 So good morning to both of you. It's a pleasure to have you here. Our next speaker is going to be Sebne Kallemi of Cannes. We are becoming as well as a sort of tradition here. He's Professor of Economics and Finance at the University of Maryland. I'm not going to refer to the multiple positions that you hold now. And you're going to have produced a very interesting paper for us about the studies on the effects of the pandemic on the euro-reinflation in comparison with the US. That I think that is a very important topic now. And afterwards, Gabriel, you will be discussing the paper of sentence. You are Director of the Austrian Institute of Economic Research in Vienna and Professor at the Vienna University of Economics and Business. Gabriel is also a member of the Scientific Advisory Board of the German Federal Ministry of Economic Affairs and Energy. So you have the floor. Thank you very much. Thank you very much. It's a pleasure to be here and to present a distinguished audience. Our paper on global supply chain pressures, trade and inflation, quantifies several issues raised by Madame Lagarde during her introductory speech. It is joint work with Juliane Giovanni, who is here. Also with Alvaro Silva and Mohamed Yildirim, who are joining us online. During the past two years, the world has been living in a pandemic cycle characterized by a series of negative sectoral supply shocks, relative demand shocks, that policymakers responded quickly and aggressively stimulating aggregate demand, unevenly across countries, which ended us witnessing large swings in economic activity characterized by collapse and rebound in domestic demand, GDP, international trade, substitution of consumption from services to goods, back to services, uneven labor shortages, and so forth. All of this played on a global scale, leading to supply chain bottlenecks and inflation. So what should monetary policy be doing in such an environment and how effective it can be are questions that we are all searching answers for. And our paper is going to argue that a first step in answering this question is about quantifying relative importance of supply and demand underlying that inflation. So our paper takes the first step at this issue. We are going to ask three questions. First, how important were sectoral supply shocks in driving inflation across countries? They are going to be extremely important, and their effect is going to vary across regions. Supply shocks are going to account for half of the observed euro area inflation between 2019 Q4 and 2021 Q4. And they are going to account for a third of the observed US inflation during the same period. Our second question is going to be, how important were foreign shocks and global bottlenecks in driving inflation in the euro area, given the larger openness of euro area to trade relative to the United States? Of course, they are going to be very important. Foreign shocks are going to account for a third of the observed euro area inflation during this period. And finally, we are going to revisit this complex phenomena of very strong recovery in trade coexisting with persistent supply chain bottlenecks. We are going to show you that compositional shifts in consumption playing out on a global scale is going to be very important for this. Trade actually didn't respond to the GDP changes during the recovery from COVID as strongly as it did during the previous episode of Great Trade Collapse and Recovery of 2008 and 2009. So our results are going to show you dynamic, consistent with a short-lived trade collapse, strong recovery, and a long-lived persistent global and domestic supply chain bottlenecks. Let me start by showing you some stylized fact before going details of our setup that will show you clearly the pandemic-related labor dislocations is bound to show up as inflation when combined with aggregate demand stimulus. Starting with inflation and employment. This is a very important figure. I'm going to show you euro area on the left, United States on the right. And although employment recovered with GDP during the recovery, you can see that purple line employment was still below the pre-pandemic level as of 2020 quarter four when the pink line, which is headline inflation, started increasing, both sides of the Atlantic. This figure is about the global supply chain pressures. This is an index plotted in black here on both figures constructed by my colleague, Julian DiGiovanni, with his colleagues at New York Federal Reserve. This index brings together several measures of supply chain pressures, such as shipping costs, delivery delays, and freight rates, and then purge demand from these indicators using new orders to understand the supply chain pressures. We plot that together again with headline inflation in pink, euro area on the left, United States on the right. You see that they increase together starting their ascent early 2021. This early increase and collapse in the supply chain pressure index is about early Chinese lockdown beginning of 2020 and then the reopening. Now, this dynamics is going to be very important because as you will recall, during this period when supply chain pressures increasing, inflation is increasing, trade also increased. This is going to make it very clear that looking at trade as a metric to decide the persistence of supply chain pressure is going to be a misleading thing to do. Trade, of course, collapsed and then rebounded. What I'm going to do on this slide is to show you the amazing country heterogeneity in this narrative. On the top, I'm plotting euro area imports and United States imports, real imports, from the drop from the 2019 peak and recovery in pink. This is the COVID episode. Compare that to global financial crisis in the purple. You can see that the heterogeneity in imports clear. There was a steeper drop in imports in euro area less so in United States. This goes back to this compositional shift started early on in the pandemic, stronger in US where consumer demand switched to goods and stayed strong and met why the imports from China. In the bottom, I'm plotting Germany exports specifically and the United States exports, real exports again. Again, the pink is COVID episode, purple is the great trade collapse of 2008. The recovery was quicker but never back to fully normal to the horizontal black line indicating of limited production capacity and supply chain destructions. Inflation is, I'm sorry, consumption, compositional changes in consumption is going to, of course, match these trade patterns and inflation is going to match these patterns too. Here's consumption. On the top, again, euro area and United States real consumption, the steep drop during COVID in pink and nothing going on in global financial crisis. Now, what is important is the compositional change driving this consumption drop. The bottom, euro area and then United States decomposition. You can see the purple is showing your durables, pink non-durables and orange services. When durables start to recovering very early in United States and then the Europe, services were lagging and as you see, we do everything till the end of 2021 and it didn't recover at all back to the 2019 levels. This is going to be reflected in inflation patterns as you see in the slide. I am plotting headline inflation and core inflation for euro area in pink, United States in purple and UK in orange on the top figures. Again, inflation start picking up in 2020, steeper and little bit earlier for the United States but what is important to understand what really happened is the bottom figures, services inflation on the left, goods inflation on the right. You see that at the end of 2021, services inflation at there 3% whereas good inflation both in euro area and UK up to 8% in United States up to 10% at the end of 2021. Let me show you a simple schematic that summarized this database narrative that will show you how global scale, supply and demand imbalances work out. Let us all remember the primitive shock here is a global health shock. When COVID started, negative supply shock due to workers contracting disease, taking care of people with COVID, lockdowns started the supply chain disruptions. Demand switch from services to goods because of fear of going to restaurants, unlimited mobility and uncertainty. Then aggregate demands stimulus came that intensified this compositional shift from services to goods, made supply chain disruption much worse because those disruptions by essence talks about a disequilibrium in balance between supply demand and ended up with inflation. So how do you quantify this? We would like to quantify this in a closed economy and an open economy. In a standard model, these type of sector level shocks are transitory and a sector's demand is high relative to another sector, a sector productivity low relative to another sector. These are no issues for monetary policy, relative prices is going to adjust and there is no need a response from monetary authority. In a world where such factor mobility is limited, factors are complements in production and some of your factors are imported, a country can have inflation because of both domestic shortages and global shortages. We are going to use the macro network model, closed economy model of Bakai and Farhi recently published in American Economic Review. This is a multi-sector two-period model. It allows all the realistic complementarities in production, a full-fledged inclusion of the full intersectoral IO linkages depicted in this figure. This shows you the linkages for 45 industries so you can see. The color coding in this figure tells you the tradable industries such as electronics in red, non-tradable industries such as wholesale and retail or construction in pinkish color. You see that, it really doesn't matter if you are tradable or non-tradable. We all saw this during the pandemic when construction sector had to stop production because they couldn't get the parts even it is a non-tradable sector. There are a lot of non-tradable services sector that is going to be sourcing from other sectors domestically and internationally. Now, take this figure and embed this in a production structure in a standard, well, not standard, a little bit complicated but a macro network model where there are layers of production. Consumers consume a variety of goods, cars, groceries, services, other durables and all. Those varieties are produced with intermediate goods and capital and labor that are sector-specific. Intermediate goods use other sector-specific bundle, a car needs, plastics, steel, glass, labor and so forth. Normally, we stick each layer of production with these elasticities in economic models, tetas, epsilons in this figure. We learned early on that applying a capitalist production function in the short run wouldn't be the right thing to do because that production function assumes substitution and will not be able to capture the realistic difficulty in terms of substituting among inputs and suppliers which we live through. So what we are going to do in our calibration exercise to decompose your area and US inflation use realistic short run elasticities estimated in the literature. We are going to focus on this period, 2019 Q4, 2020 Q4. Why? Because we would like to capture both collapse and recovery. We would like to decompose, observe inflation and we would like to stay away from the developments in 2022 with the energy and the war issue. The Bakai Fahri model is going to allow us to observe inflation under a rich set of shocks. Agate demand, we are going to pull it out by matching to observe inflation, sectoral demand and sectoral supply shocks. For those, we are going to use data on sectoral consumption and sectoral labor hours worked. The model is also going to allow us to decompose inflation into parts coming from the supply side and demand side at a sectoral level. So the key intuition can be summarized here. CPI inflation can be mapped to two objects in a full flat general equilibrium model. Agate demand shocks in a closed economy think this has nominal GDP expenditures and a minus entry of employment changes. Now these observed employment changes in the data is going to come both from sectoral demand and sectoral supply shifts, but it doesn't matter because it enters with a negative sign. When these quantities goes down, when we have low labor shortages, that's going to be inflationary. That's the key insight in this model. Employment behavior is going to be very important because we think it by sector instead of an aggregate labor market as we always do with our one sector models. There's a potential level of employment, L-bar, and that can go down because of workers getting sick and lockdowns. Then there is an equilibrium employment level L and that can change because of demand effects and downward nominal wage casualties. If wage cannot fall, then there has to be unemployment. This simple figure demonstrates this. Very simple two-period model. You start with equilibrium pre-COVID, labor equals potential labor. During pandemic, potential labor goes down. Equilibrium labor also goes down. That's what has been called Keynesian unemployment by these co-authors. Once aggregate demand stimulus comes back with the policy response, of course that red line you see there is going to go back up and after the pandemic you go back to normal. Let me show you our result using this setup. We are going to decompose observed headline inflation in the euro area of 4.69. You can see both the figure on the left using the full network of 45 sectors of the euro area and the figure on the right which aggregates them to three sectors tells you the same story. Half of the observed euro area inflation is driven by sectoral supply shocks in the purple column. The orange column tells you inflation would have been around 3% without the extra two plus percent added through the sectoral supply shocks in euro area. United States, we work with a higher observed headline inflation of 8.47. The quality of the result is the same. Sectoral supply shocks is going to be very important but their importance is less than the euro area. One third of the observed US inflation is explained by the sectoral supply shocks. You can see in the red column inflation would have been around 6% in the United States without the 3% coming from the supply side shown in the purple column. Of course, these are open economies and the networks are global. So here I show you that pink figure again linking sectors in an economy. These are structural IO linkages. I want to envision that pink figure inside the blue figure. Blue figure is the trade linkages between the countries. The color coding is such that open countries like Ireland is going to be very dark blue. Less open countries like US is going to be light blue. Size of the boxes is going to be the size of the countries and the thickness of these arrows is going to give you the intensity of trade relationship between the countries. Of course, once you embed that sectoral linkages into these trade linkages, you have this very complex global trade and production network. We are going to use this data to understand the role of global bottlenecks on euro area inflation. You can see euro area countries are going to have a very, very important place in this network. How do we do this? We are going to follow the model in my earlier work that makes the economic case for global vaccinations which extends Bakai-e-Fahri close economic model to a multi-country setup. To give the message very clearly here for euro area inflation, we are going to focus a three-country setup. Euro area, United States, and everything else, all the other countries, all the 65 countries you saw on that figure, summed together in the rest of the world category. Then we will work with three scenarios. All of these three countries get all the shocks, aggregate demand, sectoral supply, sectoral demand. Shocks are only in the euro area and shocks are only in the United States and the rest of the world. And here is the result that shows the importance of global bottlenecks on euro area inflation. They explain two-thirds of the observed euro area inflation. You can see in the red column that euro area inflation would have been 2% at the end of 2021 without the 3.26% that comes from the outside shocks. That yellow column assumes shocks are only in the United States and the rest of the world, nothing going on in the euro area. Let me close with the last point and then with the policy implications. Given the importance of these foreign shocks for euro area inflation, can we explain this simultaneous occurrence of strong recovery in trade and persistent supply chain bottlenecks? In fact, this observation in the data led to Iran's thinking that supply chain issues is going to clear very quickly, will be transitory. We are going to use an accounting framework that was used before by Bams Johnson and Eid to explain the great trade collapse and recovery during 2008 and 2009. This framework is going to give you expected international trade flows by looking at the changes in final demand that is going to initiate changes in imports and exports, also changes in the intermediate demand that's used in the production of that good that you are changing your final demand for, as shown in this equation. That will allow us to calculate export and import elasticities relative to GDP changes, both doing the collapse period of COVID, 2019, 2020 year on year from second quarter and recovery period 2020, 2021. We do the same exercise for global financial crisis, for collapse, 2008, 2009 and recovery 2009, 2010. Very interesting pattern here. The top is great financial crisis associated with great trade collapse. The bottom is COVID pandemic. What you see in blue there, this is data. The left side of the table is data, right side is the model. The data tells you, as you can see, the bottom numbers are much smaller than the top numbers. What does it mean? It means that trade responded much less to the GDP changes during COVID when you compare to the previous episode of great financial crisis. And the model numbers on the right in red tells you the same story. Model numbers are smaller than the data numbers, of course, because model cannot capture the whole reality. But the fact that this very simple model tells you there is less response of trade to GDP changes during COVID relative to great financial crisis. And we also show that intermediate goods trade played a bigger role during COVID. This is, of course, maybe not that surprising as they are very different shocks. One is a global health shock, played out unevenly by a country sector. The other one is a financial crisis. Let me finish with the policy implications and what does all mean for central bankers. We are writing the paper for central bankers at the end of the day. Aged demand stimulus in a supply-constrained world has larger inflationary effects. We demonstrated this clearly in the paper. But central banks operate in an aged demand world. A Phillips Curve-based one-sector model is going to have a very hard time in trying to separate a series of sectoral supply shocks that can have permanent effects from a one-time permanent aged supply shock that will go to white star to potential output in the Phillips Curve. We argue that central banks can make use of these macro network models that will give you a way to deal with a symmetric sectoral supply, relative sectoral demand, and aged demand shocks. Normal wage rigidity in these macro network models combined with these supply shocks leads to labor rationing. Then you have your cost plus shock starting in few sectors, but then becoming Broadway's inflation. The theoretical underpinning of this story is done by Laotapa Selai, Bakai Fahri, Gurari et al. papers. And in fact, Emmanuel Fahri was a very good friend of mine, and this was the last topic of conversation we had before he passed away in 2020. We were talking about how exciting it would be to apply their theoretical framework to quantitative approach and use it for policy making. This is what we do in this paper, and we are really proud of doing it, quantifying this for your area and US inflation. To conclude, aged demand stimulus would not have produced as high an inflation as the one we observe without the negative sectoral supply shocks. I would like to remind you our numbers. Half of the observed euro area inflation, one-third of the observed US inflation driven by sectoral supply shocks. Limited factor mobility and complementarities in production, these are byproducts of the pandemic. This was a global health shock, worked out unevenly by letting firms having difficult times in reallocating labor between sectors, finding new suppliers, switching suppliers domestically or internationally, leading to global supply chain bottlenecks and rising prices. The early belief that the strong recovery in trade is going to mean we are going to get rid of the supply chain problems very quickly, turn out to be misleading. Much lower trade elasticity is estimated for the COVID period compared to 2008 crisis, and that shows the importance of persistence of global supply chain bottlenecks. Euro area is more open to trade than the United States, which means that foreign shocks will play a bigger role in explaining the euro area inflation, two-thirds, between 2019 Q4, 2021 Q4. None of this means that central banks shouldn't react to inflation. In our model, as any other model of monetary policy, monetary policy can tame inflation in a given country by contracting aggregate demand. However, there will remain upward pressure on price growth as long as global supply chain bottlenecks persist, as Ms. Lagarde pointed out in her introductory speech. Thank you very much. Thank you very much, Stetnam. Thank you very much for your rich and informative presentation, and thank you very much for the policy implications. Gabriel, you have 15 minutes. Thank you very much. Let me first say how great it is to be here in Sintra at this very distinguished conference, and I'm grateful to discuss this paper. Stetnam, you did a brilliant job. It's very clear in the, I think, important piece of work that will inform policymaking, not just monetary policymaking, but broader trade policy, for example, as well. I would like to do three things here. First, of course, provide a short recap of what I will call the DKSY paper, some comments, and then give you a few stylist facts, some more facts, as Stetnam has already shown quite a few, and then very shamelessly use education and advertise some work of my own that has relation to what we're seeing here, namely, okay, we have these foreign shocks affecting New Yorkshire and other countries. What can be done is decoupling a way to insulate our economies from these supply chain problems. I think the key equation in the framework that we have just seen is in the first bullet point here in my list, that maps wage changes in sectors into the change in the consumer price index, and that mapping is done through the complex interrelationship of sectors, through the input-output linkages, and that means that different sectors, between services, durables, whatever, the wage changes that affect there are put together into an aggregate wage effect through the input-output connections. And what the authors then do is they use observed labor supply changes, that's the shocks that the pandemic have brought about with very different effects at the sectoral level. Again, use the input-output relationship between those sectors to come up with something like an aggregate supply shock, that's this lambda prime D log L, and they take the observed price change, the consumer price change in the economies and use that in the equation that you displayed here to pick up the aggregate demand shock. And that's the decomposition that is then put very simply used to produce the graphs that we have seen through the lens of a much more complicated model, the composition into supply chain or the supply, bottlenecks or supply shocks, and aggregate demand shocks. I think this is very useful and a very plausible way to come up to the result that Zeppelin has insisted on, namely that supply shocks are playing something like half of Corona time inflation in the Eurozone, and about 30% in the United States, and the missing part, of course, is then the role of aggregate demand shocks much bigger in the United States than in Europe. Of course, this is a stylist model, and that's a virtue, of course, because it makes it easy to comprehend and easy to communicate, but of course there are concerns that I find in my own work as well. One is that when we work with input-output data, that's the information that we have on the global scale that we need here, and data is still very, very coarse. It's much better than aggregate data, of course, but within sectors, take the chemical industry, we have, of course, supply chain relationships as well. There's a vertical structure in the chemical industry, there's a horizontal structure in the chemical industry, and within that industry, we can have very different supply shocks too. That, through the input-output linkages, are going through a multiplier effect coming out bigger through that structure than if we lump everything together within a sector. The other thing that I thought a little bit interesting reading the paper is that what is very important on my agenda, namely trade costs and what happened to trade policy, what happened to shipment costs, transportation costs, doesn't receive much attention in your paper. And so, at least I have a worries that we see a rise in protectionism. We've seen part of it in the corona crisis. We have seen a rise in export restrictions that we have not seen for a very long time in history. We see more of it now as we suffer through that energy and food price crisis and the role of the cost changes or policy changes at the links between countries and sectors. I think that is something that merits more attention. The other thing that is discussed in the public quite a bit now and is not emphasized in the paper we have seen is the role of imperfect competition. That when you have supply shocks occurring in a sector, then what we should also expect is that firms exit, smaller number of firms, less competition, less competition, meaning higher prices, leading to even more of a supply shock impact. So putting all that together, I think what we have seen is actually conservative estimates. So supply chain restrictions or supply side restrictions could even possibly explain more in a model that goes away from sectors into more granular data and also has indulgences markups. There's a lot more in the paper and I think 7MT did a brilliant job in explaining it. We have this shift towards trade-offs that was not expected and that led to a lot of supply chain issues. We have the fact that effective trade-alisticities are reduced to supply chain bottlenecks that means that the response to supply side shocks are bigger. We have that nominal wage rigidities that are very important in that model can lead to the fact that single sector negative supply shocks can create country-wide inflation. That's, the paper has a very nice short explanation for that that's easy to read for everyone. And then we have that fourth result that foreign shocks and global supply chain bottlenecks really mattered a lot in particular in the Eurozone in that Corona period more than in the United States. I think this is all plausible even if the model framework is of course stylized and it has a very clear monetary policy conclusion that Madame Lagarde has emphasized and Septim has told and think that's something we can now carry out at least from Sintra and say that we have a lot of evidence pointing toward that conclusion. And indeed it is something Septim pointed out quite nicely too. It is misleading to only look at good trade and from that think that bottlenecks don't matter or have not mattered in the crisis. What you see here is global trade, good trade that is. You see the in red, the Corona crisis. You see the V-shaped recovery of trade flows. You see that in the last data we're already 8% above the pre-crisis level. So trade is actually doing well and one could of course conclude from that what are you talking about? Bottle next in supply chains are not showing up in the data. In particular the difference to the layman crisis or GFC crisis is quite stark. Then this is the blue line situation was much slower recovery was much slower. But I think that's an important lesson from the paper. That would be a premature conclusion. We need to compare to the right counterfactual and without the supply chain disruptions actually trade, the global trade flows, the red line would have expanded even more. And that's relevant from a macroeconomic point of view because demand has shifted and it is that increased demand for durable goods that was not fully satisfied by that increase in trade. Now behind these good trade numbers of course there's industrial production and industrial production tracks the goods flows quite closely. Now let me say a couple of words on what to expect and what we see in the data right now. Are those bottle next going away or not? So there is right now I think mixed signals if you look at the ocean delivery times the latest data that's available shows that delivery times go down a little bit at least if we look at the big trade routes from China, West or Eastbound, Eastbound or Westbound. I have some own statistics that I like a lot if you look at the container ship industry worldwide we can observe container ships every 15 minutes actually see where they are, see whether they move or not. And the interesting fact is a lot of container ships are not moving. And I'm not talking about them sitting in harbors. They're not moving. Now that's container ships sitting at the Georgia South Carolina coast these are container ships idle around South California, Los Angeles, Long Beach. Then you have the North Sea. Look at the latest data that's from 21st of June so quite recent North Sea traffic jam building up and then of course we add China and we get a lot of action from there. And the disturbing feature of this data is that in the last months, in the last weeks I must say the traffic jams have again built up and it's not just China, it's also North Sea and that means we're probably not seeing that ocean delivery times movement that we have in the last weeks of that slide being continued unfortunately so that's another source of stress potentially for global supply chains. Now the global supply chain pressure index that we have already seen I think is a fantastic tool and maybe that's more for curiosity but if you compare again that index in the corona time that's the red line and in the Lehman episode you see that in the Lehman episode exactly the opposite happened at least in the first months of the shock. Global supply chain pressure actually went down after the GFC sets in. We had exactly the opposite and that shows that it was probably erroneous to compare the corona crisis to the Lehman crisis as some at least in the first months or so of the crisis did the global supply chain picture was very different. But let me look at the little bit more longer horizon. We have talked about two months in the paper we have seen and I've talked about the last weeks or months but what I find worrying is that something happened in the global economy that actually predates very much the corona crisis and even predates a little bit the onset of the GFC namely the end of a long running trend where global trade expanded faster than industrial production. And that is something that is still unexplained. Now of course we are looking here only at goods market globalization services play a large role we have difficulties mastering them properly but there is building evidence that the horizon protectionism explains this phenomenon that the economists have called the phenomenon of globalization and that is worrying because that could mean that the pressure that globalization put on goods prices is no longer there and that horizon protectionism adds to supply chain stress reduces competition reduces the capacity of the global system to respond to shocks. And I think the picture that is there since 2007 or so despite the ups and downs despite these interruptions of the crisis is a very important we should care about that and that's maybe also something where central banks should raise their voices and ask for restraint in trade policy concerns. And then let me wrap up with some shameless advertising as I said at the beginning and discuss very briefly a paper that's fresh by some quarters of mine where we try to understand whether global supply shocks the stuff that's discussed in September's paper affect our economies differently when they decouple their global value chains or not. So the question is how much would the potential of decoupling insulate the EC, the Eurozone or the US or other countries from the negative supply shocks in foreign economies. And that's a model that's very similar to DKS. Why we have also this global input-output relationship there is some break to sectoral labor relocation. We have different sets of trade elasticity as we can distinguish between the long run and the short run and so on. And we calibrate the model to a China lockdown shock in winter in winter spring 2020. So what's and we can't talk about inflation with some with some wriggling because welfare is essentially the inverse of the price index. So if the price index increases we say inflation welfare in this model goes down. And there's an important message that I want to convey here, namely decoupling would be very expensive. Now I'm showing you here the most drastic picture that they can show. That would be the coupling of global value chains even within the Eurozone. We're certainly not expecting that but that's just to show what the numbers would be. We're talking about welfare losses, what price increases that is in countries like Ireland, Luxembourg, the Baltics of about 50, 60, even close to 70%. So this would be dramatic. Of course in the paper we save the Eurozone and so on. So we can do a lot of different experiments and that would of course bring down the cost but it remains very, very expensive. And then if we stack the cards against us that's what reference have pushed us to do and ask what's the insulation effects of not being linked into global value chains. And I'm not stopping here final goods trade. It's just intermediates goods trade. Then foreign shocks would affect us less. That's obvious, it's not a big surprise but the numbers you play are very telling. The costs of decoupling on order of magnitude bigger than what the short run benefits of insulation would be. So that's a warning against those who hope by decoupling we would make ourselves less vulnerable from foreign supply volatility. Let me stop here. Thanks again for the opportunity to cast a great paper. Well thank you very much Gabriel. I think that both of you have paved the ground for the Q&A that we're going to have now. And afterwards, Gardo. We have three questions. Let's collect three questions and afterwards we'll continue. Excellent paper. So you emphasize supply shocks. I guess the other side of the coin is the demand effect which in the US I take to be very large. There's a large debate about the relative roles of perhaps excessive fiscal stimulus versus perhaps too stimulative monetary policy. Your model is very nice in that it speaks to sectoral effects and monetary and fiscal policy have very different sexual effects with monetary policy of course working more through interest sensitive sectors. Can you use your model to speak to that debate and sort of parse out how much the aggregate demand shock comes from different effects? The other question I wanted to ask is about interaction effects. So the effect of a given amount of demand on inflation is presumably going to be different depending on the state of supply. And so I wonder whether you could speak to these interaction effects and notably whether for example the monetary multiplier currently is different than it might have been historically given the state of the supply with effectively the supply curve having hopefully a more favorable slope now than it had historically. Thank you, Ricardo. So sorry, two nerdy measurement questions on your two exercises. On the first when you take out from GDP to measure aggregate demand, the formula that comes out of the model is really to use changes in sectoral hours and therefore different results between the euro area and the US are driven by differences there. But the way you build those hours has to do with using, especially for the US, sectoral employment changes. Now the US and the euro area had very different employment policies in terms of whether you had furloughs or just falls in employment or not. So could this not be biasing your results towards finding a much more effect in the US simply because there you had a lot of adjustment employment in the euro area and so you're gonna say, aha, therefore this goes into higher inflation in the euro area. On the second measurement exercise, somewhat related, again the measurement is really very much based on these changes in employment for the sectoral shocks relative to a potential. That's the Keynesian employment that you very well illustrated. But of course I think that you assume and clarify if not that any change in employment is this Keynesian because the potential hasn't changed. But again, we know or we suspect, actually we don't know, that they may have been very permanent changes in sectoral composition as a result of COVID. Insofar as those permanent compositions are there, you will again be overstating the sectoral contribution that you find to the Keynesian mechanism because you will be confusing the extent to which we're just shifting from some sectors to others permanently and you'll attribute it to this. If you'll instead pull all of that, any change you observe will say, no, no, this is gonna be an inflation on the sectoral shocks. So again, sorry for the nerdy questions, but whether these are overstating. There are enough, finally, games. I mean, it's a super interesting paper. Now, since the focus was on the inflationary impact, I feel it's important to remind ourselves that inflation is always an everywhere monetary phenomenon, the problem with inflation, the buck stops here, it stops with the central banks. I mean, I understand the point that if there are global supply chain pressures, this may have come as a surprise to central banking and maybe they were caught off guard, but it's clear that there still is a huge amount of liquidity sloshing around the system. There's a lot of money meeting too few goods and that's a very old story. If it's global supply chain pressure all by itself, we're talking about relative price movements and relative price movements by themselves don't cause aggregate inflation, they cause relative price movements. And so maybe that's a question back to you. If you see prices having to go up in certain sectors, relative to which other sectors, which were the sectors that could give? Was it wages? And maybe wages had to go down, maybe that explains why we now see the shortage of employment in many sectors. So there's an interesting flip side to what's going on here. And finally James. Thank you, Jim Bullard, St. Louis Fetso. I love this paper and I did think it was an excellent way to get at this supply shock issue and I love the granular data and so on, the input output. But I'm gonna echo I guess the previous comments which are that the ECB and that FED are inflation targets. We think inflation is up to the central bank in the medium term. So I think your statements should be that given the policy reaction, the monetary fiscal policy mix to the pandemic, we got a certain amount of inflation coming from the supply side of the economy. But there was some other response that could have occurred that would be a counterfactual that would have generated zero inflation over this era. And then we wouldn't be saying how much of it is due to supply shocks and how much of it is due to demand shocks. So I think it's taking the response to the pandemic as given, which I've in many other forms I've characterized as very, very good both in Europe and the US, but taking that as given, then the supply shock accounts for a certain amount of inflation. So maybe that's a way to reconcile what you have with some of the previous comments. Thank you very much. Step in first and afterwards Gabriel, you have two, three minutes, even a little bit more in your case. First, let me thank Gabriel for great comments and let me just clarify two points, all well taken. I fully agree that we don't have the trade costs, shipment costs, that's obviously a limitation, but that is all will go in the direction of adding up. So our estimates are lower bound. And that's exactly why we stop at the end of 2021, the new shock, war, new shock, energy, protectionism, fragmentation, all these things are going to add. So exactly in that line. So we fully agree. We do have monetary policy. There is a zero lower pound. Remember, this paper is about to get to reality as much as we can. So zero lower bound, normal interest rate in closed economy. When we go to the open economy, we do several things between exchange rates, inflation targeting, nominal GDP. So, and it's our results all robust to that. So it's not that we don't have monetary policy. We do have it, but there is a zero lower pound. And the imperfect competition comment, I fully agree, there are no markups in this model, but basically these things are a little bit long run, right? We are trying to capture this end of 2019, end of 2021, very short run, is this helpful to policymakers, real time policy making, and in a period where firm exit was stopped with the laws and regulations, right? So, but moving forward in the long run, of course, and you know, you can add markups to the model and we simplify the original model so you can do all those things in the long run. So the more long run issue. Okay, so, Annette's question. We didn't do this, but you can do all this with this model, monetary policy, fiscal policy, how the multiplier would be different. I wrote a Jackson Hole paper last year with Pierre-Livier-Gonchard, Veronika Sander, and Nick Sander, and Veronika Penchowaka. There we did fiscal policy using a similar framework and we showed that a very low fiscal multiplier in a world of supply chain bottlenecks. Here, you know, we didn't do that, you know, and that requires you have both policies and you really need this hand to mouth agents because that's what you really need for the stimulus, the entire stimulus to work its way. It isn't their original model, but Kaifahri is very rich, very general model. We kind of simplify it to the bare bones, try to get at this inflation, the composition, but it is definitely doable. And interaction effects, they are there, right? So this is exactly how that simple equation where we back out aggregate demand and then the full optimization works out where you have all the interaction effect and second order effects between aggregate demand and sectoral demand and supply and then the model gives you the composition. But going and pointing out within the context of multiplier, yeah, that's extra work. That's very interesting. We didn't do that, but it is doable if we go back to the richer model. Ricardo's question. So we use ours, we don't use employment, from BEA and from ECB. Now, you are of course right that the policies implemented during pandemic is very different across Europe and across, yes, in terms of people lose their jobs in the US here, there are the retention schemes and all that. So I fully agree with that comment. Again, this paper we did last year for Jacksonville, look at exact those policies and the implications of those policies. But you are also right, even within these hours data be used, there is the potential one and there is the equilibrium one. They are both in there, right? This is, I think it's like identification problem. You have one data and you have two things that you are trying to separate within that data and that's what the model structure helps us to do. Harold's comment and Jim Buller's comment, fully agree, monetary phenomenon, inflation ends at the central banks. Now, let me explain a little bit how we do this backing out of aggregate demand shock. In the model, it's a discount factor shock and all this liquidity put in the system, all the stimulus coming from all sorts of policies, this is a summary way of summarizing that. If you try to go and feed all that in, you have to do a much richer matching of these moments in data. A very simple way is this is a discount factor shock, aggregate demand capturing all those things, but in the model, there isn't this liquidity put in the system or the transfer of fiscal payments and all that, right? You can incorporate all that in. It's just that aggregate demand shock as a discount factor shock is a way of capturing that in a simple way. And again, we are not saying monetary policy shouldn't respond to this inflation. Of course, monetary policy should respond to this inflation, that was my last line. And I mean, it's just that inflation would have been lower. That's the way to read this result. Yes, there can be a world that could have been zero inflation. In fact, this is all again about the aggregate demand shock, exactly as you're saying. In the original Bakai-Fahrif work, it is deflationary actually. They do from January to May and everything works through deflationary because aggregate demand shock didn't work its way through yet. So we could have been in this world but we did a lot of stimulus. We are not denying that, but still without this sectoral supply component with that stimulus. Remember, we matched the observed inflation through this contract. The inflation would have been lower. That's what we are saying. Did I answer everybody? Yes, that's it. Yes, thank you so much. Ariel, you have two minutes. Yes, I don't have much to add actually. Maybe with the only exception that it would be great to use this type of framework to also look at optimal monetary policies. So what should Central Bank actually be doing with the face of nominal rigidities facing these shocks? Something has to give, as Ariel says. What is it that optimally should give? I think that would be an important and interesting research avenue. Let me say one thing on that because I just discussed this paper Friday on BIS annual conference by Jennifer Lowe. That does exactly that. Optimal monetary policy with a full macro network model and they develop a price index that Central Bank should target and the index give larger way to stick industries and lower way to flexible industry. So they say you should target services and not energy. That's the conclusion of their optimal monetary policy paper. Very good. So I think that this brings us to amend. Clare, back to you, our master of ceremonies. And the next panel is going to be important because it's going to be about inflation expectations that I think that in the dynamics of the inflation process are a crucial element.