 I wanted to ask whether this is related to the corridor system in Central Bank, targeting an interest rate corridor and exactly where you target. Different central banks choose different interest rate targets relative to the corridor. And if you, some central banks like for example Canada, they target the bottom by increasing liquidity a lot. I had some impression that that was related. The difference between the Fed funds rate, say, and the bottom of the corridor as a practical matter might be related to the question of liquidity that you were discussing. But I wasn't sure on whether that was a sufficient statistic or whether there was sort of some additional interest rate in your model aside from the Fed funds rate because of this distinction between inside and outside money. Any more questions? I'll give you some time. Yes, please. You started off the talk with this interesting picture about the netting down and the role of the asset market. I just wanted to give you a chance to talk a little bit about how active traders, should we think like trading volume matters for monetary policy, for the mix between reserves and bonds, these types of things? Any others? Yeah, the question I have is how we should think of haircuts that central banks would generally also impose on the type of collateral that you have in your economy. So maybe we answer now these questions and then we'll give you both to you and Pierre. So first thanks a lot Pierre for a great discussion. Let me start with some of the questions that you raised and then I'll turn to the others. You asked about the micro foundation of these leverage costs and that's a good question. We've assumed this basically exogenous leverage cost function that is smooth and allows us to characterize the entire solution of the model analytically much to the dislike of Pierre but you can actually do it with pencil and paper and so that's why we did that. We are now working on a separate paper that is trying to micro found these leverage costs in a model with bankruptcy costs. So banks in that model can, there's uncertainty, they can go bankrupt and in bankruptcy they face costs, fire sale costs of selling the assets and that is then generating leverage costs that look much like the C function that I was showing you. So this is how I think about the micro foundations. The question is should bank leverage be zero in the model? Introducing production is one way of doing that but even in this model bank leverage serves a purpose because households need to pay for consumption with inside money and so without banks they can't do, without the inside money issued by banks they can't do that and so already here bank leverage would not be zero. In terms of thinking about taking this model, so in this model we're studying steady states and then we're introducing shocks that temporarily make you deviate from steady state and then you go back in a short amount of time so this model because it's so simple it doesn't have interesting dynamics. So you go right back to the new steady state but the advantage is that you can think about the effect of various things on inflation, on short run inflation and so that you can analyze things like what happens when asset values collapse, what happens to inflation in the short run and then you go back to steady state. We're now incorporating this basically as I mentioned I view this model as a module that you can attach to other models so we're working on a New Canaan version that has a banking sector of this model and in that model transition dynamics are slower and so then once you hit asset values that has an impact on inflation and then you have the economy transition back more slowly and so I'm absolutely agreeing with Pierre that to think about quantitatively what happens in these markets you have to take into account more interesting transitions. So what you're seeing here is basically the simplest model that provides the intuition about how this works with these layers. Amy's question about how does this map to the corridor? The two answers I think I don't know how exactly so what is true here is that just thinking about the nominal interest rate is not enough you also have to think about the bank balance sheets how that maps into a corridor system I will have to think it's an interesting question I will have to think about so definitely your here monetary policy has to set more than one variable. Irving asked about trading volume and that's one of the reasons we built this model is to understand what are periods in which inside money is sucked out of the goods market and is used for asset trading does that lower inflation and so we're now exploring this quantitatively and think that we can clearly see episodes in the data where during episodes of high volume you have lower inflation there's less inflation and good prices but that's all coming up in future papers. The question is here cuts that the central bank sets here we have this as Pierre mentioned where you can think of this risk weight on risky assets our role parameter in the paper as a policy parameter and to the extent that central banks want to influence here cuts on collateral that would be the parameter that they would change so they can set whether risky assets are good about collateral depending on this on this role coefficient that they set. Okay thank you yeah I just ask more questions that I can that's what I can do so I understand this justification you give for bank leverage cost but what about the government yes yes it's bank leverage cost so the an implication of a model where the government basically can provide very good collateral in this world if the if the government provides reserves banks can hold these reserves and it makes their it lowers the leverage cost because this is basically a perfect safe collateral and so without if the government doesn't have any leverage costs it should provide plenty of reserves and so the model says all banks should be in should be narrow banks they should back deposits with reserves that would lower their leverage costs and in the absence of government leverage costs that's that's great of course I don't think this is reality I do think that governments in reality do face leverage costs and for example it would not be optimal for Greece right now to just flood to have a flooded system with reserves unless it's the ECB you know countries that have where governments face where outside investors doubt the the backing of bonds by governments that's the type of government that is not going to or should not issue tons of reserves because that would be additional debt and that's not that's not optimal in and I think that's the right way of thinking about it so how do you think about government leverage costs in in the model the easiest thing which brings the government on to the same footing as banks is to have an exogenous leverage costs of course again you can think of there are models of sovereign defaults that have bankruptcy costs and the probability that the government may go go bust and so that's that's the type of leverage costs we have in mind and so the the amount the the amount of commitment that the government can make depends on its tax base how much things it can tax it attacks to repay bonds and that's this is here model in reduced form but a more micro founded model would have would have for example bankruptcy costs I'm wondering if the cost of the government having the central bank basically creating more reserves you're saying it's a leverage cost I wonder if it's sort of if you had I think this relates to a question that that Pierre asked which is if you had banks lending and if there was sort of some real efficiency to their lending then there's sort of a decentralization question which is you can imagine a situation where the central bank is very very big and it provides a lot of reserves so a lot of the money is provided by the central bank but then they're gonna have to do a lot of the lending to there's you know if there's a fixed supply of government bonds eventually the central banks balance sheet gets big enough they're gonna have to be doing the lending so I think that the arrangement we have now is the central bank essentially delegates to the banks some of the money creation therefore allowing them to be the lenders in the economy because we think the private sector makes better lending decisions so I don't know what you're calling a leverage cost for the government is really sort of a decentralization or a decentralization cost is that is that an okay interpretation okay interpretation because the basically if you think about how how can you provide us what what's the payment system how can how can households and asset traders pay for their stuff and so you need a payment system and so one one system would be for for all of us to have accounts at the central banks and then also get loans from the central bank that would in that world the central bank would do everything or we can have a system where banks do the lending and they create the inside money and the government only provides outside money that we don't use I don't pay you with reserves and so once we are in that world I think it does make sense to think of both as having leverage costs because if you if the government again has zero leverage cost it should be doing everything and it should be issuing tons of reserves and in a world with lending it should also do all the the loans because if they're if the government is perfect and they have no cost in doing things they should be doing things and so this is not the world I think we live in and so I think the world is more like our model where also the government faces some costs and so now you have a trade-off between these costs and and you have economies in which the banking system is very efficient and is more efficient than the government and more trustworthy than the government and you have other countries in which the government is more trustworthy and therefore can issue tons of reserves like in the US I think we the government certainly faces a lower leverage cost and then has increased reserves dramatically that would not be possible in other countries or should and would be inefficient from the point of view of our model final question sorry first I like the paper a lot I have a question on the title of the paper you say payments credit and asset prices why are you so shy and not calling it money credit and asset prices I say that because if you go back to Irving Fisher creation of exchange it's about transactions and only a Milton Frieden replaced transactions variable with income with a presumption that money would flow into goods markets and transaction amount for money for goods markets rather than for asset markets so I would go and introduce alongside the transaction amount for money portfolio demand for money in the Tobin in the Tobin type type world and that would be say that in natural I think the natural and the natural extension a natural way to go to move away from a pure and transactions amount for money approach just looking at the history of monetary theory that are you are you prepared to go that direction it's the next paper perhaps I know I don't know maybe I'm not gonna take a stance right right now where we're working on I think the most important task that we have on our plate is to look at the quantitative implications of a version of this model and so this is in the short run I think we're gonna push the quantitative side and then we may go back to pushing additional theoretical aspects okay thanks very much I think concludes is the first session I think we are only a few minutes late so thanks very much thank you