 Good morning, everyone. Well, today we are going to talk first about real estate, real assets, the most important asset in the portfolios of the European households. And in order to do that, we are going to have, you know, our first speaker is going to be John Melbauer, Professor of Economics at Oxford University, and a Senior Research Fellow of Notfield College. You have been here since the inception, so you are becoming a tradition, John. So welcome again. So your paper is on real estate, the Transmission to Monetary Policy, also with consideration to respect to financial stability and implications for inflation. So in that respect, I think that this is quite relevant. And John's paper is going to be discussed by Giovanni Del Arriccia, Deputy Director of the Research Department of the IMF, that you have worked extensively on great booms and real estate booms in that respect. So without any further delay, John, the floor is yours. You know that you have 25 minutes, and afterwards Giovanni will have 15. Well, I'm very grateful to the organisers for inviting me to address this very distinguished audience. Let me apologise for the length of the paper, but it's not what you would call brief. But the brief given to me by the organisers was anything but brief. It was very comprehensive. I have to say the paper is built on foundations with lots of work by lots of other people. And let me just mention three wonderful co-authors, namely John Ducca from the Downless Fed, Valerie Chauvin from the Bank of France, and Junien Ahren with whom we worked on the South African financial stability mandate with a great overlap with the themes of today's paper. So here's a brief outline. Want to talk about the role of real estate in the financial accelerator as we saw it in the global financial crisis. Discuss institutional differences, why some countries had the financial crisis, other countries didn't. And my discussion has done very distinguished work on this very issue of institutional differences. I'll talk about the six channels by which housing transmits interest rates and lending standards to aggregate demand. Look at central bank policy models to see to what degree they actually address these channels. And then talk about some of the implications of the research that I've done for improving the risk scoreboard. Make some brief comments about how commercial real estate differs from residential and then conclude with the discussion of how monetary policy working through real estate has worked, the new macroprudential framework leaning against the wind and the risk outlook if there is time. So history matters. It matters not only to the analysts like us, but to the participants. And here I've selected the five largest eurozone economies plus Ireland. Ireland is very interesting, of course. And you can see the tremendous difference in the boom and bust cycle that occurred after monetary unification in Europe and of course in the run up to the global financial crisis. An immense difference. You can see that France is kind of in between it had a similar appreciation to Spain but not the collapse afterwards. And you can also see signs there of cycles in the 1980s. There was a certain boom and bust in some countries. Now the big boom in the peripheral economies was had much to do with monetary union and the fact that interest rates converged to very narrow gaps, mortgage interest rates converged to very narrow gaps with monetary union and that had a lot to do with setting off the boom and the peripheral economies. Together, of course, with the financial liberalization, the unification, the cross border financial activities that occurred. Now, there was a precursor to the global financial crisis in the UK and the three Scandinavian economies in the 1980s and that's really rather interesting and I wish we had learned more from that episode. You can see here that in Finland and in Sweden, Norway and the UK, there was a big boom and bust cycle with quite serious financial crises in three Scandinavian countries and in the UK, the Bank of England had to rescue the insurers of mortgages. The other interesting feature of this graph is go back to the early 1970s and look at the impact of the OPEC shock. So the oil price shock caused a fall in real house prices in Finland and the UK, both preceded by a big run up. So there's a bit of a warning there. So let me talk about the role of real estate and the financial accelerator. This picture comes from our JEL survey paper with John Dooker and obviously in the run up to the financial crisis, there were overly lax lending standards. So a shift in fragile fundamentals leading to overvalued real estate and then prices started falling that feeds through to construction, to the demand for real estate, into consumption which influences GDP of course and then on the financial sector in the economy, you had this big amplification mechanism which probably defaults and foreclosures affected the capital structure of the banking system led to contagion panic and therefore tighter credit standards for loans, higher risk spreads. And these tighter credit standards then feed back onto consumption, onto house prices, onto construction, into GDP and so on. So it's a vicious circle in which the amplification was really quite enormous in the US. In Spain and Ireland, very similar things happened except of course the financial system was far simpler. So let me talk a bit about the sources of heterogeneity. So thinking about the construction channel, the elasticity of supply matters quite a lot and countries differ in the degree to which the building activity responds to house prices. The share of public sector housing differs. As far as consumption is concerned, one big difference is access to home equity loans and the rates of owner occupation. Another big difference is in the degree to which households extrapolate previous gains, previous capital gains in housing in countries where home buyers are heavily geared, the incentive because of the capital gains and the reduction in user cost that comes from that, the incentive to despeculate is much higher when you're heavily geared. So you pay much more attention to past appreciation. And that of course varies with the degree of mortgage interest tax relief and whether high loans of value ratios are available and also whether property taxes are linked to current market values or not. So clearly regulation, financial market structure, math is a lot, lax regulation that permits high levels of gearing both to banks and mortgages, then you get much greater amplification. And having a fixed rate mortgage market slows the transmission of the policy mitigation that is available when the crisis breaks. In the UK for example, the ability of the Bank of England to bring down interest rates very rapidly really made a huge difference to the UK and the financial crisis. Maturity mismatch is another issue that matters a lot. Ireland being a particular case where maturity mismatch and funding mortgages mattered. Having countries where state or collective insurance schemes are widespread also makes a big difference to financial stability. On this slide, I mentioned some of the other differences which affect the transmission to the wider economy. And once again, financial regulation, financial structure, complexity, interconnectedness and all these things matter a great deal. So to illustrate some of the differences between European countries, here is a picture of how the mortgage debt to income ratio for the six economies I'm concerned with differs at the moment and has evolved over time. You can see the Netherlands is the champion. Netherlands of course has very big incentives for households to have mortgages because of tax relief. Italy is at the very bottom. You'd also see that relative to the kinds of fluctuations in the mortgage debt to income ratio that occurred in the run up and post financial crisis, the recent level of fluctuations is quite moderate. So that's encouraging. I mentioned the construction channel as an important element in the transmission to GDP. Well, here's a picture that shows how incredibly volatile residential investment can be as a share of GDP. Look at the Irish graph, the green graph. This incredible collapse from, you know, 10% or more of GDP to 2% of GDP. And so many in Spain, there was a very large fall in residential investment, also in Netherlands to a degree. This of course has very big implications for employment and the economy as a whole. So it was fairly evident from the picture of the financial, of how the accelerator feeds through to the economy that the housing channel is going to work in terms of policy from the transmission to the mortgage rate, transmission of policy to house prices to residential investment to consumer spending to mortgage debt and then potentially eventually to non-performing loans because non-performing loans give you an indication of the fragility of the banking system and non-performing loans rise. It's much more difficult for banks to extend credit. And in some countries, one can actually empirically estimate a seventh channel, which is the effect of house prices on income expectations absent in some of the European economies I've looked at. So I asked myself, well, how well do central bank models actually capture these six channels? And I'm afraid the answer is not very promising. In fact, there are only three of the six models I've looked at here that have reasonable linkages between the banking system and the real economy. So that's a major defect of many central bank models that simply haven't tried to address the linkages between the financial system and the real economy. And at a more sort of nitty gritty level, to what extent do they include lending standards? Because lending standards are really important. Poor lending standards before the crisis actually helped to lead to the crisis and the tightening of lending standards in the crisis fed back on the real economy. So including lending standards as a control in all these channels is critical. And very few models actually do that. Something else I should mention from this graph, something from this table is that nominal interest rates actually matter quite a lot. And very few of the models include a role for nominal interest rates. None of the models actually try to model MPLs themselves. So I think there's quite a contrast between the sophisticated understanding that's there in the financial stability sections of banks and the lack of such understanding that seems to be there in the modeling of the macro used for monetary policy setting. So for France, I was able to build on work that Valerie Chauvin and I did and do some further work to explore how in the case of France, these six channels actually work. And we showed that in every one of the six channels there is a significant, often highly significant crucial, in fact, role for the control of lending standards. So the house prices, consumer expenditure, mortgage debt, it's critical that you control, not just for interest rates, real and nominal, but also that you control for lending standards. So the approach that we use is the latent variable approach where we pick up more lending standards as the common factor in equations for house prices, mortgages and consumption after taking account of a full set of economic demographic factors. This is a bit like the Sherlock Holmes approach. You have a crime, you have a list of suspects, you eliminate all the other suspects and the one that remains is the guilty one and the guilty one is lending standards. Just one other feature of this under four, consumer expenditure, none of the central bank models actually control for the house price-to-income ratio, which is an important aspect of the story because if there is a housing collateral effect or a housing wealth effect, there's also a negative effect that concerns the non-owners or the renters who have to save harder for a deposit when house prices become less affordable. And it's important to control for that. Moreover, it's important to desegregate household assets into liquid and illiquid and debt. It turns out that debt is far more negative in its influence on consumption than the central bank models suggest. Here's what our latent variables look like for France. Financial liberalization in the 1980s, both for consumer credit and for mortgage credit, a bit of a collapse as MPRs rose in the mid-90s and then renewed liberalization in the 2000s followed by a moderate contraction in France after the global financial crisis. So I was really thrilled to discover that our latent variable had remarkable properties in its ability to forecast future non-performing loans. We didn't know that in advance of doing this work, but it turns out the T ratios are like 15 or something like that. So we have a model for non-performing loans in which recent short interest rates, the unemployment rates, the house price to income ratio, income growth, and average of lending standards going back one to five years previously. So history matters. It isn't just what happened in the last year that matters, but it's the lack of lending standards over a long period that matter. So what we would call a perfect storm would be years of lack of lending conditions followed by a rise in interest rates and unemployment, fall in house prices and a fall in income growth. That would be a perfect storm. Fortunately, at the moment, we haven't had years of lack of lending conditions. Just to illustrate how close the relationship is between the MPL ratio and our estimated latent variable, this is what it looks like since 1990. So just to sort of summarize what the implications are for the credit cycle, what we found is there's a strong connection between lending standards and real estate prices, credit growth, consumer expenditure, residential investment, and also eventually of MPLs. So in the short run, lucid lending standards promote growth, but in the long run, they have an impact potentially on MPLs that could be damaging. Interest rates themselves, of course, have a similar role in all these equations. Every single variable, real estate prices, credit growth, consumer expenditure, and so on, is affected by interest rates as well. So it's really important that we understand the joint influence, both of monetary policy and of macro-prudential policy and prudential regulation in how the credit cycle and the economic cycle operate. I would say that there's this contrast between the two wings at central banks, the monetary policy wing and the financial stability wing. So let me comment a bit on how our research on France potentially could improve the ESRB risk scoreboard. Now that risk scoreboard captures what the ESRB calls collateral stretch, funding stretch and household stretch, and there are 10 detailed indicators in there, which include things like the house price to income ratio and various other things. There's also information from the bank supervisors on loan-to-value and loan-to-income ratios, at least from one recent survey across the Eurozone. And the question is, well, how do you summarize all this complex information, some of it contradictory, with, you know, to try to get at what the underlying lending standards are, were they too lax or were they justified by fundamentals? And the French evidence suggests that the latent variable approach is highly successful in extracting an indicator of more lending standards, given its amazing performance in forecasting NPLs. So my proposal is that a stripped down two equation version of a more complex French model could be rolled out across the Eurozone with panel methods compensating for gaps in historical data. So I see the three benefits of the latent variable approach in this context that you get a regularly updated latent variable estimate of mortgage lending standards, which would add a valuable new indicator to the scoreboard. You get an improved specification of the house price equation that would give a more accurate indication of potential overvaluation. And finally, as we all know, it's very hard to tell whether the macro potential policy is actually working because of endogeneity. The policy comes on when credit growth, house price growth is high, but that could be driven by fundamentals. We're not sure whether it's fundamentals or problems. So these endogeneity issues make it hard to measure the success of macro potential policy. Well, the latent variable approach is designed to address that endogeneity and so should make it easier to assess the impact of macro potential policy on lending standards. Right, I said I'd make some remarks about commercial real estate. Thanks to the ECB, I've had access to a confidential database, commercial database, and that shows some information on a comprehensive measure of the price of commercial real estate, again in real terms. You can see that Ireland had by far the biggest boom and bust. The German graph is much like the real house price graph. The other countries show a certain boom and bust, but more moderate than for house prices. So there's some differences between commercial real estate and residential, which the ESRP has discussed. The larger role of institutional investors, great exposure of international capital markets, that matters. But in practice, commercial real estate is quite correlated with residential. In fact, residential rental is part of what's called commercial, though common shocks to credit-conditioned interest rates and international investors have become very prominent in top cities, residential. There are some big data issues, which the ESRP is onto. Defining even the concept of commercial real estate is difficult. And the days of problems are a lot worse than for residential. More complex opaque markets and comparability problems. So let me make some concluding remarks. First of all, about monetary policy and the housing channel. Now, monetary easing was intended to increase aggregate demand from households. But first of all, the impact has varied a great deal across countries. You're much weaker in Germany than in other countries in Spain or Ireland or Netherlands. The impact has also tended to be overstated, in my view. And that's partly because the negative longer-term impact has not been sufficiently taken account of. So lower interest rates obviously raise debt-to-income ratios, but higher debt levels have a remarkably negative effect on consumer spending. And then there's the negative affordability effect of higher half-prices on non-owners and renters. Even without the credit cycle and the possibility of financial crisis that could follow from overvalued real estate. And then there's some seriously negative side effects on inequality between the old and the young, within younger cohorts and so on. And then as this prize-winning paper from Bill and Werner points out, credit-of-fuel real estate booms have tended to crowd out more productive investment with negative consequences for sustainable growth. So that's a very serious charge. So let me comment on the new macroprinential framework. I think the work done at the ESRB, the EBA, the ECB, the country central banks, and the other regulators backed by the BIS and the IMF is really a remarkable progress in the last 12 years. I mean, it's a really fantastic progress. Obviously institutional heterogeneity is going to make it a functional necessity in the Euro area for supranational bodies and the national regulators to coordinate and each make decisions and try to coordinate. There's been a welcome push from the supranational bodies for much better data monitoring for new instruments, very much approved of that, and minimum standards on the legal perimeter for the borrow-based macroprinential measures. So I think the system is actually working reasonably well. I think the macroprinential tightening that's happened in the last few years is now validated by the current economic crisis, which otherwise might not have been far from the perfect storm that I was talking about. So just my final slide is we make some brief remarks about the leaning against the wind controversy. Since Lars Svensson is in the audience and you've all followed his debates with the BIS on this issue, you might be interested in some of my thoughts. Well, it seems to me that in the Eurozone because of heterogeneity, it's even more difficult to follow a leaning against the wind policy. So we need to think about leaning against the wind more in an international macro context. And in my view, it's very unfortunate that fiscal policies post-crisis went more expansionary because the central banks had to step into the gap. And as a result, these high real estate valuations were the negative consequences by-product. I think more generally, failures by government to correct distortions in housing markets and to promote stabilizing measures like tax relief, stabilizing tax relief, making property taxes more linked to market values, those failures have made the job much harder. So finally, on the risk outlook, the one comment, well, the two comments I'd make, one is that for investors real estate looks like an inflation hedge. And that seems to me a worry when inflation expectations are rising in the way that they have been. And that's just to be the controlling leverage on commercial real estate and by-to-let residential is pretty important at the moment. I think that's an area where macro potential policy needs to act. And then finally, I think on climate change, Europe is particularly exposed, European real estate is particularly exposed because it's such a high fraction of CO2 emissions in the euro area come from housing and other building-related activities. So holistic policy, the network for greening the natural system is doing a fantastic work on this. So I very much encourage the ECB to follow that route. But we need holistic policy cooperation between governments and central banks in this whole area. Thank you very much. Thank you very much, John, for this poly-edric comprehensive and insightful presentation. Giovanni, it's your turn now. You have 15 minutes. Okay. So let me start by thanking the ECB and the organizers for the invitation to my former boss, for the invitation to be in this conference and discuss this very interesting paper. And let me, before I forget, say that these are my views and not necessarily those of the fund. They will deny any knowledge of my existence if you attribute the views to them. So let me start with something very obvious. This is a great paper. I highly recommend everybody that wants to read the compendium of everything we know and especially everything we do know of the relationship between real estate dynamics and monetary policy to read the paper. It's long, but it's worth it. I mean, I was forced to. For you, it's more of a voluntary effort, but it's really worth it. And I think John did a fantastic job in putting together an enormous amount of information. So in this slide, I'm trying to summarize the many messages. And I hope I distilled the most important ones. So first, the relationship is far from linear. There is a lot of complexity. There are a lot of local institution, culture, mores, the matter in how interest rates transmit to and through real estate prices and mortgage markets. Second, I think our understanding is clearly improved since the global financial crisis, but we are far from where we need to be. And I think there are two aspects of this. One is the calibration of monetary policy models for so-called tranquil times. Although I don't know, we haven't had many tranquil times recently, but for tranquil times and small shocks. And the other one, I think is more of the elephant in the room and is the prudential role that monetary policy may have to play in the context of boom and bust cycles. And I'll try to talk about that a little bit in this presentation. I cannot do justice in 15 minutes to everything that John did. So I'll focus on three of his six channels. And I think these are the three channels that are more important from the prudential standpoint, which I think is the more controversial part of the policy aspect of this. And I'm gonna use several charts that come from work often called or will look 11, he was in the room and will be happy to take any questions if you have them. And you'll see that they are mainly meant to reinforce John's points. I agree with basically everything he said, which make discussing the paper a little bit hard, but luckily the organizers told me, oh, you can go up there and say whatever you want. So here it goes. In summary, we know from experience that house price boom bust cycles have often been associated with financial crisis, but there is plenty of evidence that the real problem is not house prices. The real problem is household debt and leverage that comes with that. And in particular leverage in the financial intermediaries to fund the mortgage markets and a decline in lending standards that is often associated with these booms. There is also, and this is more recent evidence, a special role for the construction sector that John emphasized at some point in his presentation. So I'll talk a little bit about that too. So going back to the global financial crisis, we are all familiar with this data from the United States at the state level. On the vertical axis, there is the increase in mortgage delinquencies at the start of the crisis. Between 2007 and 2009, and on the horizontal axis, there is a house price appreciation in the six years leading to the crisis. And the size of the Christmas balls is the change in mortgage debt during the same period. And you see there is an obvious positive relationship between both the increase in credit and the increase in house prices with then the delinquencies in the states that were at the core of the subprime crisis like Nevada, Arizona, California. Now, when you regress this more rigorously using econometrics in many, many cases, house prices turn out not to be significant once you control for credit. So what really matters is that people get indebted. And if we were all buying our houses with equity, like we do say the stock market, the problem would be probably much smaller. I'm not saying the problem would go away, it is still a big chunk of the household's portfolio, but really what generates a crisis is when debtors cannot pay debt. And this is testament to that. Now, the second child I want to show is also from work with Luke and is a relationship between the monetary policy rate, this is the real federal fund rate and lending standard. So on the left-hand side, I'm plotting that against the capital asset ratio, the banking system in the United States, and you see that when the federal fund rate is low, banks tend to be more levered. And on the right-hand side is an internal rating that banks use in the United States to evaluate the riskiness of loans. And you see the loans are more risky when the federal fund rate is low. So you have two aspects of risk-taking that increase when monetary policy is easy. One is leverage, you had to squint a little bit to see the correlation, but it's there and more clearly the riskiness of loans. And finally, you get into the special role of the construction sector. After all the work we did on credit booms, we were saying, well, not all credit booms end up badly, but I think the average is about two thirds of credit booms end up badly, meaning in a financial crisis, but there is a third that soft land and John had the example of France where we didn't have a crisis after the boom. And so we started digging into these episodes to see if we could tell in real time which ones would be good and which ones would be better, at least have some signals, some cannery in the mine. And the construction sector turns out to be a fantastic cannery in the mine. So this chart plots the differential performance of value other than employment by industry in bad versus good booms. And you see that in bad booms, the construction sector is really booming. And one economic explanation for this, it may be that if house prices go up but nothing really happens, say New York City where you can have massive appreciations, but there is nowhere to build, then you don't have this a diversion of resources from the rest of the economy. I mean, you have this balance sheet vulnerabilities accumulating, but from a real point of view, the rest of the industry probably chugs along. But when you are in a situation like, say Arizona where together with the real estate boom and mortgage boom, you also have a construction boom, then things are possibly worse. So going to the point that John made at local institutions matter, I just wanted to list a few of the important parts. Of course, the level of mortgage debt in the country and the leverage ratios that are allowed by the banking system are critically important in terms of vulnerability, but there are also more nitty gritty elements of contractual details that typically as macro people we don't look at like fixed versus variable rates or portability of mortgages. This is something people really never focus on the United States is very hard to transfer your mortgage from one property to the other. So if you have a fixed rate and the Fed titans like it has happened now, nobody wants to sell their houses because if you sell the house, you essentially have a capital loss on your mortgage because now you have to pay a much higher rate. In countries like Canada where you can transfer one mortgage, the mortgage from one property to the other this doesn't happen. So these details are really critical in the way monetary policy interact with the market and we often forget them. The way the banking system fund itself to finance mortgage is also critical. You remember the role of securitization and during the GFC. And finally, structural elements like lens scarcity and building regulation are critical. So let me now switch to where we are now and for those of us like me that are paid to be paranoid, this picture is not reassuring. So this is a median across about 40 countries. You see building permits ballooned after the end of lockdowns. Real estate prices went up very rapidly, vastly outpacing rent dynamics. So this looks a little bit like the picture with the Christmas balls, but there are some reassuring elements. So one is that so the mortgage rate matter. So there was faster appreciations in countries with low interest rates than in countries with higher rates. And these relationships seem to have strengthened over time. So monetary policy seems to have some bite, not enormous, but is there. But more importantly, the aggregate picture hides enormous heterogeneity. So on the right hand side here, for obvious reason I didn't want to put numbers, but it's a color coded table of heterogeneous effects of monetary policy on real estate markets based on the various characteristics that John listed. So there is the history dependence, what meaning how much house prices had increased in the past two years. There is the level of household debt. There is the share of households that have a mortgage and there is a share of mortgages that have variable rates. And you see that even within the Eurozone, at one extreme you have the Netherlands with a lot of sensitivity to interest rate changes. At the other, you have Italy at the very bottom of this scale. One piece of good news, I think this is consistent with what John showed, is that the twin booms in credit and house prices that we witness in the GFC are much less prevalent now. One chart here is the correlation between increase in real house prices and household debt growth at the country level, and you see it was much higher, you know, seven than it is today. On the other charts I'm showing household debt service, which at least in this monetary policy cycle is starting from much lower level, especially in the US, in the Eurozone, a little bit less, but it's still there. And more importantly, and this is data for the US, I couldn't collect that, the same for Europe, unfortunately, this time the boom in mortgages preceding the monetary tightening was primarily driven by the so-called good borrowers. So high quality borrowers with high credit scores had a lion's share this time, while they were a very small share before the global financial crisis. And this is probably what is contributing to keeping the linkances down, even if mortgage rates in the US have increased dramatically. But again, I'm paid to be paranoid, so I'm looking now at what policy could or should do, and this is a chart from the paper that John mentioned we had at the fund about five years ago, to which Lars contributed, in which we examine the cost and benefits of leaning against the wind, and essentially, in order to get the net benefit from leaning, we had to calibrate the models with all the most favorable parameters so that leaning would be a good thing to do, which meant that we were saying never lean, but we were saying, well, maybe there are better instruments than monetary policy to control risk in the real, associated with mortgages and real estate prices. And that's essentially its macro-prudential policy, but I think at that time we had very little data to see whether macro-proof work or not. Now we have much more. This is a chart from a working paper that some of our staff put together, and they look at, I think, about 6,000 coefficients on macro-prudential policy out of several, a few hundred papers, and they look at the effects of macro-proof, bought on household credit and real estate prices. To make the story short, about 50% of the coefficients have the right sign and they are significant. This sounds not particularly reassuring, but you have to consider that there is an attenuation bias in these estimates because macro-prudential policies typically tighten exactly when credit is growing fast and real estate prices are going fast, and so 50% is not that bad, and in studies that use micro-data and can control for the endogeneity, the coefficients tend to be larger and more significant. So the rest of the evidence on macro-proof is a little bit more sobering. So first there is evidence of leakages and in particular cross-border leakages, and if you go back a few years to the stories of the Balkans where they would impose macro-prud in one country and there would be Austrian banks lending across the border, and this calls essentially for more cross-border cooperation and international coordination among macro-prudential agencies. I think there should never be an IMF presentation without this sentence. So it's there, and there are also leakages over time, and this is very similar to what we know about capital flow measures or capital controls if you wanted the non-politically correct way of saying, that over time the market learns to go around these tools, and so is a game of cat and mouse between the regulatory agency of the central bank and speculators. And finally, there is evidence that these tools are more effective when they are aligned with the monetary policy stance that when you try to have the cake and eat it too, where you are tightening monetary policy and losing macro-prud or vice versa, and the effects are less. So the joke I always say is that macro-prudential policy is not a third leg of macro-policy, it's one third of a leg, but it does help. So let me leave you the takeaways, I think are pretty obvious. Let me leave you with one final chart. So I think this is an interesting question for future research. Is there a structural effect in real estate from the pandemic? So if you look at the right-hand side, these are color-coded real estate price changes in downtown New York and San Francisco, and these are cities that were booming enormously before the pandemic. And then when the rest of the country, real estate prices were going up, these prices were crashing. And on the left-hand side, there is a chart from a paper by Gupta et al, and let me move it because we updated it. And here basically you regress house prices and house rents on the distance from the city center. And before the pandemic, the coefficient was very stable and clearly negative. So further away from the city, prices were lower, rents were lower. And then when the pandemic came, the coefficient started converging to zero because people didn't have to commute. And with remote working, why do you want to live in downtown Manhattan when you can have a much bigger house in Long Island or something like that? Those of us that have property in downtown, they see very well the pain that this has caused. And the interesting thing is that at least for rents, the trend seems to have reverted and things are going back to pre-pandemic levels, even if the coefficient is not as large in absolute level. For prices, we haven't seen the reversion to the mean, maybe because prices are sluggish, but I think it's a big question because this is gonna affect dramatically the interaction between monetary policy and real estate in the future because if this is going to stay, we are gonna have a massive restructuring within the sector. And this goes to commercial real estate as well where downtowns are empty and there are commercial buildings that will have to be reconverted. And I think it's a great question for future research and I'll stop here. Thank you. Well, thank you very much Giovanni. Now the floor is open. I see our first hand raised. You have the floor and then afterwards. Perhaps we can collect two or three questions before you respond. Go ahead, please. Okay, just a very quick comment on leaning against wind. Not about my own work, but about very fine work that has been done at the ECB by Tore Cockerel and Christopher Cock. And actually, I think that started under the leadership of Victor Constancio. The first time I learned about this was in a speech that Victor gave. But anyhow, this work has now come out first as an ECB working paper and just recently as a journal paper in the International Journal of Central Banking. And it's really taking the law discussion to the eurozone and modifying, calibrating the numbers to the euro area. And it's very well done. Also, it has actually made an impact on the ECB strategic monetary policy strategy review. And it's widely cited in the paper about financial stability in monetary policy. It's really worth reading. And if there is a revision of this paper, I think John should actually include this special work on the ECB. Of course, you know the result. Costs of leaning are substantially higher than the benefits. That's very robust for a lot of different cases. Thank you. Thank you very much, Ellen. Another? Thanks a lot. It was a very interesting paper, both from an academic point of view and also a member of the French Macropodential Authority. So thank you very much for the contribution. I have a question regarding your interpretation of your Latin variables. I wonder, how did you investigate a little bit more the determinant of the correlates of your Latin variable reflecting the tightness of coded conditions with monetary policy, with regulation, and maybe a little bit less, obviously, with the structure of incentives of the financial intermediaries and the competition across intermediaries. There are some interesting stylized facts regarding heterogeneity and restaking across banks during boom times. And in particular, the fact that the leverage distribution becomes much more right skewed, so a lot more assets become concentrated in large balance sheets when the cost of funds goes down. So I wonder whether you have investigated this and whether you could say a little bit more about this because it's a key variable. Thank you, Alain. There is another physical hand raised. Thank you. So I have a question for both Giovanni and John. Both of you sounded like you were supportive of macro-prudential measures to address some of these risks, and I was hoping you could talk a little bit more about which ones. We now have a whole set to choose from, and I'll give you a little more specifics on that. For example, LTV ratios are some of the most popular ones implemented, but highly cyclical because they're based on values by definition. DTI ratios, DSTIs, less well-used, less politically popular because they're seen as being unfair to people at the lower end of the income distribution, but probably less cyclical, so it could make more sense. If you're worried about the leakages that Giovanni mentioned, then should you not even just focus on housing narrowly and do something broader, like CCYBs? Or if you talked about the risk of commercial real estate, are there any specific ones we should be using more in commercial real estate which don't seem to be used as widely? So that's just a start, but are there others I'm missing? Just hoping for people involved in this area, where should we be focusing in this broad category of macro-prudential regulations? Okay, thank you very much. I think that we do not have questions coming from online participants, but I will take advantage of my position as moderator, and I will ask you something. One of the recent effects of the fallout of the war is that construction materials are actively on the rise. And even today, I have seen that a German entrepreneurial organization has complained about shortages. Do you think that this might have any sort of impact? Do you think that this is going to fade away? So, John. Thank you. Yes, let me start with the last point. Undoubtedly, shocks like this, supply shocks, do have an impact and will have an impact on construction volumes. And that is part of the risk story in terms of the impact on GDP. Let me turn to Helen's excellent question about the underlying structure, the risk-taking and so on. So, what the latent variable does is it looks at the outcomes. It looks at the outcomes for aggregate house prices. I've actually also estimated a model that includes the Paris house price. So, I can explain Paris as well as the rest of France. And there are some interesting differences because Paris is much more sensitive to both lending standards, to interest rates, and also to international factors than France. But nevertheless, the latent variable is capturing an overall effect on mortgage growth and on house prices of all the micro stuff that's going on underneath, which is very complex and very hard to try to understand. So, the controls are really crucial here. So, I mentioned the fact that the nominal interest rate in France is very, very important. And it has a lot to do with the way in which the French regulators regulate mortgages or housing loans in general. Namely, the debt service ratio is crucial in France. It's been recently, the recommendation is to reduce it to 33%, which I think is an excellent idea. The point is that when interest rates fall to very low levels, the debt service ratio drops, and then the banks extend a lot more credit if that's all they're looking at. And I think that has a lot to do with the rise in French house prices. In fact, my model, a fall in the nominal interest rate from 3% to 2% causes 17% long-run appreciation of house prices just by itself without taking into effect all the other interest rate effects, namely the user cost effect and so on. So, that's part of the story. And controlling for that is very important to correctly identify what the underlying trends in lending standards are. Now, Kristin asks a classic question about the instruments, in particular the borrow-based instruments like debt-income ratios, loan-sur-valued DSTIs. Well, what I just mentioned on France, I think helps to illustrate that point. I think, in some ways, there is a mistake for the French to concentrate so much on the debt service ratio because it encouraged extension of credit when interest rates were low. I think it would have been better if they had used multiple, a broader spectrum of measures for regulation. I mean, generally speaking, debt service ratios, debt LTVs, there are all kinds of measurement problems. Different valuation standards in different countries. The data are very absent in many cases because not only do you need to know what kind of lending standards were banks using before a measure was introduced in order to see what the impact is and often the data on that are missing. So, it's quite hard to measure these things well. But generally speaking, as my discussion pointed out, the evidence from the international data are that they do work subject to the endogeneity biases. But, in my view, a combination of measures helps to reduce the leakages and the ESRB actually pays a lot of attention to the non-banking sector in its assessment of the risks. That's really important. Giovanni? I'm not going to speak about France. No, nothing about that. On the macro proof, I think, if you look at the evidence, there is, as John said, there is a lot of country specificity. There is evidence that what the so-called forward side measures like LTV, DTIs have a direct impact while when you operate on intermediaries, they can readjust their balance sheet. So, you can increase a risk-weight on mortgages, but that's unclear that it's going to stop boom. Now, that said, I think that the political economy aspect that you mentioned is critical. I remember Stem Fischer many years ago when he was governor in Israel was complaining that they introduced macro proof and they were killed in the press by saying, oh, you are hitting the young couples trying to start a family. Because what an LTV, a student LTV is cyclical, but you need to have the money down. Typically, you're going to hit the poor and the young. And on top of that, as for any other prudential measure, when the crisis doesn't happen, because you impose the measure, and then you get killed in the press again, because they say, oh, see, you were paranoid. There was no reason to do this. So, I think that there is a very difficult political economy argument. For whatever reason, the public doesn't focus as much on interest rates the same way. When you raise the interest rate, you also hit the poor and the young, but it's not as direct. And I think it's a little bit easier for people to accept because it's more uniform. The final point I wanted to make is, I think we have a lot to learn from what we know of cross-border capital flow measures. Really, I tease my financial sector colleagues at the fund. They get very irritated when I call macroprudential policy domestic capital controls. But really, they are domestic capital controls. You are trying to put a wedge in an arbitrage condition. So, in a uniform interest rate environment, you are saying, this particular type of investment, I want to make it more expensive. And the market is going to see a profitable opportunity there, and it's going to try to go around. And so, the answer is probably you had to use them all. And maybe you start with some and you see that they get a road, and then you had to add a few others. And this is exactly what happens with capital controls. So, the implication of this is that you should use them for very short periods of time. And you should be able to impose them and remove them enough that the market doesn't create a cottage industry or lawyers and accountants that design contracts around this. And I think in a jurisdiction like that of the ECB, where you have multiple countries and multiple legal systems, it's even more important because, you know, you start a company in Luxembourg that has the only purpose of owning your house. And good luck trying to regulate that. And that would be obviously a way to go around. And I'm not a lawyer, so I come up with that in five seconds. I'm sure they can do better. Thank you very much, Giovanni. You know, only one thing. I think that it's much more politically palatable to impose capital-based macro-pollination measures than border-based. Okay, thank you very much to both of you. Now, I think that we can bring, you know, this session to an end and go to the second paper. Thank you very much, Giovanni.