 What role, if any, do safe assets play in the functioning of strictly from a micro point of view on the financial system? Is there a scarcity of safe assets? I think that is not the point to be discussed. I think that would be the starting point of the speakers this afternoon. This scarcity of safe assets, as I say, is a starting point. It's also a fact that has been taken as given. And it's also a fact that, if I remember well, first up here, at least in my memory, in Greenspan time as a reason and something behind Greenspan's conundrum. With this, as a starting point, we will have a first presentation by Professor Ricardo Caballero here to my right. Professor Caballero is the fourth international professor of economics at MIT. He's also director of the MIT's World Economic Laboratory. And he holds a PhD from MIT. His research fields are basically, I understand, macroeconomics, international economics and finance. Today we'll be making a presentation based on a set of papers that he has authored or co-authored with some of his colleagues at MIT. And the most recent one, I believe, is one which I went through yesterday called Safe Assets, Scarcity and Aggregate Demand. I guess his presentation will be based on that paper. And I quote at the paper's conclusion is that, and I quote, policies that increase the net supply of safe assets somewhere are output enhancing everywhere. We will also have two presentations by Professor Marco Pagano here on my left and by Professor Marcus Brunen-Meier, also to the far left. Professor Pagano is well-known, he's a professor of finance at the University of Naples Federico Segundo, is also president of the Inaldi Institutes for Economics and Finance, and also chair of the Advisory Scientific Committee of the SRB. He also holds a PhD in economics from MIT. We have a crop of MIT representative here this afternoon. And he has done research, extensive research, basically I quote from corporate finance, banking, and financial regulation. Also Professor Brunen-Meier will be presenting, like Mr. Pagano paper on which I will comment, is Edward Sandler for Professor of Economics at Princeton University, is director of the Benheim Center for Finance, which is a prestigious center of finance. He holds, he's the exception, a PhD from the London School of Economics, and he has research interest in international financial markets and the macro-academy, with emphasis I believe on bubbles, liquidity, and financial stability issues. The two will make a presentation based on a paper that they have co-authored titled SBIS, safety in the trenches. If Mr. Professor Caballero concludes that creating and solving the shortage of safe assets is a good thing, the two speakers on my left, they will treat the logical implication can we increase the supply of safe assets? That's a financial engineering problem, more so if what the starting point is unsafe assets. From a bunch of unsafe assets, we will be creating a safe asset. That I think is the content of the paper they will be presenting, they will be dividing the presentation on, I don't know how exactly. And finally, we will have, and this is to my fire out, Jeremy Zettelmeier, which will be discussing the three presentations, Dr. Zettelmeier is a senior fellow at the Patterson Institute for International Economics, he's also a chief economist of the German Economics Ministry, is that right? It was. Oh, you were. Okay. That's why I was checking with you. I understand you are director of research at the European Bank for... That's also an old job. Oh, and then, well, they don't, I don't... Just Peterson. Peterson needs you. We have a PhD from MIT. Ah, yes, yes. Okay, thank you. I was going to mention that, you joined the club, you are entitled to join the club. Get out of there. Okay. I just deserved an honorary degree. We are 10 minutes behind the schedule. Well, I don't know whether the schedule included is a short presentation. So by instructions to you, you will have each of the speakers who will have, including the discussion, will have, let's say, 15 to 20 minutes, an average of 18 minutes. Hopefully we will make it that 310, 310p, no, sorry, yeah, 230, yeah, one hour, yeah. So, 14.43, and we will have 30 minutes for each speaker to reply to the questions that Jeremy Zettermayer will pose and for the audience to answer some questions. So I'll give the floor to Professor Ricardo Caballero. Thank you very much, Julio Durán. Where do you want to stand up, from here, what do you prefer? What is the strategy? Anyways. It's for you to decide. I have no question. I have the option. Yeah, yeah, you have. I can control this one there. Safer. So anyways, I'm part of the supporting cashier. Really the beef of this session is their proposal and I'm just supposed to give some introductory remarks on why safe assets are important in the functioning of the macroeconomy. Certainly they have a fundamental role for the management of individual institutions, but my angle is more of a macroeconomic angle. So there's no doubt that in general, the store of value is important for many things that matter for macro, for aggregate demand, for example, and also from the supply side. They matter for production, investment, and consumption activities. It is store of values that allows us to decouple current income from the expenditures that are needed for different variety of activities. Now safe stores of value play a very special role in this context, and that's particularly the case in times of great uncertainty, either perceived or actual uncertainty, but the safety aspect of that store of value plays a very important role. And as we have been discussing, and I pretend to say a bit more about, there is a large number of consequences that can emerge in an economy once there is a shortage of these problems. And the last point that I want to make that has to be made already repeatedly in this conference is that this is a problem that not about the future, it's not about the past, it's about the present. The present and the past and the future, but it's very important today. There I have just a quote from Narayana Kocher-Lagota in his previous job. People do move in different jobs, in which he essentially argued that this was perhaps the most important problem that central bankers were facing at this time. Now you ask a very good question, which is, what is a safe asset? Here I have some responses by institutional portfolio managers, and it varies, and it's not surprising that it varies. I can't even read what is in there, but the point is that there are many answers. But it's not surprising because what is safe depends on the individual portfolio, the particular point in time, what is the main concern for that investor, what is the regulation for that investor, and so on and so forth. The work and definition that I have adopted together with my co-authors, by the way one of them will be presenting one of the formal papers related to that one in the next week's conference. Here I will just talk about the general principles. But the work and definition we have used there is that it's an asset that preserve much of its value in a large systemic event, and what is a systemic event that can also vary over time, some points in inflation, but nowadays it's really a big financial crisis, a big recession, and that has been for us, again, the work and definition of what is a safe asset. And could a monthly safe asset, well, sovereigns, especially in emerging markets, insurance companies, pension funds, mutual funds, the financial system, could produce a safe asset with very few people, really, safe assets in the sense I just defined. Businesses, mostly in developed markets, not all of them, very few corporations can do that. And the financial system, but it does require a lot of financial engineering, really, to produce truly safe assets from the financial system. How did we get into this problem of short touch of safe assets? Well, there are many, many reasons, many structural reasons that have already been mentioned. Demographic factors have played indeed a big role. There are regulatory reasons that have increased demand for safe assets, the usage of safe assets as well. There are issues of storage of safe assets and velocity of circulation of safe assets. Both have been affected recently. But I think another factor that was not mentioned here and that has been very important over the last 15 to 20 years is the behavior of emerging markets. We used to have, especially after the Asian crisis, we saw a very unstable periphery of the world. The core looked very robust. So the solution for the periphery, one of the solutions for the periphery was to essentially to stock up on assets, safe assets from the stable core and build massive amount of reserves. Yesterday, somebody said in response to my radical acceptance points, the radical acceptance points of small open economies is buffers, well, the Asian economies took it very, very literally and they built big, big buffers as a result of that crisis. And you can see actually the effect of that crisis on rates. And this is just to corroborate the point of who has been demanding and who has been supplying the assets. Obviously, big role of China and the US is on the other side, on the supply side, all producing the Middle Eastern economies and China have played a big role, not only China, but primarily China. Let me skip all that. Now we all know that that put enormous pressure on the financial system. So that was the demand side, which was a structural factor, sort of building. And then sort of the core that was producing the safe assets couldn't take it anymore and in finding excuses to build new safe assets and eventually the system blew up. To tell you the truth, this works for the periphery in the sense that the periphery looked a lot more stable than it would have looked in the past in the presence of a similar shock. The core itself looked quite unstable. Now, the world didn't have the option of really swapping the periphery for the core and then we had the situation of safe assets shortage resolved because the size of the financial markets in the peripheries is sort of an order or two orders of magnitude smaller than that of the developed economies. But the point of this slide is to show that also the supply side of safe assets markets sort of contracted very dramatically. Now there are different estimates out there which obviously vary depending on what you think is a safe asset or not. But this is one illustrative table which was put out by Berkley's a while ago and has been sort of reproduced at the IMF and other places. But the point of that table is that essentially the supply of safe assets was cut in half as a result of the supreme crisis. A big part of it came from essentially the privately or semi-privately produced safe assets in the U.S. and the other half came from essentially European sovereign debt that was no longer considered part of the safe asset supply. So a big structural trend that hasn't gone away and a big contraction in the supply and some may argue was a perceived supply was never really safe and that connects a little bit with the points that were made this morning but it's a very important point. But that's the way the world felt after the financial crisis in terms of these markets. And the implications are of the kind we have already discussed. No, we had this structural decline in interest rates which was exacerbated. It was not start by the crisis but it was exacerbated very rapidly by the crisis, not only the supreme crisis but those that follow. There are other implications as well in an environment like this. This is something actually I talked about in an annual conference here at the ECB in 2006. It's an environment like this, you're going to tend to see bubbles popping up. But when there is a savings in general, you tend to see bubbles. When there is a shortage of safe assets, you tend to see bubbles and things that can be perceived as safe for a while. My Chinese is not very good but I think that means sort of parking spots in Hong Kong or something like that. The price of parking spots which is some real estate type thing was very bubbly in this time. But there certainly you can see in emerging markets in general real estate markets are typical sort of safe store of value that emerges in situations like that. The other implication naturally was that the private sector had a strong incentive to produce these safe assets. The corporations have a hard time producing them, especially when the natural corporations have run out of the debt they want to issue. And so then we needed a lot of creativity and the creativity we all understand now took place in the financial system, particularly in the US, but it also involved European banks, certainly buying and holding them. That industry had a hard time. That I think was a very central part of the of the supreme crisis and obviously was a very central part of the contraction in the supply of safe assets. And the implications then, let me insist on the interest rate, but also on the point that was made this morning, is that certainly safe rates, which is the lower blue, the darker blue line there area, decline structurally for quite some time, but it essentially hit zero very shortly into the supreme crisis. At the same time, however, the equity risk premium, this is a forward-looking complex equity risk premium, has not declined and actually has increased over this period of time. So that essentially puts us into what we call a safety trap. And a safety trap is like a liquidity trap in the sense that it's the safe rates behave very much as their interest rate behaves in a standard liquidity trap, but it's a safety trap in the sense that that is produced by a scarcity in a very particular kind of asset, which is much harder to produce than general wealth. And a reflection of that, that it is the safe asset that matters and not the other kind is again the fact that this premium has remained very hard. And there are many things that, so to think sort of, you want to think very broadly about what a safety trap is, is a liquidity trap that is not responsive to any form of wealth increase. You know, in a standard liquidity trap model, the role of whoever is doing policies to somehow inflate wealth, one way or the other, that's what you need, that's what it's going to boost aggregate demand and that's what it's going to take you out of it. That doesn't really work when the reason for that is a shortage of the safe asset, because it's the binding constraint is a scarcity of a particular form of asset and it's only an increase in that kind of asset that will relax the constraint that is producing a liquidity trap type phenomenon. So you can have situations in which an equity market is booming but it's not pulling investment out, for example, you know, which I would say is one of the main relationships that has shifted in the post-crisis period. And when you're stuck in a safety trap, you tend to get very much a kind of thing that happens when you're stuck in a liquidity trap, which is that it's hard to get out of economic malaise, it's hard to recover in an environment like that. Just to, I'm not going to represent any malls or anything, but just to give you an illustration of a mall that we all have in mind if it has taken at least a principle of economics course, is this famous Keynesian Cross, knowing which you have an aggregate demand, determined output, you have a 45 degree line, and then there is an aggregate demand, which is a little flatter. Well, all that changes, you can write very fancy malls, but all that really changes if you put one of the safety components in it, is that one key shifter of the aggregate demand is the availability of, it's a supply of safe assets. So you can write literally your standard Keynesian Cross mall, but now you will see that in your aggregate demand, there would be a supply of safe assets that pops up in there. Therefore, if you have contraction in the supply of safe assets, supply of safe assets or net demand of safe, net supply of safe assets, net of demand. Okay, but at certain interest rates. But what happens therefore if there's a contraction in that supply of safe assets, is that it's equivalent to in your standard Keynesian Cross mall, sort of a dropping autonomous consumption, autonomous investment or something of that kind. So the shift downward in aggregate demand. And so, and therefore, one solution to this problem is to increase the supply of safe assets, and there I have my GoSB Snow, as one of the ways of doing that, which I think is very welcome. Another implication of an environment like that is that the effects gets more or less indeterminate, actually, in that environment, for reasons that are fairly technical, but in a world in which interest rates are persistently stuck around zero, output becomes endogenous and the effects becomes undetermined. And what happens there is that the total size of the liquidity trap in the world is more or less fixed. It's not fixed, but it's more or less fixed. And the location of that liquidity trap to different regions depends on the exchange rate. Okay, so the incentives to start using the exchange rate, essentially to relocate aggregate demand, are even stronger than in the natural standard Mandel-Fleming setup. Other implications that were also mentioned in the morning is that obviously the incentive for the private sector to produce safe assets, you know, you get greater rewards of doing that. One manifestation we have seen this is emerging market world, so a big increase in corporate additions. Actually, Latin America didn't have that stuff, and now you can find it even in ETFs. Okay, so big incentives to produce now that to transform emerging markets, it's the highly rated emerging market corporations, to transform them into really truly perceived safe assets, probably needs an extra layer of CDOs, of SPVs, something like that, but the inputs at least are being created in the same sense as the subprime loans were being created that eventually were packaged and so on. Other reactions, well, central banks have reacted to a liquidity type type phenomenon by accumulating, by doing lots of things that you normally do with monetary policy beyond that massive accumulation of assets in the balance sheet of central banks, part of which I like and part of which I dislike, and that's what I want to talk about. I think that the initial reaction of the Fed was very much what you want to do in a world like this. In a world in which you have a shortage of safe assets, you want someone to take that risk that people don't want and swap them for something safe, which is what the world wants at that moment. And the initial QE policies that the Fed had were very much of that kind. Now, gradually over time that was shifting and the composition of safe to risky assets was changing to eventually end up with operation twist, which I think was highly counterproductive in my view. It was highly counterproductive in my view because, you know, at least you do the standard open market operations. You are essentially swapping a very safe asset for another safe asset. You're swapping short-term debt for reserves. When you start going into operation twist type things, you take a safe asset away, which is what you inject as safe assets, which is reserves or short-term treasuries, but you take away a very scarce safe asset, which is long-term safe assets, which are very important for insurance companies, for mutual funds, for pension funds, and so on. And in fact, it's a negative beta asset. And a negative beta asset with respect to systemic events I had in mind comes more than one for one in terms of a safe asset, because it allows you something else that is risky to become safe as well once you combine it with this thing that has a negative beta. So I like very much what the Fed did early on. I like a lot less what happened in the later stages. I like a lot more what Europe does, precisely because international central banks have done discretion at least of what they target, and they're expanding the programs way beyond sort of the type of safe asset. I don't think it makes any sense. I understand their mandate for the ACB to be buying boons, but, you know, I understand their constraints and parameters and so on. And I understand the institutions are supposed to be a little bit more rigid than we are, because that's what preserves certain things, but there's a limit. And I guess that now you're finding sort of a short touch of assets in general, you may well be reaching the limits of all that. So what's next? Let me conclude here before the real show starts. I mean, one way of sort of capturing in an image, I think what is happening, I think what is likely to continue to happen is I tend to think of the hotelling model. If you remember, it's this model, which is a finite, exhaustible resource, and the price keeps rising, rising, rising. No, because people are using it and you cannot reproduce it. Well, I feel very much that's what has been happened to safe assets when you normalize them by the rate of growth of the demand for safe assets, which includes demographic factors, which includes the fact that China grows at a faster, which demands lots of these assets grows at a faster rate at those assets those countries that produce the safe assets, but it's very much a hotelling-like view. And I think the early phases of the solution of that problem was fairly benign, which is to say, well, you know, you don't have enough safe assets, one thing you can do is increase the value of these assets. And a very natural mechanism to increase the value of those assets is for interest rates to decline. As interest rates decline, the value of safe assets as a safe interest rate decline, the value of safe assets sort of went up, up, up. So that was a very benign resolution of the shortage of safe assets. Eventually, we hit the zero-lower bound and that benign mechanism went away, yet you didn't have the valuation channel anymore. And now the next valuation channel in line is effects. You can increase the effects, appreciate the currencies of those countries that produce these safe assets, Switzerland, the U.S., and so on. And that in itself solves a little bit of the shortage of the assets, but it starts having much more macroeconomic consequences because it affects certain countries more than others. We also can affect quantities, and certainly the issuance of more public debt from those countries that can issue it, not anyone, those countries that can issue it, is a way of increasing, you know, the supply of safe assets. Now we all know that there are political limits to those kind of things. And some of those political limits are justified, although I think there is a space in many places to expand it. I think they are justified because the argument that, and again was said a little earlier, the argument that, you know, as you want because you're not paying anything for this debt, is a typical dynamic and efficiency type argument and that requires coordination and coordination can go wrong. I mean, you know, in a world in which people vote Brexit or elect Donald Trump, I wouldn't trust so much in backward induction. We'll get as always to the equilibrium we want and so on. So it's good to be a little concerned about things that are somewhat fragile. Funding infrastructure makes a lot of sense to me with the capacity we have because it improves both sides of the equation in the sense that it produces the type of debt we need and it also has a potential to increase growth, which is something you also need. You need to score economies to grow a little faster in order to close that imbalance. Private substitutes, I think that we need them, but we know the fragility. We already went through a round like this. I think that one way or the other, we have to use the capacity of the private sector to generate this safe asset, but understand that they cannot really ensure the systemic tail event, the truly tail event. And for that, you need some sort of public-private cooperation. I think it makes little sense that that you issue just a government debt for every single instance, but I think there are certain events for which you need them. And it makes sense, I think, to involve the private sector taking over all that part that you don't need and just specialize on something. And of course, there are issues of moral hazard, how much you charge them, and so on and so forth, but those problems don't seem enormously large relative to the other enormously large problems we're facing. And the last point I want to make is that, well, if that's it, it's very scarce, don't waste it. And I think that so you can also reduce demand for safe assets. And I think when you start looking for reductions of the main asset for low-hanging fruits there, I think central banks look very, very apparent to all of us. In emerging markets, there is a good reason, there are effects reasons, but there is also precautionary reasons to buy these assets. But there is a lot of space for pooling. And I think that, you know, swap lines and all these kind of things that the AMF has been working on and it makes a lot of sense to me in the world, as a way of reducing the waste of it doesn't make any sense that every single country has a big pile of reserves, it would make a lot of sense to do a little bit of pooling there. We also use the facilities of the develop world, for instance, for a very extreme instance. And in the case of develop markets, I think that what I said before, I think they should focus on risky assets. I understand there are lots of mandates issues, institutional issues, but if there are issues to work on, are those issues because it makes very little sense that, you know, a very large share of the US Treasury is really being held by central banks. Is international central banks, the Fed itself is holding a big share of those instruments, and I think it's very wasteful. Thank you. Professor Cavallero. Very interesting. I guess that Jeremy Zittalmaio will have questions on the conceptual framework, which I believe is a MUNDEL type of model where you have an additional constraint. And very interesting also insights on monetary policy from the Fed and fiscal policy. And the policies of central banks with regard to investment of reserves and how many reserves do we have. I guess the audience will have many questions on that. Now we go to Professor Pagano who will be illustrating us on the financial engineering problem to construct a safe asset out of a, I guess, set of less safe assets. Right. Okay, thank you for having me here and as was already said that I'm drawing on a joint work not only with Marcos who will present the second part of this paper but with some Langfield, Stein, Neuberg, and Dimitri Vajanos. So essentially the issues I'm going to touch upon are three. First of all the rationale of this object, SBIS, that I will explain about in my talk. The second is how safe can we really make them depending in particular on their so-called trashing point or subordination level and how much safe assets are produced this way. And the third, since I will answer the second question by producing the results of some simulations that are in our paper which take, as given the probabilities of the fault of euro area sovereigns, I will ask a kind of look as critique type of question which is what if actually SBIS themselves were to affect to some extent these probabilities of the fault as well as the fact that on that front and I will try to address that by presenting the results of a very simple model. Now, the rationale for the introduction of such a safe assets in particular with reference to the euro area is twofold. First of all, the so-called diabolic loop that we have been experiencing during the sovereign debt crisis in the last few years in the euro area is the risk by which a shock to sovereign solvency affects the value, mark to market value of bank equity leads them to the leverage generating a credit crunch which lowers tax revenue and therefore has repercussion on sovereign solvency as well as generating potentially the intervention of the sovereign to counter the banks with negative equity value which also has a negative effect on sovereign solvency which feeds back upon again banks. So, now this why the introduction of a safe asset can mitigate or even eliminate this diabolic loop, well to the extent that banks rather than holding risky sovereign debt were to hold with asset not sensitive to sovereign risk this diabolic loop would be mitigated at the very least. There is also another implication of the diabolic loop which we saw at work in the euro area which is essentially a flight cross-border flight to safety. So, essentially this is like the international dimension of the diabolic loop by which when a sovereign is prioritised in a country the investor seek safety in other safer countries. So, this is similar to the slide that Ricardo was putting up, it's just more geographically determined so to speak. Essentially the flight to safety as we all know was essentially during the crisis a flight to Germany. It was the country that was and still is safe sovereign in the euro area and so that's why it was a flight to Germany because safe assets are asymmetrically supplied besides being relatively scarce as Ricardo said also they are asymmetrically supplied within the euro area. And here again, ESBIS can help because first of all they would reduce the likelihood of the diabolic loop and therefore also this flight to safety might arise but also to the extent that still there would be sovereign crisis, the crisis would take the form no longer of a cross-border capital flight but actually a portfolio shift from high risk to low risk European assets. So, what would we desire from a safe asset? What are the desirable features of a safe asset? First, in the context of the euro area we would like to be more similar. First of all we would want such an asset to be at least more symmetrically provided than currently it is in the euro area but we also would like it to still be safe at least as safe as the German Bund and hopefully also more liquid than the German Bund and more abundantly provided. Also importantly for political reasons of political acceptability we would want this asset not to feature joint liability. So, essentially this rules out all existing proposals of euro bonds which all involve joint liability among euro area sovereigns we would like also this not to be too politically costly in case it kind of reveals to be a flop in the sense that if investors don't like it we want actually to be able to go back much too much cost. So we would like for the same reason it not to involve have to involve treaty changes, European Union treaty changes and also to be useful for monetary policy as well as for asset managers currently there is not a liquid and risk free benchmark yield curve in the euro area and this the introduction of a safe asset in the euro area such as the one that I will discuss in a second would provide that feature as well. So what are we proposing essentially? It's not a new proposal it was already advanced by a somewhat larger group of us the Euronomics Group in 2011 so we are now proposing it again so to speak putting it again on the table the idea of SBIS is quite simple it is a kind of plain vanilla form of securitization of sovereign debt, euro area sovereign debt which exploits both the diversification that comes from pooling on the asset side and the protection that comes from trashing so essentially seniority on the liability side essentially SBIS are the senior tranche or senior bond if you wish which would be produced by such a securitization which would enjoy protection in case of default of any of the issuing sovereigns of given by the junior bond essentially junior bond would stand first in line to take the losses coming from any sovereign default now of course the bigger is the buffer the safer would SBIS be but also the less of them would be naturally available so the question is as you move the trashing point essentially how safe can you make them and how much of them can you produce and these are essentially issues that we first of all try to look at by way of simulations as I said before and we define a safe asset as one which has half of a percentage point 0.5% like expected loss rate over a five year horizon so it's equivalent roughly to the requirement of a tripled bond now we simulate these problems over 10 million draws and we have two reference scenarios a benchmark scenario and an adverse scenario actually we have several adverse scenarios where we increase both the default correlations between sovereigns the probabilities of default losses given default and so on we have a thick appendix in the paper where we experiment with all kind of more adverse scenarios that are being called the results are broadly still consistent with those of the benchmark scenario so in the benchmark scenario the idea is very simple we have three macro states in which the economy can be one is a very bad state is a crisis state which arises with 5% probability in which the probabilities of default of all the sovereigns are quite high in Italy has in this state a probability of default of 65% and the loss given default of 80% then there is a bad but not so bad state what we call milder session which arises with 25% probability in which the probabilities of default are still elevated but lower than in the very bad state and also the loss given default is 80% of the value as in the crisis state and the good state which arises with 70% actually the probabilities of default are quite a bit lower and losses given default are 50% the value lower than in the crisis state and essentially we the way we choose these probabilities of default are such as to make sure that our expected loss rates over a five year horizon are broadly in line with those which are implied by CDS premia on this currently by these sovereigns and what do we do with this simulation essentially we use them to compare as this is the thing that I showed you before with alternative possibilities one of which is essentially the status quo so essentially the current debt which is issued by these sovereigns another one is what we call pure pooling which is essentially just pure diversification forming sovereign bonds without actually trashing them as well another one is a country level trashing that is neither pooling at the aggregate level not trashing at the aggregate level but trashing at the individual country level to get an idea of how far we can go with each of these now this is the status quo essentially this is the five year expected loss rate on existing sovereign bonds according to all countries in the euro area we have that the criterion for safety that we have chosen is met these are Germany the Netherlands, Luxembourg, Austria and Finland where the expected loss rate over five year horizon is less than 0.5% all the others don't are quote unquote unsafe according to this criterion but also that horizontal line that you can see in this picture which is the level of the expected loss rate given by pure pooling tells you that also simple diversification does not yield give you an expected loss rate below 0.5% so it's also not safe so just diversification is not enough according to our simulations if instead we start using trashing what do we get extra well first of all look at these lines here these segments what we do here is we choose we look at different subordination levels at these trashing levels from 0 to 50% and those segments that you see in this picture give you the expected loss rate that you get from as this so essentially when you not only have pooling but also trashing of course with 0% you get exactly 2.79 which is the same line that we saw in the previous pictures but as you can see the loss rate of SBIS drop dramatically as increases the subordination level and in fact are already below the benchmark of 0.5% with 25% subordination level and they're already below the German expect the loss rate which is 0.13% if you because you go to 0.09% if you choose a 30% subordination level if you go further of course you get practically to 0 in contrast the bars that you see in this diagram here these bars they represent the expected loss rates that you can achieve with the senior tranche at national level so if you do a trashing at the national level for Germany, France, Spain and Italy first with 0% then 10%, subordination level as you can see the decline is much smaller essentially you increase safety but far less than you can do if you both have pooling and trashing. How much extra safe assets and this goes in the direction of what Ricardo was saying before how much do we increase the supply of safe assets by using SBIS the supply of SBIS is plotted in this diagram for different levels the subordination level as the black bars now you can see that there are no black bars corresponding to 0 and 10 this means essentially you don't get until you get to 20% subordination level you don't get any increase in the supply of safe assets with SBIS all of a sudden you have the senior tranche of the subordination so SBIS becomes safe if you achieve 20% subordination level at that point you get the maximum bank for the back in terms of the supply, increasing the supply of safe assets you get almost 5 trillion euro of SBIS essentially on the market it's more than twice the supply of existing safe assets it is triply sovereign in the euro area if you go to 30% of the subordination level which is our benchmark our preferred subordination level of course you get the decline you get less of a supply but you get that this stuff is safer than German boons what are the grey bars the grey bars indicate how much what is the supply of safe debt that you would get from tranching at the national level as you can see it doesn't initially because of course you apply a more stringent criteria in terms of tranching level it only increases a little bit towards the end because you get France essentially also into the game so to speak of producing safe assets by increasing the subordination level sufficiently but anyway you don't get as much of an increase in the supply of safe assets as you would with ASBIS this slide illustrates what is the 5 year expected loss rate with the riskiness of junior tranches you can see that the expected loss rate also decreases and these are the red segments in this picture decreases dramatically as you increase the subordination level and if you choose 30% subordination level it stands at 9.1 9.1% which compares close to that of Portugal which is almost 9% or a basket of Italy-Portugal Cyprus and Greece GDP which is 9.3% and of course it decreases even further if you choose a higher subordination level instead if you again look at the numbers which look at the junior tranches of national security stations you effectively have no effect from raising the subordination level in terms of safety now let me come to the last point of my talk which is whether ASBIS can actually not only redistribute risk because essentially so far we have the amount of risk that is currently present in the various sovereigns in Europe is redistributed across the two tranches if instead of doing that we ask the question what would happen to these probabilities of default if actually due to changing behavior of the system through general equilibrium effect the introduction of ASBIS where to not just reallocate but actually reduce the risk so if this could happen there would be a second reason why ASBIS would contribute to the supply of safe assets not only in the sense of actually creating this tranche which is safer but actually making even potentially the junior tranche safer and the reason why we think this may be the case and we try to make this point through a very simple model is that as I said before it can reduce this diabolic loop operating essentially the idea is that if you couple the introduction of ASBIS with an inducement to banks to actually hold ASBIS instead of domestic sovereign debt you can reduce this diabolic loop parameter region so to see this consider how the diabolic loop may arise in a multi-country setting and we have a very simple model to capture this point so we consider a setting a very simple setting with two symmetric countries where investors are subject to independent sunspots that is they can suddenly panic about the loose confidence about the solvency of their sovereign with probability in both countries now if the sunspot occurs investor pessimism causes the price of domestic bonds to drop by reducing the market to market value of bank's equity in so far as they call domestic sovereign debt which makes in turn equity negative and pushes the government or at least may push the government to bail out the banks which in turn has a negative effect on the government's solvency and this can validate essentially initial pessimism of investors in this sense the diabolic loop can be an equilibrium instead of holding all of their sovereign portfolio in the form of domestic sovereigns banks where to hold alpha s where s lower bar is the face value of domestic sovereign debt in the form of domestic bonds and beta s of a pool security formed by 50-50 mix of domestic and foreign sovereign bonds so essentially this is a kind of if it's like our pool portfolio in the simulations so that the total sovereign portfolio of banks will be gamma s which is the sum of alpha s and beta s. Now beta can then be seen as an indicator of the degree of diversification of banks in terms of sovereign risk so if beta is equal to 0 you have complete home bias if beta is equal to gamma so alpha is equal to 0 then you have complete diversifications that are holding 50-50 the domestic and foreign sovereign bonds. Now what we show in the model is that if you raise beta this indicator of diversification and you do not do trashing so essentially here we are considering a pure pooling as a way for banks to diversify their portfolios if you raise beta this has two opposite effects a virtuous effect which comes from diversification but also a bad effect that comes from contagion because essentially the diversification effect comes from the fact that domestic banks are less exposed to domestic sovereign risk but the contagion effect comes from the fact that they are by the same token more exposed to foreign sovereign risk. So we can capture both of these effects in this picture you have on the vertical axis beta which is this measure of diversification which ranges potentially from 0 to gamma. Gamma is perfect diversification 0 is complete home bias but you have another quantity on the horizontal axis and that quantity is the degree of capitalization of domestic banks so how much equity they have. You can see that if E0 the initial equity of banks is sufficiently high anyway irrespective of the degree of diversification of banks you are outside of the diabolic loop region. You don't have this diabolic loop equilibrium does not exist if however this E0 is below that gamma pi tau of the L0 point so it's to the left of this vertical blue segment you have two possibilities. If banks are still highly capitalized so they are in the interval between this point and this point as you raise beta and you go up in that direction you move from a region where a diabolic loop may arise only if there is a local sunspot to a diversification region where banks where the economy is no longer exposed to a local sunspot there is no longer a diabolic loop that can be triggered by a local sunspot that is by a confidence crisis at the local level at the country level because banks are sufficiently well diversified and so not very sensitive to a local drop in the price of domestic sovereign debt if actually you are below this point instead you have this kind of bad side so to speak this curse coming from diversification which is contagion because at this point if you increase diversification since banks are not very well capitalized you also import instability from the foreign sovereign essentially you will have a more likely diabolic loop coming from contagion now this is where actually ESBIS come in because ESBIS are able since they exploit not only diversification but also diversification so they have offer better protection against the sovereign stress then simply pooling does the introduction of ESBIS as you can see makes the region with no diabolic loop larger and compresses the contagion region as well as the diversification region and does this the more so if you increase the trashing point that is subordination from the upper panel of the figure to the lower panel you see that the contagion region and the diversification region shrink even further and in the limit for sufficiently high trashing point you have that the contagion region totally disappears as well as diversification region also disappears and the notabolic region actually becomes highest and what is the intuition for this intuition is that trashing effectively default risk to junior bondholders which are outside by assumption in the model outside the banking system so that's the key thing and therefore it kind of lowers the likelihood that this diabolic loop will rise and notice there is a little miracle here that in the no diabolic loop region also the European junior bonds so the junior tranche of the securitization becomes safe because there is no longer a diabolic loop okay so to conclude essentially the points that I tried to make is that SBIS are in a way superior as a way to produce safe assets in the euro area and probably this is a kind of more general principle than pooling because they also exploit not only diversification they also exploit trashing that is a benefit of seniority and if we take as given probabilities of default and losses given default they would more than double or about double the supply of safe assets in the euro area and still be at least as safe as the German bonds at the same time for that level of subordination 30% the European junior bonds would be about as risky as currently Portuguese sovereign bonds are if banks were encouraged to hold these European senior bonds instead of the domestic sovereign debt that they currently massively hold due to the extreme home bias that currently is a feature of European euro area banks in particular if they were encouraged to replace these domestic sovereign bonds with SBIS their introduction according to the model would be able to break the bank sovereign diabolic loop which relative to the results of our simulations would make SBIS even safer and AGBs less risky than our simulations suggest. So in a way if there is some truth in our model in a way our simulation results are quote-unquote too pessimistic relative to the potential achievements of SBIS. Thank you professor Pagano. Thanks a lot for having me here to be back here at the ECP like I guess everybody is now convinced that we need SBIS and my task is to show you how to do it and a lot of details to be figured out and I want to go walk you through many of these details so that's the implementation part of the papers or the last part of the paper is a big chunk of implementation details and the devil is in the details and let's go with the devil to the details so what I will do is step back a little bit I will give you a definition of that and I will contrast three definitions then I will talk about sovereign debt and banks and there are some conflicting views within Europe I might jump over this because the time is more limited than I thought initially and then I will say if you introduce some regulation on sovereign debt how should SBIS be regulated so there should be a certain principle that SBIS should the only thing what is needed is that they are not disadvantaged and we are also proposing an SBIS that will be fixed on described or specified and finally I will say how would you go in the transition phase from the current regime to the SBIS regime so what's a safe asset there is one which I say or safe is something which is risk free at a particular horizon so you have some guys or you know might be very a worse and they want to hold the basket some central banks or something and that's one definition of safe asset goes a little bit with Cavalier in his model he enlarged it a little bit so he's closer to the definition that will come later on the second definition of safe asset is Alahomström and Gordner would say it's like a safe is an asset which is informationally insensitive so in a sense that you don't have to worry about that somebody has more information than you and you face a lemons problem you know you get 100 euros back whatever state of the world realizes hence it's safe and I don't have to worry about asthmatic information I have more so the lemons problem won't appear that's an advantage of a safe asset and then I have worked with Valentina Datto so we push two attributes of safe assets and the two attributes are the following we say a safe asset is like a good friend it is around when you need it so when you need it at a random horizon it is has high liquidity has high value and you can actually use it it is there like a good friend and then what the second aspect we call the safe asset autology often as asset is safe because it's perceived to be safe and a classic example is like in August 2011 when the US Treasury was about to default because of a conflict within congress and the Treasury or the White House actually the price of the US Treasury appreciated so the fundamental got worse but the price appreciated so there is some multiple equilibrium story and there's even a bubble component so safe asset in a model is often a bubbly asset so it doesn't the fundamental might move in the opposite direction than the price of the safe asset and then running with this that's I think the definition of a safe asset of course it's still open in the year I think there's no the literature has not converged to our final definition now let me jump over the next thing was in sovereign debt and banks how we regulate sovereign debt there are holdings by banks there's some conflicting views and this also relates to different views which is you know much bigger in context discussed in my book which just came out you have to read it to get it let me move on to the regulation of aspes essentially the question is if you were to introduce some risk weights on sovereign debt how would you how much risk weights would you put on aspes and what we argue in this paper is that a fair way is this look through principle so the aspes again that's what Marco has shown you have some portfolio of sovereign bonds they are up to 60% of GDP and based on some moving average of GDP weights so you have some and they have some risk weights so you take the aggregate risk weight and then you concentrate the aggregate risk weight on the junior bond and the senior bond because it's safe doesn't get any risk weight so in a sense the whole construction if you buy the senior bond and the junior bond which is like holding the whole portfolio you have the same risk weight whether you buy a senior bond and a junior bond or you buy the whole portfolio essentially this is the so-called look through principle where essentially risk weights are pushed towards the junior bond that's it's a better arrangement having this aspes with this than having some let's say exposure limits because this disadvantages small countries if you have exposure limits what happens is that all the banks they will say oh we need a diversified portfolio but we pick just a big sovereign bonds and we will not buy smaller countries sovereign bonds because their market liquidity is lower so an aspes arrangement is a fair distribution it's a union wide sovereign safe asset rather than having not a union wide safe asset and again as Marco has mentioned that's important because if you like to save the capital flows and not cross border anymore they're actually from European level to a European level that will flow from a European junior bond to a European senior bond and both are European bonds so money is not flowing out of the periphery anymore to the core but it's actually for one European entity to another European entity so you might ask you who is buying this junior bond so this is you know all the risk is concentrated at the junior bond who will buy this junior bond that's one of the objections we often get oh somebody has to buy this and who is taking on this risk the first thing and that's what Marco pointed out very nicely it's more linear in Miller fail so of course we repackaging existing risk so essentially if like in a simulation we do just repackaging you see you create a lot of safe assets but because of this repackaging individuals people's behavior will change in particular banks behavior will change they will hold not the junior bond and this essentially makes the whole risk smaller so in dutchness risk is actually going away this diabolic slash doom loop is reduced and that's one reason why the overall risk is going down so more people can hold the junior bond also at the lower yield the second thing is why people might want to hold the junior bond like our friend from pension funds and insurance companies because it essentially gives them embedded leverage so the asset the junior bond has embedded leverage in it so essentially it is as if I would buy a portfolio of sovereign debt and I would level it up by issuing 70% of debt no what does it mean I as an individual investor cannot level up this debt I cannot raise that at the SPS interest rate but if I buy the junior bond I get the SPS interest rate when I level up that's a safe bond safe interest rate so this is a huge advantage so this embedded leverage is way more attractive than a straight out leverage investors could do too so it gives you an advantage compared to what we have at the moment so for both of these risk reasons there is there should be some demand for the junior bond now what should be in the junior's handbook so there are many aspects which should be specified in the SPS handbook you know certain you know standardization and all this so one of the standardizations is of course because SPS will be a much bigger market than the current fragmented national markets there will be an arbitrage margin so one thing is because of this high market liquidity where who gets this arbitrage margin we assume here in this specification that there will be a fund and accumulate and accrues we accrue the benefits from this and if they're everywhere silver and debt restructuring you can use this additional fund building up now there are also some market liquidity challenges we are facing so one is this low debt level problem so if you buy up a lot of this national debt and then issue a senior bond and a junior bond there might be certain states certain member states in the European Union or in the newer area whose debt to GDP ratio is not high enough so remaining you have to essentially suck up all of the debt so we don't have a price signal anymore from them so that's one challenge we face and the other challenge is there might be a small country problem I said already it's a big advantage compared to the other proposals that the small country problem is much less of a problem in this proposal but of course it might be that we absorb a huge chunk of some small countries debt and the remaining part becomes less market liquidity is going down so that these are two things which you know have to be debated as well so how what do we want to do in order to minimize that so we first of all in the SPS handbook we want to say that there's some standardization of SPS it shouldn't be that anybody can just create SPS and you know do slightly different versions of SPS and then we don't have any standardization and the market liquidity advantage would be lost so we would like to have if you do is if you issue an SPS whether it's a public or private bank or institution that the subordination and the dronging point is the same you only can call your thing an SPS if you have 30% subordination and dronging point the same thing is so the portfolio shares of the sovereign thing should be the same so we propose the GDP weighting and the GDP weighting should be based on past GDP a moving average in order to avoid prosycicality so if you just use the current GDP it would be a problem let's suppose a country has a goes through a crisis the GDP weight would go down and then there would be a smaller weight absorbing its stats so this would be prosycical column we have switched this off that's why we had this moving average specification the other thing is in the dronging we don't want to have any maturity mismatch or very limited maturity mismatch in this dronging vehicle whoever does it for public entity there's very little maturity mismatch in the sense that whenever you buy 10 years sovereign debt you also issue a senior bond and a 10 year junior bond and you do this for all different maturities in order to have a nice yield curve what else is specified in the handbook it's specified how the national debt offices have to coordinate how they issue debt so there should it can't be that one country is only issuing one year debt and the other is issuing 10 year debt then it's hard to build this no maturity mismatch vehicle so there has to be some coordination across the DMOs and in order to manage that if they insist on this you can still do some timed dronging that some security choice you can still do but it would be nicer if you don't have this maturity mismatch in the first place the other thing is they have to coordinate somehow when to issue new debt or they just have to say we issue something so it can't be that France is issuing March 15th and Germany is issuing in March 28th and there's a two week gap and we don't know how to build the portfolio there's some additional warehousing risk popping up the way to deal with that is to be announced where you do the secretization in advance said okay France and Germany will issue something end of March and if you want to buy senior bond there will also appear end of March in advance to be announced end of March everything will happen but we do it already you know March 20th and then March 31st everything will occur simultaneously there's no warehousing risk to reduce the warehousing risk and it's also important that no country says oh by the way now we issue the new bond which is senior to the SPS before we serve the bonds which is in the SPS structure we serve some other bonds so this would undermine of course the whole we also think it's very important that should be also specified in the SPS handbook that the SPS issuer in particular if it's a single one or competing private ones should have the right to buy in a secondary market so it's important that the DMOs you know cannot corner squeeze the SPS issuers in a sense but you know there is a possibility that if there's a shortage of particular bonds and the price goes wild they can go in a secondary market as well now I touched upon already who should issue these bonds this SPS and the junior bond should it be a public entity should it be a private entity or both we don't specify in the paper which way to go we just outline the advantages and disadvantages and that's a political debate we wanted to stay away from and I can just tell you what are the advantages if a public issue does it versus a private issue or perhaps ideally perhaps both can do it so the danger with a public issue like the Europe ESM or the ECB or the Euro system are generally because we don't have so many ECB could do it for the part of the bonds they're holding which are part of the loss sharing or for the other part there's no loss sharing that's more complicated or the European Investment Bank could do it we think the ESM is probably or ECB would be the more natural ones to do it so the danger is of course if it's a public issue there might be some legal challenges when there will be a debt restructuring who has the power to decide what and how to restructure the debt and you might not want this or there might be some market expect some implicit bailout guarantees and all that the other thing is there might be some legal challenges because it's not clear whether this would require a treaty change or what regulatory or legal changes you need to go through not a treaty change but perhaps some legal fees because the private sector might charge a higher fee the advantage of the private sector is more arms length relationship and this is more important in sovereign debt restructuring arrangements and the private sector might also retranch the junior bond not if one junior bond and then you tranche it again and you have fine tune it depending whether the amount lies but nevertheless I think the issue is whoever issues certifies it can't be just anybody coming saying we create SPs and with the private of course you always have the counterparty credit risk the previous session is focused a lot to counterparty counterparty risk so it's counterparty credit risk but you can manage that with bankruptcy remoteness so it's only the warehousing risk and the warehousing risk you can manage with it to be announced the privatization then there might be some counterparty legal risk if somebody is issuing SPs under German law and the other one is issuing SPs on some other countries law there might be not totally identical so there might be some we agree on a common law under what law these SPs have to be issued under what legal jurisdiction so it could be that all the SPs have to be issued under X countries law that has to be agreed upon but the good thing is in this case you don't have really compared to other securitization you don't have a moral hazard select what assets to put in because it's always in 19 countries sovereign debt you don't have to monitor the only issue is when it comes to a potential default who has the voting rights we do some debt restructuring or not that's why we had a whole subsection how do you structure the governance doing a restructuring who has the voting rights and that's something to debate and discuss with you so what we think is that the SPs issue the power should not be totally concentrated if there's a single SPs issuer in this one SPs issuer because that would be too much a concentration of power and it should be essentially a second look through principle so essentially you take the votes which come from the sovereign bonds and you distribute it to the European senior bond and the European junior bond according to their shares you have to balance the conflicts of interest between a junior and a senior bond holder and you have essentially the same look through principle we applied also for regulation more generally we would also apply to the power distribution in case of restructuring or if restructuring has to be decided but that's something really for the worst cases which are important to deal with how the governance structure would look like now finally I conclude is the question how should we finally introduce it the whole thing so we have specified the long steady state what are the details how to do it and as Marco mentioned what's the nice thing about this proposal about many other things it has no downside risk so let's suppose you propose it and it goes online and nobody picks up on it nothing happens you just go back what you had before you go back to the previous situation and nothing happened now what we want to propose here is a three stage introduction the first stage would be some limited experimentation so there's some entity public or private or both we'll see that issues buys in the secondary market some of the sovereign debt and once the sovereign debt is bought it issues the senior and the junior bond later it goes on perhaps already to the primary market but once we one sees it works well and all this in this stage you can always revert very easily in the stage two you introduce a big swap and essentially the idea is everybody who has sovereign debt at the moment they can participate on this big auction and this auction they can say you can swap your sovereign debt for European senior bonds the combination of European senior bonds and European junior bonds if you have some bonds which are more risky like let's say Greek bonds then if you swap them you get much more junior bonds and very few senior bonds if you want to swap the German bond to something you get actually more senior bonds and fewer junior bonds such that everybody reflects the current value of this thing so people submit demand schedules everybody submitting demand schedule you see depending on all the prices the bond price the Italian bond price and so forth how many of these things I want you put everything in the market clearing mechanism and then you see what equilibrium exchange rates are across this bond and then accordingly everybody can swap these arrangements and it's very similar what you have in spectrum right options where you bundle things and you bundle whole things various spectrum rights and here is different national sovereign bonds and then swapping it for juniors and senior bonds then the other thing we thought is that before you do this that's all in the old regime there are no risk weights on sovereign dot once you have done this then you can phase in some risk weights and as I said the look through principle the risk weights for there should be no risk weights on the senior bond it should be all concentrated on the junior bond the ECB can play a crucial role as well on top of it but being potentially one issuer it can conduct monitor policy primarily with the senior bond that's natural to have a benchmark yield curve and also at the short end it will conduct the monetary policy of course it has to set certain haircut rules and all this but of course given that it's such a safe bond and the simulation I've shown it will be as safe as a German bond that's actually very natural to have similar haircut rules so let me conclude with this slide on the details and the implementation I first started off what's a safe asset and what we like or what we termed this phenomenon a good friend analogy and the safe asset autology I think these are two marks of a safe asset I jumped over this bank sovereign risk holding about the risk weights where the conflicting views are in Europe and again it's one dimension where the different perspectives in Euro and there are some underlying reasons for that deep reasons and different economic philosophies and we talk in the book about these economic philosophies a lot and this leads to these different attitudes which I think we can overcome I talked about who will buy the European junior bond and what's the advantage of this embedded leverage and that's a big advantage in terms of regulation for ASPs by the look-through principle that's essentially a fair treatment there's no advantage nor disadvantage in the margin if we can give ASPs a little bit of an advantage it would be kick-starting this because we have to get this whole thing started at some point and the ASPs handbook will then specify the whole standardization, what's the dronging point, what are the portfolio weights and all this and how to harmonize the national debt issues in terms of frequency or how frequent we have to issue that more frequent smaller processes and then we also have to decide is the ASPs provided by a monopolistic public issuer or is it can be many private issuers or both in this and importantly is the governance structure in terms of if there's a potential debt restriction of one of the member states who is devoting right stand and as I mentioned once we have this, this is a long run steady state and we can get there to three stages I just outlined before I believe it has thanks Thank you I'll give the floor to Jeremy Zettelmeyer I believe he will not be short of questions to address and I will also ask the organizer maybe to give us more time because I'm a lousy chairman and we are close to exceed in our time but I think Jeremy has the right to do the same time as the other speakers Thank you so much Mr Chair, so just to clarify my role here so I'm going to be a discussant but like Ricardo said Ricardo to much greater extent myself I just garnish in this session right so the main dish is indeed the ASPs and we have been sort of strategically placed around it so you know Ricardo provided the appetizer then you got the main course and now I'm going to give you some sort of dessert or if you like digestive that makes it easier to see and you know I I don't need to be a digestive to the to the appetizer and so I'm not actually going to discuss Ricardo's presentation also I thought it was very easy to digest in its own right because he presented it very very nicely so I'm I'm basically going to play the role of sort of a nice stress test if you like so nice stress test is one that's not so easy as to completely lack credibility ex-aunt but in the end doesn't really get you into trouble right so this is essentially what I'm going to do you said that you said that not me so let me first give you a little bit of context like Marco said the idea the SB idea has actually been around since late 2011 it was presented by a somewhat different and bigger set of co-authors in late 2011 to the to their credit the set of co-authors has run because some of them have gone into public policy roles and I believe that they still believe in SBs even after having gone into public policy roles so in that sense that's probably the stress test you should take most seriously more than anything that I could say now now in spite of the fact that the idea received a lot of attention it wasn't really taking very seriously by policy makers and what we are now seeing is sort of a second much more thorough attempt to argue the idea than the short paper that the authors produced in 2011 and also possibly in a more receptive environment because things have changed in those five years so five years on from 2011 there's still dissatisfaction with the Eurozone but we are no longer in open crisis and this may create a greater willingness to experiment at the margin the lack of safe asset motivation which was not the motivation that was originally pushed by the paper it was there but it wasn't pushed so hard it has picked up stream as a second leg that the proposal can stand on and this is why we kind of flew in Ricardo to make that a strong case of course you know one of the problem in 2011 was that people really had the great financial crisis in their mind so anything that sort of smacked of financial engineering CDOs, SPVs they actually called their idea an SPV in 2011 they probably quickly realized that was a mistake and so this word SPV only appears right at the end of the paper in a particular context so all these ideas are no longer that credited you know there's an official attempt at the European level to revive the asset-backed security market and so you know they fit in there and then finally a particularly important point back then in 2011 they had a lot of competition from alternative proposals right there was the blue bond red bond there was a sort of European safe bills idea there was the exemption pact of the German consulate for economic advisers there was of course you know just plain Euro bonds and so this essentially all these rivals have run into walls and the reason is that they all involve redistribution of some kind and so this is the only proposal that at least by design own motivation is choose redistribution right if it works as intended it should be more hazard free and because distributional tensions in Europe have if anything become more entrenched within the Eurozone since 2011 it might help their idea right it might it might be a solution to a problem that doesn't impose a cost where people are very sensitive these days about that cost and so the question is will the idea now finally break through and I'm going to try and answer that question and you know always I'm not going to give you very clear answers essentially yeah they have a good chance it might but they are quite there yet they might want to make a few changes and that would make it even better and the way I'm going to approach this is I'm going to go through the original set of objections and then ask to what extent the objections still stand both given the passage of time and given the new version as I recall it right I've been sort of following this from the outside there were essentially four objections maybe the fourth one is more my interpretation okay the first main objection by you know many people was you know this may be a neat idea but it's just way too complicated to academic there's just this ton of practical issues that is rises and is left unaddressed okay the second objection is and this is sort of more the Germanic objection is probably one of the slides that you would have shown if you had had more time is that SBs are just a way of introducing mutualization through the back door so they're not as innocuous as they seem why because the great financial crisis has taught us that you know making claims about instruments being risk-free based on financial engineering are really fraught with risks and if this goes wrong you know these may infect default also for losses and then someone is going to be liable who is that going to be well the part of Europe that doesn't cause that default which is us okay the Germans now in my case I'm German even though I do not work for the German guy okay so there's a lot of talk about the devil in this paper you know there's the devil in the details there's the diabolic loop but you know people may think that in fact the paper the idea itself is pretty diabolical because it has this you know kind of hidden thing you know it's not openly mutualization but it may have that implication then the third one was actually a point that was made a lot and you mentioned that that no one will really want to hold the junior tranche in other words you create something safe but at the expense of creating a junk bond that cannot stand on its own feet and if it cannot stand on its own feet then of course the senior tranche cannot be sold sold either and then finally maybe my spin so the SB idea is very closely tied and you know you haven't really emphasized this very much to the idea that you want to definitely give risk weights or impose some sort of regulation on bank exposures to national sovereign bonds so it involves it's a double bill it involves you know a tougher regulatory regime for sovereign bonds than exists now and that of course triggers a whole set of issues basically there are many people who out of interest or because they fear that this may cause a crisis that this is going to lead to big shifts and asset holdings are skeptical of that idea okay so where do we stand with respect to these four objections just to give you a quick overview I think the paper gets very very far I would say it's 95% 90% in my non-practition of you in overcoming this first objective that this is a neat idea but they haven't really thought it through it really really thinks it through this version of the paper I think and I think what you also have conclusively defeated is this argument that there would be no interest in the junior tranche I think that you have done a very good job of so within this set I think there's only one objection left which is fairly obvious and the authors are actually aware of it but it's not in the paper which is related to how you're going to price the underlying national bonds once this thing takes off I'm going to come back to this on the issue of mutualization through the back door I think the paper makes a lot of progress but it doesn't fully succeed in overcoming that potential objection and similar with respect to the last objection which is whether this proposal in a sense makes too many enemies in a politically economy sense to be successful so I'm going to elaborate on these three points so the first one very quickly so the worry is once you get to the like Markus explained you can start introducing this at the margin but at some point you have to make the jump to this big debt swap and at that point depending exactly on how you set the upper bounds of what you're going to include in the SPV or in the portfolio you could render a whole bunch of national debt markets illiquid with one stroke and this is obviously true if you decide to include all that but it's also true if you set some upper threshold of GDP so they mentioned 60% of the GDP as a possible idea whether a whole bunch of countries in Europe that have debts less than 60% of GDP and of course you know the EU fiscal framework actually tells all of us to get back to 60% of GDP so in the state everyone is illiquid in your proposal it doesn't have to be that way if you define the caps differently so if you say 60% of national debts are included now we briefly discussed it and you said well the Germans wouldn't like that but that has to do with the fear that there's something diabolical after all if it's truly this is not going to be a source of moral hazard or transfers then there's really no reason why what would have to set the cap at 60% of GDP rather than 60% of the debt stock now having said that if you do end up rendering some of these secondary markets illiquid there is in fact a solution and so my suggestion as a discussant would be that you take up this issue and discuss it in the second part of the paper in the next version and the solution is to basically rely on primary markets to do the price discovery and so if you have a sufficiently competitive set of licensed private SP producers as purchasers you can in fact do that I mean there's having deep liquid secondary markets in government debt is not something that most countries in the world have and yet most countries in the world do issue sovereign debt and the reason is because we know how to structure primary markets Governor Ingvers was head of the MCM department at the fund and a lot of what that department did was to go around to teach essentially central banks on how to do that so you can actually overcome that objective but it has a strong implication which is that it kills the idea of using a central debt agency as the only or the main SP issue I mean that it seems is inconsistent with any form of SPs that would generate this liquidity problem so I think it does affect the discussion at the end okay then quickly on the issue of financialization through the back door so I think the paper has made progress here because the simulations of expected loss rates are much more thorough than the in the early version you do have an adverse scenario with some fairly massive contagion assumptions and also the European debt agency is no longer at the center of the proposal which is you know this potentially more hazard inducing institutions now I still think that you know to the Germanic mind this is not going to be quite enough and the reason is that you are following basically with the SPs you do two things first you give us a recipe of how to construct a safe instrument via financial engineering but secondly you are also labeling this as a safe instrument you are calling it European safe bonds now like you point out in your work with Haddad you can justify this labeling in terms of a multiple equilibria story but you know any sort of labeling can also create an implicit guarantee that the instrument is really safe by construction that's not a problem but we know of course that safety via financial engineering relies on certain modeling assumptions particularly you assumed of default cross-sectional correlations loss given defaults and if you haven't gotten that quite right then we are back to that that world so the question then is how credible is the simulation exercise so I looked at it I think it's quite credible you know they look at reasonable assumptions strong assumptions I mean in the sense of being tough but there is one objection that could really hurt you which is that when you do the calibration of course you are using either creative ratings but more important CDS spreads that are based on I think end 2015 prices and so there are many people with the Germanic mindsets arguing that these things are distorted and that they actually understate the true sovereign risk and the relationship between two sovereign risks in Europe and so then if you have that mindset you still are going to be worried whether your SPS could really withstand a multiple default and so you know just out of interest I sort of check this in a very primitive way can I just bring this up no I do have one maybe this one oh yeah really do you have the little okay so this is a version of a table that's in the paper so what you see here is Euro area countries except for a group of smaller Eastern European countries plus Malta that I've grouped into one just to make the table more manageable and they are ordered in if you like in terms of credit rating with the poorest rated countries the riskiest countries ordered on top and then what you see in the in the second or third column is just the outstanding debt stock percent of GDP and billions of euros now I've made an assumption on whether they would be included in the portfolio or not and so for simplicity I've here assumed Greece would be out because it doesn't have tradable debt at this point it's all official almost all official everyone else would be in full I have another table which shows what happens if it's only 60% of GDP which I'll show in a second so if you do that then the portfolio weights in the SPV would be given by this column percent of included so you would get roughly Italy France and Germany in the same proportion right Italy is the biggest national component in the basket but only by a little bit you can make some loss given default assumptions and so I have made some assumptions which are essentially taken from the paper but they're not the most drastic loss given default because I've actually worked on sovereign debt crisis and I think these are extremely high loss assumptions they made in the most toughest case in the paper so I've taken some sort of average of scenarios but they're still very high loss given default assumptions so in the case of in Italy in default we assume that they assume as their central value that's 69% would be actually lost so then the question becomes based on these assumptions could the save bond withstand so to speak a correlated default and so in one scenario I've assumed that all countries rated 9 or above default at the same time so this is essentially all the countries that had a problem after the great financial crisis Slovenia plus Italy and you know my intuition was well you know this instrument can be maybe withstand one or two big default cases but not this complete set well surprisingly it comes out of the beauty of multiplying these things with losses given defaults it could right so Estonia, Paris, Portugal, Italy, Slovenia, Spain and Ireland all default and investors recover somewhat between 45 and 25% and the rest is gone then the share of defaulted debt in the total debt included would be less than 25% right so their 30% tranching would do it so how far can you take this then so this is the next one I've added now a whole group of small Eastern European countries and the pecking order doesn't do much because they have such a small weight so then you add the next one the next one happens to be Belgium Belgium is a little bit bigger even with Belgium defaulting we are still below 30% so in order to get above the 30% we need France to also default then we are at above 40% now if you're worried about that and maybe the Germanic mind might be worried about that then you have two options you can either make the junior tranche thicker you could raise it to 40 or 42% if you like or you can do something that they suggest in passing but they don't actually explain it very well and this is another suggestion that you should explain this more which is rather than including everyone in full you take everyone up to 60% of GDP so that's a beautiful Germanic idea because you are penalizing the high-dead countries they don't get into the basket and so if you do that you would get slightly different portfolio weights does this work here so you now see that basically what this does is it pushes the portfolio weight of Italy down from 23% to 16% it doesn't do much to the French weight and it pushes the German weight up to almost 30% and of course then the consequences are that the share of defaulted debt if we have a crisis in this 9 above case would be below 20% and even in this catastrophic case where even France defaults we would be at 35 so basically one recommendation might be actually to show some of these things as a complement to your simulations okay I get to my final point so the final objection is even if one can deal with this mutualization argument maybe the proposal would not survive because of its redistributional effects of forcing risk weights on banks maybe it just gets too many people upset with you and so here make no mistake I'm completely in favor of imposing these risk weights that's my Germanic mind but not everyone agrees now here is where I find the paper a bit sketchy so you have this whole story where you're trying to back load regulatory reform you're arranging for the centralized swap mechanism and so what is not clear to me is to what extent does the proposed auction mechanism have redistributional implications for European banks or not can you really design it in a way that doesn't have redistributional implications how big would that be I think to discuss that would be really crucial for the political economy of the proposal if it doesn't have huge redistributional implications this might be this might be a good idea so to conclude you know I'm one of these European economists who likes to think of European reform in terms of grand bargains and I particularly like to think do that before I work for the German government so you know the Austrians, the Dutch, ourselves and the Finns we don't like mutualization but we may have to give a bit on mutualization and in return you know we get a new regulatory regime for banks that involves risk reduction throughout Europe that's a sort of nice grand bargain I've also learned in my brief time in government that these grand bargains have the problem that you may end up with absolutely no support because both sides hate you and so the question then is there a way of you know structuring your proposal that avoids this dilemma and I think you're going in that direction by saying let's back load the regulatory change and introduce the SB at the margin I think this is the right way to go I think you may need to go even further in this direction in the sense that you need to come up with a structure of the SB that would be acceptable to the Germanic mind even if in the end they do not get the regulatory reform in return at the end and that's a little harder but it may not be completely insoluble thank you thank you well it seems that we have a discussion that was almost convinced so it was not quite as critical as I imagine it would be but we have a little time but I think we have to give you the opportunity Professor Bruno Meyer or Professor Pagano maybe to have a short reply since he has the Germanic mind it's a better sell he's better suited to reply let's have a big German here so let me say this of course even if you don't if you can include all the debt I think perceptions still matter I think I would be from a Germanic mind perspective I would be very worried to say we include all sovereign debt put it together of course the liquidity problem would be the market liquidity and the price discovery in the national debt would be more complicated as well but I really think these for the peripheral countries a big benefit to have this arrangement because essentially it switches also the flight to safety across borders now it stabilizes their yields dramatically without a little cost I think that's something the Germanpund because the Germanpund loses its monopoly position in as a safe asset so there is something Germany gives up and I think if then we do something which stabilizes the whole system and we get rid of this diabolic loop I think that's something everybody should say it's a public good and I think that's why I'm still hoping for we achieve both and we hope to achieve this with a very smooth transition that's why this big swap arrangement has only banks to reposition themselves at a fair price and then subsequently we face in some risk weights whatever they are but we don't want any advantage some big advantage for ASP we just want a fair treatment through the look through principle you could do it without risk weights totally but you have been in German government you probably know better than me yeah if you give something big on the one side probably they want something on the other side too and it's something which will cost something but in the long run it will make Europe safer let me leave it at this do you want to add something no no please we're short of time we have unfortunately exhausted the time for questions we cannot have the opportunity to have questions from the floor and from some Harvard PhDs I see around there but we can continue our discussion over coffee so that to give time for the next session thank you