drop. So the ranking will switch below the less carbon intensive loan. And so then at that point because you have a ranking switch you actually get to a situation where the carbon less carbon intensive loans are funded first and only then. The more carbon intensive loans and so here then you have a relatively big effect on how much of this stuff is funded because it basically jumps from here over all the way to the to the left side. Green supporting factor looks similar I'll show you both graphs right away except of course now we're moving the Greenland it's going to move up and it's going to get shorter because you need less equity per carbon less carbon intensive per clean loan. And so it's going to mirror in effect the results I just showed you except that now you are initially crowding in at the margin right so this moves up the ranking doesn't change but the line gets shorter which means some of these loans that weren't funded here can now. Find space on the bank balance sheet up until the point where at some point the ranking reverses and then the less carbon intensive loans are all funded. And then if there is still space on the bank balance sheet the now lower ranked carbon intensive loans can be can also be funded in this example you can see actually to get this shift or switch in ranking. You actually have to leave quite a lot of balance sheet space here and so actually quite a lot of the carbon intensive loans are still funded. Right so one way to summarize this is it's a little bit like substitution and income effects when the line switch their ranking it's like you've induced a substitution in fact it's relatively cheaper to fund clean or dirty loans depending on what you do. But there is also this thing that's like an income effect which is this crowding in or out at the margin and one of the key insights here is that this income effect has opposite signs for green supporting and brown penalizing factor so they're not equivalent. In terms of how they affect what gets funded. And you can do a similar thing in this general more general graph that I showed you earlier on you will also get a relative change in ranking of the loans and the lengthening here's the brown penalizing factor and lengthening of the overall demand curve. But to leave time to show you some of the policy analysis. Let me not spend more time on this I just wanted to show you that kind of the general effect survive also this slightly more if you want realistic setting. Okay, so in my last five six minutes let me talk about what we can do with this model in terms of thinking about policy. And I want to start from what I think is what I would say is if you want the natural starting point which is to say if we keep the regulatory objective function constant if you want so the regular bank regulator thinks about prudential gold if you want. Then this objective function is going to look something like this. The regulator likes the fact that the bank banking sector facilitates positive NPV investment so that's something that enters with a positive sign. But they don't like the fact that every once in a while the deposit insurance fund has to pay out that is captured by this deadweight cost maybe a shadow cost of public funds or something like that. So you while you want more credit you want to make sure the deposit insurance put is not too large. You can then and I won't really spend time on that here you can then you know put all the model in rewrite that objective function. The only thing that that's useful for looking at optimal policy on the next slide is to see that if we now think about how could climate related financial risks. In fact how the regulator looks at these loans so this is basically the prudential regulators ranking what's you know the surplus generated per bank equity unit of bank equity by a particular loan. And they will look at OK you know the NPV might be affected by climate related financial risks the certain firms become less profitable and this put value if suddenly say dirty firms become much more risky. And the insurance put will move right so it's these these two components that will be affected by potential climate risks. I want to show you one example of what we can do which is here I'm going to look at a scenario where you've calculated optimal capital requirements and then you say oh there is now this additional risk in this case in this example for dirty firms say carbon risks for carbon intensive firms how would you change your capital requirements. And so I'm plotting here the capital requirements for the dirty and the clean firms as a function of aggregate capital in the banking sector. You know there's some region when there's very little equity where only one type only the more profitable type which is a dirty type will get funded and then at some point both get funded and so on. But the key thing to take away here is that the original capital requirement is the dashed line and now you realize oh these firms have become riskier. And perhaps intuitively what's going to happen is you're going to be increasing as a regulator the capital requirement then for these dirty loans. The interesting in my view thing is that if these transition risks are moderate so you increase this risk a little bit then what's going to happen is that mainly you're interested in a bigger loss absorption cushion from a regulatory perspective but the regulator is not going to try to change the ranking of those two loans. And then in particular in this middle region you're increasing the capital requirement for the dirty loan a little bit like a brown penalizing factor. You will actually under this optimal policy crowd out some of the clean loans and under this objective function that's an optimal thing in fact to do. Then if I increase the transition risk even more than at some point you get to the point where the regulator actually wants to not just have a loss absorption cushion but also change the funding decisions. In other words induce banks to first fund the clean loans and then only then after that's been done if there's some space left from the dirty loan. So here if for these large risks you want to set a large brown penalizing factor not just for loss absorption but to induce the way banks rank their loans a change in the bay in the way banks rank their loans. So capital requirements can deal with the risk side of climate change in this way and depending on how large those risks are you might do different things but you can incorporate that in that say traditional objective function. So in the last two three minutes let me try to talk a little bit about why if you wanted to lower emissions in addition to your potential loan you might run into difficulties. So there's a kind of less optimistic message on that part of the paper. So I'm adding here same objective as before the potential stuff. I'm not pushing it out. It's still a project part of the objective function is the prudential regulator regulators objective. But now you are taking into account also say the carbon externalities or emissions directly. I'm going to focus to illustrate why this is difficulty on what I think is the most interesting case here which is suppose the oil firms or whoever the dirty firms in this in this firm I suppose you think actually there's no social value these carbon externalities are so large we actually don't want those firms to be funded. But of course the banks if their profit maximizing not thinking about these externalities then these loans are profitable in fact even if you were to set the capital requirement up to 100%. And so then you get into what we call here the limits of green capital requirements which is in an extreme scenario where the banking sector is actually very well capitalized then you can raise the capital requirements trying to push out carbon intensive loans but actually you won't be able to do so because if there's sufficient capital the banks are happy to still fund them at very high capital requirements. Now if 0.2 here bank capital is more limited then you cook you can prevent the funding of dirty loans but still you're you're going to be constrained. And the reason is that there's now an incentive constraint you have to make sure that banks will rank the clean loans above the dirty loans and in some cases you're not going to be able to do that without also lowering the capital requirements for the clean loan in other words you put the capital requirement for the dirty loan at 100% that by itself will not be enough so now you have to lower relative to your potential optimum what you're charging on the clean loans and so there is going to be the sacrifice in terms of financial stability that you're going to have to live with if you want to use capital requirements to reduce emissions something that would not happen with a carbon tax. I'll just refer you to the paper if you're interested in what happens if there's non bank financing you know what happens if banks could raise equity at some cost what if there could be some abatement incentive so you know the paper has a bit of a discussion on that but I want to conclude given that my time is up but just spending one more minute on something that might be on all of your minds which is okay but now suppose we think about capital requirements but also carbon taxes our favorite tool as an economist and how would they differ and what can we take away from that. First conclusion is that carbon taxes because they don't work indirectly through the capital requirement but directly lower the profitability of dirty investments in this model I've assumed them away but they would work right independent of the source of financing for a firm independent of you know substitution and bank capital scarcity and so on so they would work. And so they're going to be better than capital requirements here with which we just saw in a sense ineffective as a direct tool to reduce emissions. But the one role and that's my last point that the model suggests capital requirement might have is an indirect role as something that we call a facilitator say for a transition. And what we have in mind is that in a situation where we think that carbon taxes might be too low because governments are worried that they're going to impose losses on their banking sector and they cannot commit to raise those carbon taxes and stricter capital requirements might set the preconditions for a government led transition in terms of stricter environmental regulations right so by raising capital requirements. The financial sector will have a cushion against those losses safe on stranded assets so this first bullet becomes less important for for governments and so then it can become credible or incentive compatible for the government actually to follow through and increase carbon taxes. One of the things we like to point out here is that of course this is not a blank check for using capital requirements I think it it it outlines specific conditions under which this kind of facilitator role might play a role. And I'll conclude here and look forward to join us comments. Thank you. Thanks a lot for the perfect timing as we say here and we can now go straight to the discussion. Yes. Thank you. Thanks a lot for inviting me to discuss this paper. It's always a great pleasure to be back in Frankfurt. Okay so my discussion. Okay so I think the motivation for the paper is is is very clear so you know an alternative to carbon taxation could be to make the financing of polluting firms more expensive. There is a lot of work on ESG finance and in general but you know financing of firms is still mostly bank based. So actually I would assume you know that most of the research on you know green finance should be actually based on banks and so for this you know the paper by Martin is very very welcome. Okay. Okay I will first show you a Mickey Mouse version of the model so very simplified version of the model. Here I have to apologize to Martin for a very specific reason which is that I like the paper a lot. Actually I like it so much that I thought it was a good idea to give my students a very simplified version of the model as an exam. Turns out it was not simplified enough and now there are about 100 very angry investment bankers in London who are very angry at Martin and at me. Okay so but let me show you a very simplified version of the model so it does not really do justice to the to the full model but it's useful to get the intuition. So we have two types of loans clean and dirty and the dirty loans you know financially they are better. Okay they repay a higher interest rate with priority P the economy grows then all the loans are repaid no problem. Priority one minus P only a proportion one minus Q of the loans are repaid. Okay so there are two states in the economy there is a good state and a bad state. I have one bank okay so the bank is going to lend either to the clean firms that's LC or to the dirty firms that's LD the bank is financed with deposits and equity. Both both are perfectly priced. The lending of L cost gamma of L the deposits are perfectly insured by the government and at no cost sorry so the deposits are well priced but they are priced at zero basically because they are insured by the government. Equity is well priced so also to show that you know the results in Martin's paper don't come from the fact that banks cannot issue equity it works as well if if they can. Okay and the regulator is going to set some risk weights. Okay so the key thing is what the shareholders of the bank are going to get. So with priority P everything goes well they get you know the loans are reimbursed and then they reimbursed the depositors with priority one minus P. The economy is in a recession and then you know if the bank did not take too much risk then everything is fine it just makes some losses. But if this quantity is negative then the bank does not have enough money to reimburse the depositors then the bank defaults on its debt and the shareholders get zero and the depositors are reimbursed by the deposit insurance fund. Okay and so that's a very traditional model of bank reshifting in which because the deposit insurance is for free basically or you know marginally it's not priced well enough not not resensitive enough. The bank has an incentive to potentially take too much risk because of these deposits and that's why we have capital requirements. Okay very very simple model of bank capital regulation. Okay in this model what is the impact of having a brown penalizing factor. So assume that the capital requirements are such that you know if you have a risk weight a high risk weight on dirty loans high risk weight on clean loans the bank can never default. Okay in that case if the bank does not default actually the Modigliani Miller CRM applies there is no friction there is nothing. And so in that case the bank is just going to invest in the loans that are the most profitable which are always the dirty loans in this very simple setup. Okay so what this clarifies is that this is what Martin called you know the limits of capital requirements that a brown penalizing factor you know it's not the same thing as a carbon tax. Okay it's not a tax on dirty loans is the removal of the deposit insurance subsidy. Once this subsidy is zero you cannot go further okay you cannot make this subsidy negative. Okay so that's a very strong intuition from the paper. Okay so if the banks are safe or if deposit insurance is well priced basically this tool is useless. Okay then we can look at the green supporting factor so assume that we have a high risk weight on dirty loans but we have a low risk weight on clean loans. Okay so we allow the bank to lend more to the clean firms. In that case you can do some not super complicated mathematics but it's not very interesting it's just what you see is that there is this new term when you lend to the clean firms which is the deposit insurance put. Okay if you can take a lot of leverage when you lend to the clean firms then in the bad state of the economy your bank is going to default and the losses are going to be borne by the deposit insurance fund. Okay and so that's basically an implicit subsidy coming from the deposit insurance fund and then in a sense you know it works if this subsidy is large enough the bank wants to invest in clean loans. Okay but why is that? Well you know it's not really that we are regulating and improving the incentives of the bank to lend to green firms and so on. What we are doing is that actually we are subsidizing these clean firms in a very indirect way. Okay we are basically using the money in the deposit insurance funds or the single resolution fund or whatever in order to subsidize indirectly these clean firms. Okay so it's really a public subsidy, a very very disguised one. Okay so it works but you know it's basically a subsidy. In this toy model I cannot have one super interesting effect there is in the paper which is this equilibrium effect in particular the fact that very counter-intuitively what can happen is that if you have the brown penalizing factor actually what's possible is that the banks you know they have to lend less to both types of firms and they might end up lending less even to clean firms. Okay which I think is an unintended consequence of this type of regulation very interesting to think about. Okay so that was the intuition of the paper. Okay so my assessment very very very quickly the policy messages are important and powerful. I appreciate as a theorist I appreciate that you know the trade-off between having a more theoretically pure model and having a more applied and relevant model is solved very elegantly so I think we are the bliss point between the two. So overall I don't see much to improve upon. Okay so I think it's just you know ready to be published. I have one question so for some political reason in the model we assume that there is no carbon tax to correct the externality. Okay otherwise we would just have a carbon tax and that would be it. Okay so we are in the world of the theory of the second best. Okay there are some constraints on the tools we can use we don't really understand where they come from and we have to deal with them. Perhaps you know there is a symmetric constraint that some bank subsidies need to be maintained. I think this is more or less what Martin and his co-author have in mind at the end of the paper that there are you know political constraints that we cannot punish banks too much. If so I think that the green capital requirements you know the intuition that you can get from the model is that there are sort of a Faustian bargain. Okay so it's public support to banks via you know not well-priced deposit insurance versus in exchange directing credit towards socially desirable objectives. I'm not a big fan of this type of faustian bargain but by the standards of banking you know I think we have examples of things that are much worse in which you know governments really extract subsidies from banks and so on. Okay so maybe that's one way to think about this. If we wanted to go further in this paper if we want to develop the model a bit further I see two possible directions. I'm not sure this should be done okay so this maybe for another paper or for follow-up papers. One possibility would be to have the model even more general equilibrium. Okay so in the model actually there is no reason to have banks okay we just assume that they are the only intermediary between firms and households. One could think of a follow-up in which you know there would be bank finance and market finance so there is an extension going in this direction already. There could be a rationale for bank finance that may be different for clean and dirty firms so that would be much more complicated. The additional insight one could get from this model I think is to study new equilibrium effects like can dirty firms substitute with market finance are their cause of pushing clean firms towards banks this kind of things. Okay but again there is an extension that already goes in this direction quite a bit. Okay the second possible direction but I lost my pointer so there's not much I can do. Okay good thank you. Okay so the second direction would be political economy. So the debate on green capital requirements reminds me of Calomiris and Haber who say you know the banking system is the outcome of political deal making. So very cynical view of green capital requirements could be I think that you know the middle class is not ready to pay for green transition via taxes. Okay so we tried this in France and it failed miserably. The elite allows banks to finance green transition with deposits if things go wrong the middle class would be forced to bail out or recapitalize the banks exposed. Okay so again I think the green capital requirements the only scenario in which they have an impact is when there are these guys subsidy but someone has to pay this subsidy. Okay and so I would strongly advise against you know this type of policy because to me this seems a very reckless move. Okay so what we are saying is that we would allow banks to be under capitalized when they lend to clean firms and then in a few years time you know some banks may default. What will happen if the next SVB is a firm that was allowed to lend to very risky clean firms and operate with very low capital requirements and it has to be recapitalized by the general public. I mean that would be a political disaster and also a disaster in terms of you know having people on board for the environmental transition. Okay so I think the political economy aspect of this is very important. I would prefer applying the Thinbergen rule. Okay so we have a carbon tax to address environmental externalities. We have an income tax to redistribute the impact more evenly. Maybe that's the point that people have missed so far and we have capital requirements to ensure that banks are solvent. We have three objectives, three tools. The Thinbergen is happy and you know that's I think better policy making. Okay but having a framework to integrate all of this I think is something necessary and very useful for thinking about these issues and so we need more economic theory on this. Okay let me conclude because I'm out of time. So I think the policy takeaways are first order very important and very robust. The main two results I showed you I mean you can write any sensible model of banking capital regulation you will reach those conclusions. Okay so very very strong conclusions. The theoretical framework is very nice very useful you can extend it in many directions and so you can use this to think about a lot of issues on the interaction between banking regulation and the environment. I think it's a great paper touches upon even more fundamental questions in banking regulation in terms of political economy, general equilibrium and so on. So perhaps interesting for future research. Okay so I look forward to seeing this paper published hopefully in the best possible journals. All right thank you very much. Thanks Jean-Rédois for further elaboration of the issue of whether and how capital requirements, green ones can be good or not so good as we are right now. Before ending over to Martin I think it's best we discussed that he sort of takes all questions discussed and plus additional questions. Any question from the room? Maybe one or two? So we have the two already so we have two slots for from the room. Thank you very much. My name is Vanessa Riedek from the Financial Stability Department of the Austrian Central Bank. I just want to reinforce what Jean-Rédois Collier has just said at the end to give it a broader political perspective on supervision because the green supporting factor is not the first example for a rather political act to supervision. This reminds me of the SME supporting factor, the supporting factor for small and medium sized enterprises, but also the favorable risk rates for sovereign debt and then looking back on basal regulation on a more broader perspective. Would you say that this political intervention into regulation given your new insights of green finance is a problem and should basal or capital regulation stick to the prime objective of risk mitigation and how much other political objectives should enter regulation? Thank you. Thank you, thank you. I'm Skander Vanden Heuvel, Federal Reserve Board, very nice presentation, really enjoyed it. I think of these potential changes as being fairly long run so I realize you had limited time but could you say a little bit more about what happens when bank capital can be adjusted through equity issuance or retained earnings? There's a last one from the floor understood from Andreas, we will have another one from the chat and that would be over to you, Martin. And Patrick has from the ECB SSM, just a question out of curiosity. For me it was a bit surprising that these factors, I can't remember which of the two can crowd out clean loans. Do you know, I mean your paper has not yet been published, we just got support for it being published and this is something new that you found or whether this is already common knowledge because there is for example in the UK quite some hesitation because of all the different effects of green capital requirements, thank you. So Andreas? Yes, I have two questions. I have two from one person. So I start with the first one independently of being of lower or higher emissions, where is the threshold? Under what criteria? How can a criteria be objective if everything is relative or independently of the consideration of being a dirty firm? These firms can be financially very healthy and not having any financial concerns for being granted a loan. Thus, why putting a penalty on this lending while pushing lending to other clean companies that could potentially be of high risk from a financial point of view? And the second question from him is, are not the credit risk departments of the banks themselves who should do a financial risk analysis of the debtor with the inputs of the business risk and the financial health of the firm? Why a regulator should introduce a tool, policy skewed, that will create inefficiencies in the financial system? Just an example, is gas and coal clean in the light of the lack of other energy sources because of a disruption in their provision? Okay, great. I will, in the interest of the next paper, try to be brief, but lots of very interesting stuff. So thanks first of all to John Edouard and I will follow up on some of the smaller points, you know, bilaterally with you. But one of the things I really liked in the discussion was you had a great intuition that, you know, when it's about removing the deposit insurance friction, you can always do it with a 100% capital requirement. But the externality will not go away at that point. And I think that's a good way to summarize why you run into these difficulties when you try to use capital requirements in order to affect emissions. I like all the political economy suggestions and I think we'll think about some of that perhaps in follow up work, but I agree with you that that's very interesting to study. Let me start from the back in terms of the question and hopefully I noted them all correctly. The question that came from the WebEx chat. So, the second part of the question was, well, shouldn't banks credit risk departments take care of this if something is riskier because say of climate shouldn't that be priced in. And one thing, and I didn't have time to clarify that during our presentation is that in this model banks actually do price this stuff correctly. So to the extent that there is climate risk that affects the financial risk or profitability of a firm, they take that into account. But of course, the incentives of how to take it into account are distorted because of the deposit insurance put right so we cannot rely simply on the credit department. Of banks, just like we use capital requirements for other types of risks to try to align incentives in a world with deposit insurance. And I agree with some of the skepticism voiced by by the same person who I think of how how easy is it, you know, I'm in a easy position here. We have a simple theory model. They are clean and dirty firms. Of course, in practice, what is clean, what is dirty, what does that taxonomy look like is something that we have not answered in that paper. And I think it's clear to be upfront about that. One thing I think I would say is that given that even in this model where you can distinguish clearly between these two types, capital requirements are not a good tool to reduce emissions. In that sense, I think stepping into the world where you have a difficult time telling people apart might be even harder. Is the crowding out common knowledge? I don't know, but I hope that that's something that we learned that from doing the model. And so I hope that's perhaps also something that other, you know, researchers or policymakers can can take away from that model. I think it's one of those things that once you write down the model, at least then we were like, okay, of course, of course, you have to look at what's the last loan that was funded. That's going to be pushed in or out, but it's not something we had in mind when we started writing this paper. So from that perspective, I hope that it is something new, something we haven't done, which would be interesting is look empirically. Can we say something about, you know, for the average bank, what does their marginal loan look like, right? And so I think that would be something interesting to do either by us or other people or, you know, people in the research department here to figure that out. There was a question about adjustment in equity and this is actually, you know, can banks raise equity via retained earnings or via issuing equity. The first version of the model we wrote had that, and then we realized it doesn't really change very much as long as raising equity is not something that's completely frictionless. So if you can raise costlessly as much equity as you want, then you are in, you know, what I guess we economists would call the Modigliani-Miller setting where capital structure becomes irrelevant. You might as well have 100% equity on everything. But as soon as equity has some cost, then what's going to happen is that straight line that I had, the vertical line for the equity supply will be an upward sloping line. And then you can do the same kind of experiments. You're moving up and down those lines or extending them and you recover the same qualitative effects. But sometimes some of them will be attenuated or amplified via the endogenous reaction now in terms of how much equity the banking sector has. The last or first going in reverse order question was about a broader political perspective, the SME factor. So I agree with this analogy that if we are thinking about a green supporting factor, it does remind us to some extent of the SME factor that was, I think, in place for a while. So I think there is an analogy there. Now, how much should political goals affect regulation? I think that question goes a little bit beyond this particular paper, but I think what I would say based on, you know, what our model says is that, Yeah, even if your goals are quote unquote the right ones, right? So suppose it is actually what we need to do, reduce carbon emissions, which probably many of us will agree here, bringing that into the regulatory capital regulatory objective function. I think in this paper, I would say based on the results after shown is not a good idea, right? And I think there is a role for this indirect role that I discussed at the very end. But I think in that sense, my interpretation of what we got out of the model is, let's say, a somewhat positive one for what can we do in terms of climate risks if we can measure them, but more negative on can we do much beyond that in terms of reducing emissions? So that tells me perhaps it's better to be cautious and not introducing particularly if you're not going to be very effective some of these tools into the capital regulators objective function. I hope I answered everything otherwise. Happy to talk in the coffee break. Thank you very much.