 Hello and thank you for rejoining for our second session of this afternoon. We know that when debt levels are too high, it stifles investment and may sow the seeds for future crises. Yet corporations, many short on cash during lockdowns, had little choice but to load up on debt during the pandemic. Economies are now reopening and some of the crisis support put in place by governments is slowly being removed. As we return to something approaching normal, how worried should we be about the macroeconomic implications of a debt overhang? Professor Moritz Schulerich of the University of Bonn has assessed this question and is now going to tell us what he found out. I'll hand over now to the session chair, ECB Executive Board Member Fabio Panetta. Mr Panetta, over to you. Thank you Claire and good afternoon to everybody. It is a pleasure to chair this session and introduce the paper by Professor Moritz Schulerich from the University of Bonn. The paper brings together two topics that are widely debated in the current environment corporate indebtedness and macroeconomic stabilization. I'm sure that Moritz will explain very clearly the findings of the paper which answers the question of whether corporate indebtedness creates key risks for a quick economic rebound. But let me spoiler this for you. The answer of the paper is no. Corporate indebtedness is unlikely to weigh heavily on the recovery. This doesn't mean, however, that corporate indebtedness does not matter for policy making during the recovery. I will take a couple of minutes to give you an example of why it does matter. In the initial phase of the pandemic, the choices facing policymakers were relatively narrow. Both monetary and fiscal policy had to support the economy on a massive scale. This support has sheltered firms from the liquidity crisis, but at the same time it has significantly increased leverage. In 2020, the corporate debt to GDP ratio rose by eight percentage points. Such headline increase in turn has exacerbated a key feature of the euro area economy, one that policymakers will have to take into account in the recovery phase. I'm referring to heterogeneity. Firms financial conditions have become more heterogeneous across countries, as the paper shows very clearly. Heterogeneity has risen also within countries, across firms and sectors of activity. Leverage has increased especially for those firms that had weak financial conditions already before the crisis. Another source of heterogeneity is the financial instrument used by corporates. During the crisis, most firms turned to their banks to cover their liquidity needs, but an increasing number of companies started issuing bonds on public markets. Taking into account these heterogeneities is crucial for center banks. We know that firms with different leverage and financial structures respond differently to monetary policy impulses. Heterogeneity also matters for other policymakers. I'm thinking, for instance, about the need to reform insolvency regimes in a post-COVID world in order to identify and resolve non-viable debt while facilitating the restructuring of viable debt. This distinction between viable and non-viable debt is crucial when financial conditions are highly heterogeneous. So, although the conclusions of the paper are relatively benign, the subject of the paper is extremely relevant for policy reasons. But let me stop here and hand over to Moritz, who will have 20 minutes for his presentation. Thank you very much, Fabio, despite the spoiler. It's a great honor and pleasure to speak here to you today. Thank you for the organizers, for the opportunity to present my work. I want to talk about corporate debt, its role in driving macroeconomic fluctuations and the implications for policy, as you said. I will take a macro perspective and a long-run perspective. I want to turn back, if you will, in order to look ahead. And that's very much the idea of this paper, namely to aggregate our knowledge about corporate debt dynamics over the business cycle. Let me start with what you also mentioned at the beginning at the introduction, Fabio, is the pandemic hit economies after a decade-long corporate debt boom and debt to GDP has risen very sharply in the recession as well. The questions I want to address today, what does this mean for the economic outlook? Especially, will corporate debt overhang restrain investment? Will roaming zombie firms, we heard about them from Victoria, will they slow down the recovery? And as you mentioned, there is a sense of history repeating in the air. Household debt overhang slowed down spending post-2008 and the experience of Japan in the 1990s where corporate debt overhang and zombification contributed to a very slow and drawn out recovery from the crisis. We're well aware of that as well. So what I'm going to do in this paper is I want to study corporate debt and output dynamics across the near-universe of modern business cycles. So it is a big picture view based on novel long-run data for corporate debt in 18 advanced economies since 1870. The data cover loans, bonds and lending for non-bank intermediaries to the entire corporate sector. And for further details, I recommend you have a look at the paper with Oscar Martin and Alan that this work is also based on and will have this data available at the JST database macrohistory.net. So with this introduction, let's take a look at the bigger picture. Let's look at take a bird's eye view of corporate debt over GDP ratios in the long term. And they are marked by financial deepening in the early industrialization, a collapse in the Great Depression and World War II and a slow and steady increase ascent after World War II. And you also see on the right-hand side quite an acceleration in the recent period and a jump about 10% points, 15% points in the last couple of years alone. We can also see is there is quite some cyclical variation in corporate debt, although this is a median across these countries, and it is that cyclical variation in the data that I want the corporate credit booms the surges, if you will, that I want to study explicitly in this paper. Another way to look at the data is to divide corporate debt to GDP changes in five-year periods and plot their distribution. And this is what I do here in this chart, which shows the distribution of all five-year changes in corporate debt over GDP in these 18 economies. And you also see vertical lines that mark the position or the distributional position of a number of important economies, France, Germany, Italy, Spain, and the US, and where the last five-year corporate credit boom fits into this distribution. As Fabius said, I think there are two main messages here from this chart. One is that for some countries, especially for the US and France, indeed the last five years, have been quite a large credit boom, I've seen that is on the right of the distribution. On the other hand, and that's the advantage of these long-run data that we can use, it's clearly not out of proportion. We have quite a few other observations if you look at the density of the distribution here. Other countries, if you look at Germany, if you look at Italy, if you look at Spain, for those countries the past five years, even included the COVID-related increase in corporate debt, have not witnessed particularly large increases in corporate debt over GDP. The idea of large heterogeneity is also confirmed. We zoom in on the past decade, on the decade since the global financial crisis, for seven economies here. You see there's three economies that we will talk in greater detail, I guess, in the following and also after this conference, which are the US, France, and China, where the corporate debt increase in the past decade has been particularly pronounced. There's a group of European economies plus Japan and including the UK here that basically have corporate debt to GDP levels at the very same level as they were at the end of the global financial crisis. In the next step, I want to look at corporate debt dynamics over the business cycle. I called it a big data approach because we have data for almost all business cycles in these 18 advanced economies. We can say something about how corporate debt shapes business cycle dynamics. The first two charts give you the main idea of the paper, namely on the left-hand side, we show how corporate debt build-ups in over five-year periods are related or correlated with real GDP outcomes in the following three-year period. Another way of putting this is we ask the question whether corporate debt changes, blacked corporate debt surges and booms predict future GDP outcomes. On the left-hand side, you can see this for, as I mentioned, the near-universe of business cycles and advanced economies. There is basically no correlation in the data. Corporate debt does not predict GDP outcomes. If you look at the right-hand side chart, however, where we repeat the same exercise with household debt, you see a clear negative correlation, which meshes nicely with recent work by Artif Mian and Amir Sufi and Emil Verna, who also point to the predictive power of household debt booms for negative GDP outcomes ahead. In the following, I will look at this core results and become a little bit more formal. We're going to estimate local projections that show us how, that gives an idea how dynamically corporate debt build-ups shape the recession severity and the recovery speed across this large picture, large set of business cycles in the modern era. We're starting with a very simple chart that shows you how a normal business cycle looks from a long-term historical perspective. We start in year zero at the peak of the cycle. In year one of the recession, we lose one or two percent of real GDP. By year two and three, we're back to the previous peak level of output, and then we continue to grow along. In the following chart, I will show you the estimates of how a large household debt boom in the preceding expansion period shapes or alters effects this average path of economies going through a business cycle. It looks like this, the red line here with a relatively tightly estimated 95 percent confidence intervals shows you how two standard deviation household credit to GDP increase in the previous business cycle, in the previous expansion phase, slows down, first of all, increases the severity of the recession, and then slows down the recovery so that between the post-household credit boom GDP path and the normal GDP path, quite a large gap opens up after four and five years. We're talking easily here about four or five percent of GDP. An experience that I think many of you in the room have also noticed after the global financial crisis where sort of persistently our forecasts for the recovery speed were too optimistic. So now in the next step, I'm going to look at the same chart, but not look at household credit booms in the preceding expansion, but at the effect that business or corporate credit surges have on the recession path. And the picture looks like this. The blue line is the two standard deviation increase in the corporate debt to GDP ratio in the preceding five-year period. And essentially the path is the average path of the economy going through, of economies going through the business cycle is hardly affected. The blue line is virtually on top of the black dash line. So there's a large difference here between household and business credit. This key result is going to be conditional is that business, corporate credit booms do not shape business cycle dynamics and that corporate debt booms do not typically leave large traces on output dynamics is going to be conditional, however, on three main caveats. And I'm going to go through these three caveats now one at a time. The first caveat is that the sectoral composition of the corporate debt boom matters. We owe this insight to recent research by Karsten Müller and Emil Verner, who showed that tradeable sector corporate debt booms do leave traces, do predict bad really economic outcomes. This is reproducing the charts from their work here for a large sample of 116 emerging and advanced economies. So it's a slightly different sample and this is really taken from their work. But you see the difference between the outcomes of the aftermath of tradeable, non-tradeable corporate debt booms and tradeable corporate debt booms. The idea being that the non-tradeable debt booms resemble very much the household debt booms. They're typically related to real estate boom and bust cycles, whereas tradeable sector booms enhance productive capacity and have positive growth outcomes down the road. With this recent research in mind, we can look at the composition of the corporate debt boom, the sectoral composition of the corporate debt boom in the past decade. And my reading of these two charts on the left hand side, again the non-tradeable sectors, this is indexed to 2015 is 100. So you can see the economies coming in and going out of the 2015 level of non-tradeable and tradeable corporate debt to GDP. My reading of these two charts is that the boom in the past decade was not particularly tilted towards non-tradeable sectors. We have not seen a classical real estate credit fuel boom bust cycle, also not in the two economies that continuously stand out here as having had particularly pronounced corporate debt dynamics, namely the US and France. Even in these two economies, the dominant part of the debt increase happened in tradeable sectors and can be expected to have more benign economic outcomes. They also happen, this is a dimension I talk about in the paper, the non-tradeable debt booms also happen to be less concentrated in asset-based lending and more concentrated in asset-based lending. So the feedback mechanism between declining and asset prices and declines in net wealth are much more pronounced whereas tradeable sector lending tends to be cash flow-based and hence not prone to these dynamics. The second caveat is that, and we heard this from Victoria earlier, is that debt reorganization frictions matter. What this graph here suggests is that business credit booms become painful or have leave traces on the business cycle and have this negative aftermath if and when they occur in economies where debt reorganization insolvency and bankruptcy frictions are high. So when it's costly and inefficient and takes a long time to resolve corporate over indebtedness and insolvencies, the real economic effects of corporate debt booms show up. You see this here in this lower line, the purple line, where we trace the effects of the aftermath of business credit to GDP to incentivization increases in high-friction insolvency regimes. Again, in the low-friction corporate debt booms basically blow over in these high-friction regimes, output costs reappear. So now what we can do is we can use these coefficient estimates coming from the long-run sample and apply them to the current context. We can say we can trace how relative to the baseline trajectory going forward from 2021 to 2024, the combination of corporate credit boom and bankruptcy frictions, which by the way are taken from this junk of hard McLeish and Schleifer paper and then updated in recent years, how these frictions combined with a corporate credit boom can be expected to affect the economic outlook going forward. And the result for looking at Germany, U.S. and Italy is there's very little we have to be worried about. In the case of Italy, the coefficient is even positive, which has to do with the fact that corporate debt in Italy was actually reduced in the U.S. and Germany. There's very little deviation. The one country to watch possibly is France. These indicators are imperfect and they measure institutional quality only with a very large noisy background. But according to these indicators, France would be a country where the GDP outlook could be a little bit negatively affected. We come to my last caveat for this overall benign view of the aftermath of corporate debt booms and corporate debt surges. It is banking supervision and the zombie problem. I'm showing here data from a recent paper by Ryan Banerjee and Boris Hoffmann from the BIS who defined zombie companies, these are publicly listed companies, zombie firms as companies who have insufficient earnings to cover the interest rate expenses and have a low stock market valuation. The point here is to show that the evidence is actually quite mixed. There are some countries, if you look at the UK and the U.S., where zombie shares seem to have increased quite substantially in recent years. If you look closer, the increase happens to be in the mining and energy sectors, but then there are a few countries as well where zombie shares in Germany, for example, trend downwards or don't do much. So the third caveat will be what is, and the zombie phenomenon obviously is very closely linked to evergreening of loans by banks and incentives for banks, as well as weak insolvency regimes, as we heard in the first paper, to avoid loss realization and reorganize the balance sheets. So what we can do is we can also ask, is there evidence in the long run data that in regimes where banking supervision is weak, the aftermath of corporate credit booms becomes more costly. And once more, if you look at the left-hand side, the purple line, there is some evidence. It's not statistically highly significant. Some evidence that after a business corporate credit boom, in a poor supervision environment, the GDP outcomes are far worse than average. And there's also some evidence that this goes hand in hand with an increase in the zombie share as measured by Bernadier and Hoffman in their paper, potentially suggesting that there is a link from the emergence of zombies, the inefficient allocation of capital and productivity, lacking productivity growth to the real macro outcomes that are plotted on the left-hand side. So let me sum up and use my last two and a half minutes to talk a little bit about the implications of what I've just shown you as I see them. First of all, the long run view summary is indeed that the aftermath of sharp increases in corporate credit is typically benign. Since this is a borderline Panglossian view, let me also come back to the three caveats, but overall concluded, none of these three caveats currently raises big red flags. First, in advanced economies, the sectoral composition of the post-global financial crisis corporate credit was not tilted towards non-tradable credit. We have not seen the boom, bust, real estate dynamics that we've seen, for example, in Japan. Again, we've also talked about the caveat that debt reorganization frictions might lead to larger output costs than the average path that I showed you. And it's true, and we've heard this earlier, that weaknesses in debt reorganization regimes persist, and Europe has some way to do that, to go there, to reach a truly federal regime. But an important point is that the countries that don't score well on this indicator that are often named as countries that have somewhat sluggish, inefficient, and slow debt reorganization and solvency liquidation regimes are not the ones that boomed in the past decade. So Italy, Spain come to mind where corporate debt was actually decreasing. The main worry that I continue to have is bank supervision. It's definitely much improved, and we're not in the 1990s Japan scenario. But zombie lending is undead in Europe, as recent papers have shown, and I think that's a part to keep in mind. And of these three caveats, the one that, in my view, would be the one to worry most about loss realization and balance sheet strength are very important. We've learned this from the sluggish cleanup after the GFC. What are the implications of this? Well, the fears of a post-pandemic headwind to growth caused by corporate data wing are likely unfounded. Efficient reorganization liquidations framework are crucial. I can just underline what we heard earlier. In terms of corporate zombies, the evidence is mixed, but supervision and loss realization is crucial. And the last point is that is important to me that this debate about monetary policy in zombies is not only about zombies. A very nice recent research has shown that aggregate demand conditions are paramount for the success of new firms, of startups and firm formation. So to the degree that monetary policy creates the conditions for stable and strong aggregate demand, it effectively helps to create the environment in which new businesses can thrive. And that's very much the opposite of the zombie fears that are often raised. Thank you very much. I close here. Thank you, Moritz, for this very clear presentation. I'm sure that your findings will spur an interesting debate. To kickstart this debate, we will have a 10-minute discussion by Egon Zakraszek from the Bank for International Sediments. While Egon prepares his presentation, let me remind participants that wish to ask questions at the end of Egon's presentation, they should get ready to raise their virtual hand. And please do not be shy. With this, Egon, the floor is yours. Sorry about that. So thank you, Fabio, for invitation and to the organizers for inviting me to discuss this very, very interesting paper. As was pointed out, in the years leading to the COVID pandemic, corporate debt levels in advanced economies have risen substantially. And not surprisingly, the nature of the COVID-19 shock has led to a further significant increase in corporate debt. Just to put some context around there, the panel to the left, the red box and whiskers plot shows the distribution of the 2019-20 change in credit to non-financial corporates normalized by the 2019 GDP. You can see for the median advanced economy in this sample, that increase was 10 percentage points, as a lot of heterogeneity with some countries registering very significant increases. By way of comparison, the corresponding median increase in the credit to household sector was a lot more muted, around six percentage point, and it's also coming from lower levels reflecting kind of post-GFC household sector deleveraging. The panel to the right looks at this as slightly different and asks, you know, which countries borrowed most during the pandemic. So it just shows on the X-axis the 2019 credit to GDP ratios for both sectors, against the corresponding change between 2019 and 2020. Okay, and as you can see, there's clearly a positive relationship, so saying that countries that came into the pandemic with high levels of debt on average borrowed significantly more. In particularly, there's countries up in the upper right corner, Belgium, France, and Sweden, that registered very, very sizable increases, pandemic-related increases in corporate debt, and those were coming on the top of already very high levels leading into the pandemic. So what this session is about, and what Morris's paper is really about, is how concerned should you as policymakers be about these recent developments? As we heard, according to Morris, not too much, and why is that? So the real, the answer that comes from 150 years of macro-financial history for 18 advanced economies, and what the, you know, striking finding that emerges from this research is that that unlike household debt booms, corporate debt booms do not influence both peak GDP dynamics. In other words, corporate debt booms are not systematically followed by deeper recessions or weaker and more sluggish recoveries. That's a very striking and a new finding. There are, however, important caveats, and Morris discussed that, so I'm just briefly summarized that. The first one is that composition of corporate debt matters, right? So if you have, if the corporate debt buildup is very concentrated in a non-tradable good sector, that should be worrisome. Those things, those kind of corporate booms look a lot more like household credit booms. The other thing that matters, and we heard it also in Victoria's presentation and in earlier discussion, is that insolvency regimes must be efficient, which makes sense because it's the restructuring of the firm's balance sheets that's going to get you out and allow investment and output to recover after a shock. The other point that's also discussed in the paper is that bank-centric, as opposed to market-based financial systems, also tend to be more vulnerable to corporate credit booms. And this point has been made before by Alan Greenspan, the so-called spare tire argument, and there was some interesting research by my colleagues at the BIS that has looked at those things. And those things are kind of a tight to banking supervision in the sense that a weak supervision allows banks to kind of extend and pretend policies that can lead to zombie lending, which then have deleterious effects on economic dynamics. I think this is a very interesting thought provoking in a timely paper, obviously. And I do want to call out that the construction of this credit series that Moritz and his co-authors have done is a really, really important contribution to the macro history database. It's a great public good. So, Moritz, thank you for doing that. I will have two comments. The first comment is best summarized by a quote by Mark Twain, who said, history never repeats itself, but it often rhymes. The question is, how well does current situation rhyme with past history? And then my second comment will be a little broader. This is that I will argue that fluctuations in aggregate credit quantities, those are the kind of things that Moritz and his co-authors are looking at, provide an incomplete account of credit cycles. And I'll be more specific when I get to that slide. So, to give you just kind of my bottom line, what is my view is that I am somewhat less sanguine than Moritz. And I think that the current corporate debt buildup presents a downside, tail risks to the economic outlook. So, how informative can history be about the COVID-19 fallout? I would argue that maybe not too much. And one of the reasons is that the COVID-19 shock is very, very unprecedented, right? It has a very strong supply side dimension, reflecting all the economic and social restrictions imposed by the authorities to deal with this health crisis, public health crisis. On top of that, monetary and fiscal responses were coordinated. They were unprecedented. They're still ongoing. And importantly, the support programs, they were enacted different countries, different across countries. But in general, they were of very broad scope and limited conditionality. So, I think this just kind of really complicates how to think about this buildup and how we're going to get out of that for the near-term business cycle dynamics. And I think, and this point was echoed by Fabio in his introductory remarks, is that I think a much an approach, while history is certainly, I'm a big fan of this kind of approach, but I think for this, looking at with firm level heterogeneity, looking by industry size, financial conditions would be really, really important. Now, what's interesting is, is there has been some very interesting work done, and it's a very important, I think, French case study. And France is particularly interested, because if you remember from my first graph, right, it's a country that exhibited a big pandemic induced buildup in corporate debt, and it entered a crisis with a high level of corporate debt to GDP. So in March 2020, the French parliament established a committee which chaired by Benoit Corre, who's your former colleague, is now colleague of mine at the BIS, and they looked to evaluate and monitor, you know, the financial support available to French companies during the crisis. And what's, I think, really great about these studies is, is that his team combined firm level information on the take-up of credit support scheme, and in match that with their corresponding firm level balance, income and balance sheet data. So very large comprehensive database, 3.5 million French firms, and they're analyzing what happened, you know, to these things during the first two waves of a pandemic. There are a lot of interesting results. I sort of pulled out, I think, three points that I think are very germane to today's discussion. So the first one is, is that the intensity take-up rate, so the amount of support a firm received relative to its turnover was highest for financially weakest firms. So that's not great news, right? So that means is that the pandemic-induced build-up was very much concentrated in a part of the firm distribution that was already financially weak. The second point is, is that the share of the money that was paid out to small businesses was higher than the share of employment, okay? So it means that, you know, the credit schemes were, you know, the amount dispersed went very disproportionately to small businesses, which could very well make sense since they were in industries or sectors that were most hit by the pandemic. But it also brings up a point that was discussed earlier that you have a lot of small businesses. If there is an avalanche, a cascade, a highly correlated set of defaults, this could potentially overwhelm the court systems and, you know, impede an efficient resolution of these things, and that could slow down the recovery. Now, on the good news, is that pre-crisis zombies, so defined in a traditional way, did not make a disproportionate use of the credit support scheme. So that's on a good news. But, you know, I think as, you know, Victoria's research showed, and there's other stuff that has been looking into this stuff, defining zombies is not completely trivial. And, you know, the way you define it can give you very, very different results. But at least, you know, in this case, I would say this is a little bit of a silver lining. In my view, a lot more similar analysis using firm level data across different countries is needed to ascertain the likely impact of this corporate dead buildup on post-COVID macro economic dynamics. So history can teach us something, but I think I would be much more comfortable if we had many more French case studies done. My last comment pertains to what I would call an integrated view of credit cycles. The post-GFC research on credit cycle over which Schuler, it was Moritz and his co-authors were instrumental to kicking off, you know, convincingly demonstrates that at low frequencies, so three to five years, rapid sustained credit builds up, you know, predicts economic downturns. A point that is less appreciated, maybe not quite well, you know, no, is that investors' sentiment in credit markets also carries negative information about the future economic growth. Above and beyond that contain in credit aggregates. Okay. So what do we mean by credit market sentiment? So what this literature refers to, it refers to variation over time in expected returns to bearing credit risk. So when we say that expected returns to bearing credit risks are too low, it's equivalent to saying that credit is priced too aggressively. That's manifested to a narrow credit spread, lots of high yield bond issuance and easy laying standards. So it's not just the quantity, but it's also the pricing. Now what's particularly interesting, and there's a lot of interesting theoretical work and empirical work done on this is that exposed investors are predictably disappointed. In other words, this credit market sentiment is mean reverting. Okay. So credit is priced too aggressively, it will mean reverting. And that tends to lead to abrupt and large evaluation of credit related assets. It leads to supply in a pullback in the supply of credit and the downturn. Right. So by just focusing on credit quantities, this part of the credit cycle is missing. I do think that that's important. I don't have historical data, but I did a little very simple exercise just to finish up is I looked at the U.S. In the U.S., we have a reasonable measure of credit market sentiment over the last 50 years, something called the excess bond premium. And the way you want to think of this is it's a corporate bond credit spread net of default risk. So being net of default risk has a natural interpretation as a measure of credit market sentiment. So what I have done here, the panels to the right, just show what happens if you have a standard, if you have a sustained corporate credit boom that's in red, okay, that tends to be followed, okay, predictably by a reversal in credit market sentiment, okay, over the near term horizon. There's less so for the household credit booms, but this is only U.S. data where you don't have very many household credit booms over the last 40, 50 years. So another risk that I think policy makers should particularly be attuned to is that a sharp and sudden reversal in sentiment for these highly leveraged economies could be problematic. And there's very nice recent work by Arvin Krishnamurti and Tyler Muir who looks at this dynamic in this interactive way, employing long history data and finds important nonlinear interaction effects. Thank you. Thank you, Egon, for this thought-provoking discussion. Before opening the floor for the Q&A session, let me remind the House rules. First, participants wishing to comment or ask questions should raise their virtual hand. Second, participants should limit their intervention to 90 seconds maximum to allow for an active discussion, okay? We already have four persons from the audience who have asked the floor. The first one in my list is Governor Francois Villouard de Gallo from the Bank of France. Francois, you have the floor. Thank you very much, Fabio. I hope you can hear me. Very well. Okay. And thank you to Moret. So in 90 seconds, one remark, one strong interest, and one question. The remark is about methodology. On your first graphs, Moret, it's highly dependent on the definition of debt you use. If I take the case of France, which was often quoted, if you take consolidated data with economic groups and not only individual data, the picture is quite different. And if you take net debt and not only gross debt, the picture is completely different for the last two years. In our case, gross debt increased significantly indeed. But as cash holdings increase in the same proportion, net debt remains stable. Second, my strong interest is for your cavert number one. And I think you have something highly interesting there to say that that intradable sector does not matter that much. Trade in non-tradable, linked to households, probably matters to diminish the GDP potential. And the housing sector and mortgages is a very interesting test for your hypothesis. So we should dig upon that. And my question to conclude with, I would tend to agree with your idea that evolution with the caveats is not that important. But the absolute level of corporate debt to GDP matters and still more its opposite, so to say, or its complementary, which is the absolute level of equity to GDP. And here, clearly, a firm which is financed through equity will be more risk friendly and more read it in a way. And this is not only at micro level. We see it, unfortunately, for us Europeans at macro level, corporate equity to GDP is twice higher, at least twice higher in the US and in Europe. And the capacity to innovate is obviously stronger. So I would like you to elaborate on that. Thank you, Francois. Maurice, you want to address this question? Yes. Thank you very much, Governor. I'll point on the data as well taken and also think that maybe I can blend in what Egon in his discussion also mentioned. Thank you, by the way, Egon, for that great discussion, is that the absolute level of debt and the question of tail risk. So these are two things that we looked at, what I looked at for the paper. It turns out that the tail risk dimension, if you look at quantile regressions, is also not elevated after corporate debt booms. So corporate debt booms don't make the bad recessions even worse. And the level of debt, so controlling for corporate debt to GDP levels at the onset of the recession, just very little. It matters a tiny little bit, but it doesn't change the overall picture dramatically. But this being said, these two little clarifications, I think your point is very important and very well taken that the gross debt perspective that we take in this research is an ideal world. We could complement this by having the complete view of the balance sheet structure of these companies and even look at the distribution ideally of financial vulnerabilities and balance sheets across the firm distribution. I think that's very important research that has advanced in recent years. We know, for example, that some companies and the tax sectors are very cash rich and look very healthy, whereas others have increased leverage quite a lot. Last point, maybe also on Egon's point about sentiment, I'm absolutely, I think this is a complementary and a very important way. I'm a great fan of that line of research as well. And it is true that in recent years, many indicators of credit market sentiment before COVID have flashed red in terms of covenant light loans, in terms of spreads, in terms of the leveraged loan business. So I guess this big picture view that I present here, as you said, always has to be, this time could be different. But I think it's important to just say like, yeah, it would be the exception from the rule. The rule is more that corporate debt booms are certainly not as drawn out and not as harmful as household debt booms tend to be. Maybe I'll leave it here to leave time for more questions. Thank you, Moritz. The next quick question for a quick answer is from Professor Kalemli Oskar from the University of Maryland. Thank you very much. Excellent paper and excellent discussion. I'm going to make three remarks. First one is on COVID. And as the introduction pointed out, heterogeneity is very important. And if you want to do something that makes sure of that is the COVID shock. So purely, we're going to meet Sandra and Verneka-Pancheloka. We did several papers, one including in the Jackson Hall Symposium of this year, very similar to the French paper that Econ mentioned, but we did 27 countries, sectors, and firms balance sheets. And what we find is actually under COVID because of a tremendous policy action, neither corporate debts or rank nor zombification are issues. But this doesn't mean these weren't issues historically. So this brings me to my second comment and Moritz's main conclusion that this wasn't an issue historically. My paper on corporate debt or rank in Europe with look 11 find a very serious impact on sluggish recovery and low investment and output of corporate debt for the 2008-2010 crisis in Europe, which was, of course, a financial crisis. So I would like to get Moritz's views on that. I mean, basically coming to my last comment that the importance of bank-based systems, financial frictions, and the fact that small firms are extremely important. Many of the papers mentioned like Benerjian Hoffman only use listed firm data. These firms are not really financial constrained firms. And they have that, but they are other sorts of financing too. When we think about SMEs, which constitutes a large part of corporate debt in Europe and also in US in terms of bank loans, and also more than 30% of employment and more than 50% of active output. As policy makers, we should be very concerned about both loans and bonds of SMEs and need to think about ways how once we bring that into the picture, are we still going to get the view that these things don't matter for active outcomes? Thank you. Moritz, you have the floor. I would can be asked to have quicker questions and quick answers. Thank you. I'll be quick my answer. Hi, Shevnam. Thanks so much for the comments. They are very well taken. I know you work with Luke and others was very important. And I think it's it's fair to say there is a certain tension between the micro and the macro evidence. And I'm a great fan of these micro studies. And I think on the firm level, these effects are measurable. And there's a certain tension that apparently, at least from this very like bird's eye, high flying altitude perspective that I have in this paper, these micro effects do not translate into measurable macro effects that could have all kinds of reasons that could be general equilibrium effects. Yes, there are zombies, but then other companies come in and invest more and take up their business. Prices can adjust. I think I guess we're not going to solve this today. But I think it's important for policymakers to be aware that the micro evidence, yes, I would absolutely agree. There is from your work and other workers evidence that on the micro level debt overhang on the company level is a thing. And inhibits investment and firm growth. On the other hand, there is a gap between the micro and the sort of the macro. And we're not quite sure why these micro effects do not generate larger macro effects. On the second point, bank based systems, I agree small firms that regenerate, all these things matter a great deal. And I think all of those would probably argue that it's important for Europe and the European supervisors and policymakers to keep a close look at the small firm situation, at the situation at banks. As I mentioned in my talk as well, I think that's the main of these three caveats, likely the main risk right now for the outlook. Thank you. Thank you, Moritz. Next question is from Professor Emke from the London School of Economics. Professor Emke, you have the floor. Thank you. Thank you. I had trouble on reading myself. Sorry. My comment follows up, I think on what was just discussed also after Shannon's comment, you know, in the work that at the ESB ASC we did on bankruptcy, I think one of the worries that came out of our studies was that if a large proportion of this falls onto a small very small firm's stand and this, you know, links back to the previous paper we saw, then perhaps the, at least the formal, but perhaps also the informal bankruptcy and solving the solvency procedures that we have may not work as efficiently there. And so I wonder whether Caviar, I think it was Caviar too, in all of this paper, is something that deserves particular attention this time around that we interested in your views on that. Thank you very much. Thank you, Moritz. I'll endorse that. I think that's an important point. And it comes back to sort of, let's, we need to try to get the micro and the macro together, I think in the current context. And we have it from Victoria's excellent paper earlier. Insolvency and dead reorganization regimes are crucial. They are, I think, understudied in the literature. And but they matter a great deal, as is the question of asset-based versus cash flow-based lending in the corporate space where I think we don't have very good cross-country data. There's a recent, very nice paper for the US, but for other countries, we don't know that. And to the extent that we get these feedback cycles between declining asset prices, hopeful recovery, evergreening loans, you could, you know, you could have the suspicion that this is much more prominent in asset-based lending situations where, you know, you wait for recovery. So there's a lot to be done, but there's no margins. Coins are very, very well taken. Thank you. Next question is from Professor Sissova from the Catholic University of Luven. We don't hear you. You may be muted. Now, do you hear me now? Yes. Yeah, great. Just a small correction. I'm not a professor at Kyiv Luven, but I'm a PhD student, so I'm one of the participants of the Young Economist competition. I have a small question about the caveat theory of Professor Sholaryka presentation, and in particular, the zombie lending. So in that respect, it may be also related to the first presentation of Professor Ivashina. So my question is about whether we should see zombie lending as only a negative thing for the economy in the sense that I know there is recent evidence using European data that actually zombie lending can have a disinflationary effect in the sense that due to this extension of credit to additional firms, we actually create excess productivity capacity, and this pushes the prices down. So I just would like to hear your opinion on this potential effect and whether maybe policy makers in Europe could actually use zombie lending, which seems to be present in Europe, to decrease the inflation. Thank you. Moritz. I think that's maybe the last part of the more questionable policy makers for you, Fabio, but I think the idea with zombie lending is that it inhibits restructuring and the efficient allocation of capital and thereby slows down productivity because I think there is evidence for that, and that's why we're mainly worried about it. But at the same time, I think if I can make this point again that I made at the end of my presentation, these phenomena to the degree that they're often linked to accommodative monetary policy or low interest rate environments, I think they miss an important point, and that's come out of recent research by my former colleague and a student from Bonn that aggregate demand conditions are extremely important for business formation and the success of startups. So this opposition that's often created between low interest rates, inhibit business dynamism, firm creation, and creative destruction has it all wrong. It has the science wrong in the sense that you need a good aggregate demand environment to help startups thrive and thereby lead to the very creative destruction and productivity growth that we all want to see. Thank you. We now have the last question from Francesco Nymphole from the newspaper Milano Finanza. Francesco, you have the floor. Thank you. Good afternoon. My question is about caveat number three, because in your research, you have underlined the importance also of precautionary capitalization along with stringent supervision. Based on this conclusion, would you suggest an enhanced role for precautionary capitalization and maybe a more flexible framework than the one we have seen in Europe? Thank you. Moritz? Yeah, thank you very much for the question. It is important and I discuss it in my paper in greater length than I could in the presentation. I think the evidence is quite clear that weekly capitalized or weekly supervised banking systems see more evergreening of bad loans, see more zombie creation. Victoria added another important dimension, which is, again, there is also the role of bankruptcy and insolvency regimes that can also contribute and that then prevents the exit of impaired businesses and depresses productivity growth. I think we or ourselves as Europeans a very close second look at the situation this time around, because I think after the global financial crisis, the cleanup of bank balance sheets was not optimal, was not fast enough. And again, I think the worst policy, and that's true also from my research that I presented today in terms of the aftermath of corporate credit booms, the worst possible policy is a policy of kicking the can down the road. To the extent that precautionary recapitalizations are a way to deal with this, their political economy of those is very difficult, obviously, and credible stress tests are the important and best tools to deal with this. But I can just I agree with you that we have some homework to do I think on this front. Thank you Moritz. So we have no more questions. And as we come to the end of this panel, let me share with you my personal takeaways. In the coming months, it is paramount that we prevent with our policy actions the emergence of an adverse feedback loop between insolvencies, financial conditions, and the real economy. The problem, of course, is how to do this. I believe that the paper suggests two policy conclusions. The first one is for the European Central Bank. The best way for monetary policy to prevent adverse feedback loops in the recovery phase is to effectively support aggregate demand, avoiding premature actions that could drive up bankruptcies and start a vicious feedback loop. In other words, the benign results of the paper are conditional on policy. The second conclusion is for other policymakers. Enhancing corporate insolvency frameworks across the European Union is crucial. I thus hope that we could move fast on the initiatives on this front foreseen by the action plan for a capital market union presented by the European Commission in 2020. On this note, let me thank once again Moritz, Egon, and all participants who intervened for this very rich discussion. Over to you, Claire. Thank you, Mr. Panetta, for those very insightful closing remarks and the excellent chairing of the discussion today. Thank you to Moritz and Egon, too, who, despite the chair's spoiler, managed to, I think, provide much additional detail, which I think is going to be of interest to our participants. I'd also like to thank the participants, too, for asking so many insightful questions in both this and the previous session. We are going to now take another break. We're going to be back again at five Central European time to try and unpick one of the big topics of the day, which is the future of inflation. Join us again then. Thanks a lot. Bye.