 Welcome back everybody to the last session of the day in fact, unfortunately of the whole conference today, we have to a great presenters. George Degali and Yuri Gorotnichenko, who need no introduction, will take us to the, I would say the core intersection between monetary policy making and three factors that shape the environments in which monetary policy needs to be designed and executed financial shocks, fiscal policy and adjustments in inflation expectations. This is very, very topical, very, very important papers. George Degali will go first with a really intriguing paper and I'm saying that because, well, the title is Monetary Policy and Endogenous Financial Crisis, I'm saying it's very intriguing, very fascinating fact, because it seems to challenge what many of us and many people in the world view as the received truth on monetary policy making, that is that monetary policy would do best concentrating on its macroeconomic objectives and in setting policy. And rather than say setting policy with a view to dampening boom and busts in financial markets because that might cause inefficient macroeconomic fluctuation. So the paper as an opposite results, at least in my reading and it will be super interesting to understand what makes the difference in results. Then Yuri Gorotnichenko will take over in 45 minutes more or less, drawing on much of his very influential work on a number of things that are very close to central banks, particularly has drawn attention on the inattention of the public to whatever central banks have to do and to speak. But he will also speak about fiscal policy, fiscal policy influences the public's behavior and on inflation expectations, how inflation expectations shape firms behavior in particular as price centers and employers. So very important questions and that will be, there will be a lot of interesting answers that these two leading economists are going to offer. Jordi, I'm going to give you the floor in 30 minutes, you have 35, sorry 30 seconds, you have 35 minutes for your exposition then 10 minutes we reserve for questions, please put your questions in the chat function so that I can wrap them and convey them to the to the presenters. Jordi, please start. Okay well thank you very much for the invitation and also for your introduction. Yes, the title of this paper is monetary policy and endogenous financial crisis, and this is joint work with Fredric Wassey of Risco art and Christina mania, and for the Rican Christina work, respectively at the BIS and the goodness bank so the usual disclaimer applies. And let me say that this is very much work in progress or any comments and suggestions would be extremely welcome. Now, the, that this paper is motivated by a question that it's certainly many central bankers minds, which is what should be the role of financial stability considerations in the design of monetary policy. Now the conventional view is that central banks should focus on macro stability that should focus on stabilizing the output some measure of the output gap and inflation. But there is an alternative view, especially since the great financial crisis which holds that central banks should also care about preempting trying to preempt financial crisis and of course limit their damage exposed. Now the problem when it comes to trying to address this question is that the standard model of monetary policy analysis ignores financial factors so it has little to say about financial crisis. Now, of course, we know there are plenty of extensions by now of that model that incorporate financial frictions in the picture. However, in many of those extensions, most of those extensions crisis are triggered by exogenous financial shocks and or the presence of financial frictions just amplifies the effects of non financial shocks. Okay, of course, that's, you know, that makes it possible to do some interesting analysis, however, those models offer no room for monetary policy to preempt financial crisis and hence cannot even answer to the question that I raised at the beginning. Okay, so what we need is a model that has an endogenous overtake that at least allows for the possibility of endogenous financial crisis. And this is what we try to offer in this paper. So we develop a simple model. It's a version of the New Keynesian model with particular type of financial frictions that may allows for the possibility of endogenous financial crisis and in which monetary policy can influence the probability of such crisis. And there's an interesting trade off that emerges in the model between the short run macroeconomic stability on the one hand and medium run financial stability on the other hand. Okay, so let me here I have the main findings but before I go over the main findings I will just list the key ingredients of the model. So the model has nominal rigidities, like a conventional New Keynesian model, and those nominal rigidities imply non neutrality of monetary policy. So with the basic New Keynesian model we allow for endogenous capital accumulation. This is key in shaping the channels which financial crisis emerges. So you can think of the model as being part of this new wave of monetary models with heterogeneous agents in the sense that firms rather than households will be subject to deosyncratic shocks, deosyncratic productivity shocks in particular. And the existence of those shocks makes it desirable in principle for financial markets to play a role in relocating capital across firms. Okay. But those financial markets will be subject to two frictions, a symmetric information and imperfect enforcement of loan contracts, a perfect limited commitment by the part of boards. And as we will see this will give rise to the possibility of an endogenous collapse of financial markets. Okay, so here let me go over the main findings of the paper. Okay, so that you have them in mind while I described some of the ingredients of the model. So the approximate cause of a financial crisis in our model is the fact that returns on investment may become too low as a result of a capital overhang after a protected boom. Okay, so capital may be accumulated excessively. And that the lower returns on investment implied by that will raise borrowers incentives to channel financial resources to not productive activities and to default. Okay, the anticipation that that this may happen will tighten but effectively is an incentive compatibility constraint, which takes the form of a maximum leverage ratio, and these may lead to the collapse of loan markets, and hence to a misallocation of capital in the economy and a decline in aggregate productivity which will will lower output considerably. Now in this environment deviations from price stability may be desirable. In the run up to a potential crisis. Okay, the reason is that by tightening monetary policy, the central bank may tame. Booms that may bring about excessive capital accumulation. Okay, at the level of capital that cannot that may become non sustainable. If, if fundamentals. You know turn direction. Now, a policy that is rule based, as opposed to discretionary, and that stresses output stability can help avert those crisis can at least it can reduce the probability or the incidence of those crisis. And this is in contrast as I said with exposed discretionary interventions which may instead enhance instability. So, let me not say much about the literature because I have limited time. Let me just say that this is part of a few papers in the literature that developed micro founded models of indigenous financial practices. But the difference between our paper and those papers and you see some of those listed here is lies in the particular channel. And I'm not sure which or the mechanism through which the, the, the financial crisis emerges and the implications for monetary policy. Okay. So let me just point out the sum of the differential ingredients relative to a basic new kinds of model that I'm sure most of the audience is familiar with. Well, as in the basic model, we have an infinitely lift representative consumer that maximizes preferences given by this utility function, which is completely standard sees a bundle of consumption goods and these hours of work and subject to a sequence of budget constraints which here that in addition to consumption expenditures and investments in the nominally riskless bond. Households can purchase shares in intermediate goods firms at the real price QT and ST is the quantity of shares and JD notes is an index for the different intermediate goods firms. That investment in the following period will yield some dividends. Okay. And so the return on investment is given by this. Okay, these firms these intermediate goods firms that households can. Who's equity households can you invest are live for one period. Okay, so there are no capital gains if you will. Okay, this is the is the return is given just by the dividend in the in the following period. Okay, so this problem gives rise to a number of conditions, which I'll go over now. And we just say that this ZT in the in this other equation is one of our exogenous shocks and can be interpreted as risk premium shock. So this is similar to the risk premium shocks in the Smith's and Valters. Okay. So, on the supply side, we have a continuum of final goods firms that are monopolistic competitors and infinitely lift. And those firms transform an intermediate good into a differentiated final good. Okay. So they set prices are differentiated good subject to quadratic adjustment costs. Okay, so they maximize an objective function like this. So this is a dynamic problem because of the quadratic adjustment costs. And this gives rise to an optimality condition, which in a symmetric equilibrium. This is the form of a new Keynesian what effectively is a new Keynesian Felix curve capital by T here is gross inflation and this gross inflation depends dynamically on the gap between the average markup in the economy and the average markup is given by the price of the intermediate goods. Okay, times one minus tau tau are just a subsidy introduced for technical reasons. Okay, the gap between this actual markup that I just mentioned that the desired markup which is given by the usual function of the price elasticity. Okay, so this is pretty much a standard. Now let me turn to where Keynesian is, which is in the intermediate goods firms. Okay, so we have a continuum of intermediate goods firms, but now those firms produce an identical good and they are perfectly competitive so they take the price of that good as given. The price of that good is the little PT which was an input, which was the price of the input of the final goods firms now. In some period, they are exactly identical, but they are subject to idiosyncratic and aggregate productivity shocks exposed. Okay. So in particular, for a firm that draws an idiosyncratic shock to this is the production function. Okay, so this is the output and capital and labor that the firm will employ. This you see here Q is the shock and in addition to this idiosyncratic shock, we have an aggregate technology shock that follows in the ER1 process. Now for simplicity, we assume that Q can take just two values, one or zero. If Q takes a value equal to one, the firm is productive, if the Q equals zero, the firm is unproductive. Okay, and there is a mass mu constant mass mu of firms that are unproductive each period. Now, at the end of period T minus one, think of a firm now an intermediate goods firm that will operate in period T. Okay. Now, at the end of period T minus one, the firm is born, if you will, and issues equity in order to finance capital. Okay, now each firm, all firms are exactly identical. So they will get the same amount of equity. Okay, which will be QT minus one. Okay, we normalize the number of shares per firm to be equal to one. Now, at the beginning of period T shocks are observed. Okay, demand shocks and the technology shocks. Okay. And then each firm determines its optimal level of capital and each law optimal level of employment on the basis of those aggregate shocks and the idiosyncratic shock Q. Now, the gap between the desired capital and the initial equity funding is financed through the loan market. Okay, there will be a loan market. And the firms in principle can borrow and lend in that market at a real interest rate RTL. Okay. Now, during T or at the end of T, if you want, they will produce and sell their intermediate good at the price PT, which is taken as given. And they will sell the undepreciated capital one minus delta ATQ at the price PT at the end of the period and then they will distribute all the proceeds to the shareholders. Okay, that's the dividend that entered the budget constraint. Very good. Now, with one could compute the term in what's the equity return for a firm with productivity tool. Okay, and they just came as I said earlier is the dividend divided by the initial equity injection. Here is an expression for the dividend which is which revenues minus labor costs minus financial costs of, you know, borrowing the capital gap. Okay. Plus, the proceeds from selling the undepreciated capital now in equilibrium can be shown that the equity injection has to be equal to Katie so we can replace QT minus one here by Katie and can write down this return on equity in terms of equity and other variables. The differences. Okay, very good. Now monetary policy is just given by a Taylor type role. Okay, so the nominal rate responds to inflation and to deviations of output from steady state according to this rule. So Bang will choose these two parameters. Very good. Now, let me start by describing the frictionless benchmark. So here all potential lenders observed to and contracts are fully enforceable. Okay, and rather than going through the. Okay, the, you know, these algebra, let me just show you some. The implied aggregates of loan supply aggregate loan supply and aggregate loan demand functions because they are quite they are quite intuitive. Okay. So this is the loan supply. This is the loan market. Okay, this is the interest rate on loans. Okay. And let's look at the loan supply that's given by the blue line. So if the interest rate on loans is less than minus the depreciation rate. It doesn't pay even for unproductive terms to lend their capital in the loan market. Okay, so they they're the supply. They would they would instead they may want to borrow as much as possible. Keep the capital idle and resell it at the end of the period. Okay, so the supply of loans will be zero. Now, if the interest rate in the loan market equals minus the depreciation rate, then those firms are indifferent between keeping the capital idle or lending it out in the loan market. So we have this horizontal line here, and this is given by the, you know, up to this point, this is KT, the initial capital that all firms have times you which is the mass fraction of unproductive firms. Okay, so for any interest rate in the loan market above that unproductive firms will be willing to supply all their capital on the loan market. Now, if the interest rate is below is above sorry this threshold that you see here for this stress or this threshold is just that the what we could call the net operating return on investment on by productive firms. Okay, so if the interest rate is above this threshold, and this includes, you know, this this object here is just the output capital ratio. So the numerator is the marginal product of capital. This is the average markup in the economy. This the whole expression again as I said it's it's the net return on investment. So if the interest rate is above this threshold, then even productive firms will not to put their capital into work and they will instead supply it on the loan market. That's this additional segment of the loan supply. The loan demand the loan demand is straightforward so if the interest rate in the loan market is above the return on investment then no one will want to borrow from in the long market in order to increase their capital. If it's equal to the net return on investment then we will have this indifference. Okay, and if it's below the, you know, the demand for loans will be infinite. Okay, now if we combine the loan supply and the loan demand and we examine and we look for the competitive equilibrium in this long market we see that there's only one one equilibrium. Okay, which is given by this point P in which, you know, all unproductive firms supply their capital, which they don't really need or want in the to in the loan market and productive firms borrow in the long market in order to increase their capital and interest rate in equilibrium in the long market is given to the return on investment. Okay, so we have again we're looking at the frictionless benchmark still in equilibrium the interest rate in the long market will be equal to the return that unproductive firms get from investing in the loan from lending out their capital in the long market and will be equal to the return of productive firms and it will be equal in turn to to the return on investment. Okay, now output in equilibrium then will be given since all capital has been reallocated so to speak to productive firms will be given by these production function and the equilibrium can be shown to be equivalent to that of a standard model with a representative firm. Okay, so nothing particularly interesting happens. Now let's look at the case of frictions, the same model so now we are going to assume there is a symmetric information so firms only on no potential lenders can observe the productivity of a firm and this limited enforceability of long country contracts borrowers can just run away with the money. Okay, if they find it is optimal to do so. So now let's think about the options that an unproductive firm has in that context, it can borrow to increase its capital. Okay, this is an unproductive format, it can borrow to increase its capital keep it idle and sell it at the end of the period and run away with the money that's gone. Okay, now what's the implied payoff in that case remember the firm is in productive so it will not really produce or sell so what it will do is just to sell the unappreciated capital that it borrows and it will borrow as much as a productive firm would want to borrow that is it will pretend to be a productive firm. Okay, when it shows up at the loan market. Alternatively, it can lend out its capital in the loan market in which case the implied payoff depends on its initial capital KT times one plus the interest rate in the loan market. Okay, obviously, we want to avoid and the possibility that the firm finds it incentive to run away with borrow to borrow and run away with the money so that requires that the payoffs have that the second payoff is no lower is no smaller than the first payoff and that implies a maximum leverage ratio which is given by this. Okay, so notice that the lower is the return on the loan market, the lower will be the maximum leverage ratio because in that case, the lower is the return on the lower market, the higher is the incentive for an unproductive firm to pretend to be productive borrowing the loan market as much as it can and run away with the money. Very good so that gives rise to an aggregate demand for loans. Now the aggregate supply for loans will take the same form as before that has a change but now the aggregate demand for loans is given by this and it bends backwards if you will. Okay, and the reason is that for interest rates below the return on investment, the incentive compatibility constraint is binding and it takes this form as I mentioned earlier the lower is the return, the interest rate on loans, the tighter is the maximum leverage ratio and the smaller is the amount of funds that productive firms can borrow. Okay, and productive firms now will not want to borrow because again because of the incentive compatibility constraint. Okay, so now there are two possible cases. Suppose that the return on investment is high. Okay, and have a high with what we mean is that this condition is satisfied. Okay, so that depends upon other things on the market and it depends on the marginal product of capital. So suppose that the stock of capital if you want is not too large. Okay, so the return on investment is still high. In that case, we will have the equilibrium of the loan market can be represented by a diagram like this. And we will have three possible equilibria. And there will always be an equilibrium without trade. This is the autarkic equilibrium in which no one borrows our lands. Then there is an equilibrium which is unstable and which we are going to ignore. And instead we will focus on this equilibrium which corresponds to the efficient equilibrium allocation and equilibrium that we would observe if markets were frictionless, in which all the capital that firms, unproductive firms has is reallocated to productive firms. Okay, so that's the case where, you know, this term, the return to investment is sufficiently high. But what happens if the if maybe because there has been excessive capital accumulation, the return on investment on new investment is low. Okay, then, by the way, in the previous case, if there will be a change that corresponds to what we call normal times. And again, since all the capital is reallocated to productive firms, again, we're back to the case of equivalent to the case of a representative. Now, suppose that the return on investment is low, below this threshold, then we will have a situation like this in which, you know, the demand for loans and the supply for loans don't intersect except for this point. Okay. And again, the reason is that the return to investment is so low that the interest rates that in the loan market that productive firms can pay are also low and are too low. And the fact that they are low implies a very tight maximum leverage ratio and demand for loans that for any interest rate on loans is always less than the supply for loans. Okay, so the only possible equilibrium here. So if you want, there will be an always an access supply for laws that will put downward pressure on the interest rate, and it will drive the economy to this equilibrium in which there is no, no, no trade. And in which, you know, that's what we call a financial crisis, and in which aggregate output is now reduced by this by this wedge here one minus me. Okay, because the capital of productive for productive firms will not be reallocated to produce. Very good. So now let me examine and briefly comment the condition under which a crisis will emerge and then I will just I'll show you the anatomy of a typical crisis based on some numerical simulations. Okay, so this is the condition under which the only equilibrium will be the equilibrium without with with a collapsed loan market. Okay, so again, this is the aggregate marginal product of capital if you want that this is the, this is the average markup. Okay, so you can see that given KT, even the capital stock, which is predetermined. Okay, a crisis can be induced by a lower why lower aggregate output or a higher and or a higher average market. Okay, so typically say a negative demand shock will, we can bring about a crisis by increasing the market or reducing. Okay, now in that case, a monetary policy should seek to stabilize both in the short run. Okay, so that's the usual, the conventional approach to monetary policy that would say look at the response to demand shots and stabilize output stabilize the market. Okay, but notice that this condition may also hold and the crisis may emerge if the capital is too high. Okay, but of course the capital accumulates over time gravel. Okay, so, and in that case, the larger is the capital the smaller will be the shock demand shock or a technology shock whatever that may trigger the crisis so what can, what can monetary policy do in that case well it should prevent the capital from accumulating excessively, because otherwise the financial fragility of the economy will increase because the smaller shock will trigger collapse of financial markets. Okay, but of course this is a process that takes place in the medium. It's not not in the short run. In addition, there are some feedback effects because anticipation of a possible crisis raises precautionary savings by households and that increases capital accumulation and hence the probability of a future crisis. Okay, so the reason an additional reason why the central bank may want to preempt that excessive capital accumulation. So now let me show you a picture that kind of describes the what is a typical financial crisis this economy. Okay, so we calibrate the model in a standard way all the parameters are standard with some standard table rule as a baseline with a fraction of unproductive firms being this relatively small 2.4%, which implies a crisis incidence of 8%. That's the only non standard. We simulate the nonlinear model over one million periods using a global solution method. We identify the crisis and the starting dates of crisis and then we record the values of a different of the shocks and the variables around those crisis starting dates and we report the average values around the simulated crisis and here you have the you have the those average values around starting dates of crisis simulated in our simulation. Okay, so zero is the starting date of a crisis. And here you have, for instance, the supplies of the supply shock around, you know, quarters around that the starting date of the demand shot capital stock and so on. Okay, so you can see that in the typical crisis. Oh, by the way, and the horizontal line gives the steady state problems for all the variables. The typical crisis is one that because of positive supply shocks and positive demand shocks, especially positive supply shocks. Okay. That leads to a large accumulation of capital well above the steady state. Okay, and what happens is that those supply shocks just go away and that's the same for the demand shocks. Okay, so this high capital becomes unsustainable. And, but then a small shop technology or demand triggers a crisis and that's what we see here. And after the crisis, you know, the capital stock will be will decline but only gradually and then you can see that during the crisis there was a significant decline in. Okay, and also quite, quite persistent. Okay. So, what should the monetary policy do here we were assuming a standard table table. Okay, so what are the options for monetary policy. Now, in the absence of financial frictions, the optimal monetary policies straightforward in this model, because the only the only imperfection or distortion we have is the existence of sticky prices so the optimal policies is Okay. However, with financial frictions. This is not optimal because what the strict inflation targeting does is to replicate the flexible price equilibrium allocation. And this is not efficient because they may involve, you know, too many inefficient crisis. And the inefficiency of those crisis is the result of individual agents not internalizing the consequences of their decisions on financial fragility of course like in many months. So a strict inflation targeting in that context fully neutralizes demand shocks. Okay, so that's good. So because it will eliminate demand driven crisis but it amplifies output and capital fluctuations driven by technology shocks. Okay, and that will be a sort could be potentially a source of crisis because of the excess capital accumulation that they may generate. So what may be more desirable out of policies that aim at stabilizing output then so let me show you a table. Okay, I don't know how much time I have a maximum maybe you can give me. I don't hear you. Sorry. Two minutes, two minutes. Okay, so let me show you this table and then I will just conclude. Okay, so here we have, you know, think of the model with without financial frictions. Okay, here you have different values of the coefficient on inflation and the coefficient of an output. We know that in that model, you know, the optimal optimal policies given by a strict inflation targeting and any policy that deviates from that will generate some, some losses. So these are the concept consumption equivalent losses permanent consumption equivalent losses. Okay. And in particular, you can see, as is well known and because some of the intuitions are driven by technology shocks. The larger is the coefficient on output. The larger are the welfare losses. Okay, so in that we know that you know in that model without financial frictions it is optimal for the for the central bank to single mindedly stabilize inflation because that will also stabilize the output gap. Okay, which is what is optimal as opposed to stabilizing output. But now let's consider the case of frictional markets. And you see here what happens as we increase as you know starting from the strict inflation targeting. We adopt the standard table rule and alternatively we creep increasing the coefficient on output. You can see that the welfare losses become welfare gains relative to strict inflation targeting. Okay, and the main reason is given by this column here you see the incidence of financial crisis under each rule. And you can see that as the coefficient on output is increased. Okay, the incidence of financial crisis gets reduced. Okay, and obviously that's desirable because those financial crisis into a misallocation of capital inefficient output losses. Okay. So, let me skip this part. Well, we have just to, we show that we introduce in the last part of the paper we introduce monetary policy shocks and we show that an additional reason why financial crisis may emerge is because of monetary policy that is to lose. Okay, by the result of sequence of shocks that are on average imply on average interest rates that are below the interest rate implied by the table rule, but also an interestingly, okay, and the crisis may be triggered at the end of the run up period if there is a positive monetary policy shock. Okay, so we interpret this as a warning against the temptation by the central bank when it's late in the crisis and capital has accumulated excessively to increase the interest rate. Okay, because then it's too late. It will not have any impact on capital accumulation anymore. And instead, the economy is sufficiently fragile that it may trigger a financial crisis itself, as we see in this case. Okay, so let me just conclude. So we've shown a simple extension of the basic New Kingdom model with financial frictions and indigenous financial crisis. We focus and this is important. Our model focuses on one dimension of financial crisis. It's not meant to be a comprehensive model of financial crisis in particular on the misallocation and loss in productivity resulting from financial markets not doing their work. Okay, and there we have, there are many papers that provide evidence of such a misallocation taking place during the particular during the much recent financial crisis. And the lessons for monetary policies that there may be a rationale for deviating from price stability of a single focus, and in order to have a financial fortune. Thank you for your attention.