 So I was glad to hear Jeremy talk about Piguvi and taxes. Remember discussion we had many years ago, right after the crisis where we were, Javier and I were talking about Piguvi and taxation as a way to control the sort of money creation, the excessive aggregate liquidity risk by the price. And at that time, Jeremy told me, you know, the problem is when you're dealing with someone who's addicted, you cannot just rise the price of coming into the casino, you just have to stop them at the door. So I think that's maybe the point of having quantities where you put limits to disposal and maybe at the aggregate level or so. But if we, now if, and this is a if, if we've done enough of a good work in giving them enough buffer, or if we feel that we can track better the creation, then we can go back to a less dramatic and in fact more efficient as you pointed out, system of managing excessive short-term creation, the private sector. The alternative is indeed to supply on the public side because it's clear that the demand is there. But to the extent you want to divide the task, then I think some instrument there, such as liquidity tax would be appropriate. That's not really a question, but Jeremy, do you want to comment there? Maybe I'll just use this as a quick opportunity to comment on something that Olivier said before, which is, I mean, of course, completely correctly, that it doesn't really make sense even if you bought all this stuff about like maybe the Fed should do it or the trade. It doesn't make sense to think about a target for the size of the Fed's balance sheet. If you were gonna get in this business at all, it would probably make sense to think of a price kind of thing, like let's do it in your preferred world where the treasury implements it. I would say the treasury should pay attention to the slope of the front end of the yield curve and they should organize their issuance accordingly. But that makes much more sense than saying, I have a target for how many T-bills I'm gonna have. If they just think of themselves, if the treasury would just think of themselves as basically trying to issue at low cost, but be a little bit more adaptable than they've been in the past, that might get you some of the way. But surely I don't think a quantity target so we'll make a whole lot of sense. Why do you think it's a good idea to go in the direction which the Fed is going now, which is to reduce the balance sheet because you seem to be agnostic. I mean they still have 20 basis points to be eliminated from increasing the balance sheet of the Fed. So we could decide that we're going to do it until that wedge is gone, in which case we should be increasing the balance sheet of the Fed now rather than starting to decrease it. Why do I think? No, again, my basic preference, all else equal, would be to have more, and whether it's the Fed or the treasury, would be to have more short-term claims in there. I mean that argument has been, the history, the sad intellectual history of this is we started actually with the treasury trying to urge this. But you seem to think that it's better if it is a central bank to do it rather than the treasury. Well, you know, I mean. If so, then I think you have to be agnostic as to whether Fed has the sign right. Whether they have the sign right right now. Whether they are decreasing their balance sheet where they should be increasing it because treasury is not going to do it. Okay, so, okay, let's stipulate that. I don't think necessarily the answer was ever to get it to zero because there's a trade-off. And so, you know, I think some of the slope right now, as you said, there's about 20 basis points of slope. I think some of that, you know, to appeal just to the expectations hypothesis might get you some of the way. There's probably one hike at least baked in over the next six months. So, you know, my guess is it is flatter now and that there is some effect. I wouldn't want to overstate the effect. And I wouldn't, given that the effect is not overly powerful, I wouldn't want to say that the job is to get it all the way to flat. But, you know, I would like to not see it reemerged with the sort of kind of steepness that you saw before. Yeah, and if it were me, if it were me, I would be in much less of a rush to shrink the balance sheet, you know, that's not my call. Let's say, that's a fascinating discussion. I would just like to encourage you not to lose totally sight of the fact that central banks are also in the business of doing monetary policy. So, if there is anything you would like central banks to do for financial stability reasons, you've got to explain how it squares with the monetary policy mandate. And I guess the answer then in your model would have to be about the composition of the balance sheet, which is what you said. But let's not lose totally sight of the monetary policy discussion. Please. Any other question? Please. Let me ask the obvious question. Elu Funtaden, University of Mannheim. Let me ask the obvious question to Jeremy. So, if 4,500 billion is the right size of the balance sheet of the Fed, we clearly got it very wrong before 2008. And so, is there a reason to assume that the lack of regulation before 2008 resulted in the private creation of those short-term assets in a dangerous way? And could by expanding the balance sheet before 2008 to 4,500 billion, we have avoided that type of private supply of short-term assets, which was dangerous? So, is regulation complement to balance sheet size or is there some subsidizability in terms of now going back to smaller balance sheets as many people seem to want? So, again, I think that this is very much in the spirit of Benoit's question. I think the pre-crisis thing was characterized by a combination of no regulation or inadequate regulation and a strong economic incentive to fund at the short end. And the right answer, I think, in an ideal world, you do the best you can with regulation. That's in some sense the first best instrument. If you think that it works very well and that there's no sort of evasion around the margins, then that's it. And we don't really know yet right now because I would say right now we've done the regulation but the incentive part has not really re-emerged. And again, we'll learn, we'll basically learn as the yield curve steepens. I suspect, just sort of based on the fact that there's regulatory arbitrage is sort of everywhere, that the regulation by itself is not gonna do as well as some complementarity between regulation and trying to keep the incentive somewhat moderate. Maybe to your point, Jeremy, maybe one additional remark is that the regulation doesn't apply the same way to banks and non-banks, as we know. So the constraints on the maturity transformation has not as yet the same say in the money market universe on the investment fund universe, even though the FSB is working on it. So that would be the worry, that would be precisely the worry, the yield curve steepens and we start seeing various shadow forms of maturity transformation, that's exactly right. Another question. It's a question to Professor Einstein. Can you recall in your model where the demand for a short term bond from the private sector comes from? Is there a friction, a very high risk aversion for example? And if with the crisis and now that the crisis is gone, is this friction gone? You know, it's a very good question and there's sort of a deeper take on it. In other words, a superficial answer would be to say something sort of institutional, to say well, you know, there's a lot of money market funds and there's three trillion of money market funds and so they need to have short term assets and there you go. And then I think a pretty powerful criticism of everything I've been saying so far would be to say you're taking all that demand as sort of given and in some sense as legitimate. You're saying well, we should be sort of organizing society to sort of accommodate that demand either via the private sector, the public sector but maybe that demand is in itself somewhat excessive. So just to give you an example, where's all this money market fund stuff come from? Some of it probably comes from all the cash that corporations have accumulated. That in turn has something to do with taxes and repatriation and all of that and maybe we shouldn't be doing all these gymnastics just to accommodate that demand and the right answer is if we sort of did something with the tax code so that guys repatriated some of the cash, there'd be less cash and money funds and some of this demand would go down. So I think that's a very fair point and I think it's sort of related to Benoit. There are other instruments in some sense that one might want to bring to bear that would be closer to first best instruments than essentially implicating the government in just uncritically accommodating this money demand. So I think that's completely right. Well, at point, there's an argument which has been made against not Jeremy's specific argument but that line of argument which is the people have incised which have focused on the wall of collateral and have argued that when the Fed does this it basically removes good collateral by basically holding it on the balance sheet of a central bank rather than where it should be. And I'm not sure I can make sense of this because I mean the Fed is actually providing amazingly good collateral which is money, except that it goes to the banks and people who need the collateral are not the banks typically are not primarily the bank. So it's something interesting. I suspect that the answer is that there should be a way of using cash as collateral outside the banks and so I suspect that the way this liquidity is provided should probably be extended to more than banks which would be the solution. But I mean in a way if I listen to the collateral argument I would conclude that what you're suggesting is completely wrong but I think it's not correct. I mean it's all right. So let me just totally agree with one thing which is here's I think an easier, lower hanging bit of fruit which is whatever we agree the size of the Fed's balance sheet should be I would prefer that the liabilities not be reserves but be something like their repo. You mentioned central bank bills. If we think that the world has a demand for safe short-term assets, why are we creating all these things and saying oh only the banks can hold them? That just seems like it's a violation of every Friedman rule type of principle you would want to have. You should make the stuff freely available to whoever wants it. So I always liked the RRP program in that sense and I would have loved to have seen them essentially put the rate on that higher and so thereby you'd be using your liquidity a little bit more broadly and I think that would answer your point and would not make their footprint any bigger in the aggregate but there seems to be for reasons I don't fully understand a bit of traditionalism associated with doing it more in the way of reserves and less in the way of repo. Can I just comment on this one, Beno? I think these two comments I think raised the bigger issue of who should have access to the central bank balance sheet and Beno, you're very familiar with this debate. One debate that's been going around has been on the role of central bank digital currencies, how widely they should be, what form they should take, who should have access, what's the elasticity and so on. I think that raises a whole host of questions which have to do with the nature of intermediation, what are the financial stability risks during periods of flight to safety that are going way beyond the narrowly focused question about the demand for safe assets. So I think we should try and take a step back and take in the larger picture here and perhaps in a position within a larger context. I was going to say exactly the same as you, so it must be some kind of central banking DNA somewhere. That if that becomes a broader discussion on who's having access to the balance sheet of the central bank, then we should look it through. We should look at the consequences for the future of the banking system or the kind of financial structures that we want to see looking forward and sometimes have the impression that we are jumping a little bit too fast to the conclusion that, well, the banks are no longer performing the social role that we would like them to play, which is maturity transformation that essentially is a social role of banks. And so we've got to move on and write them off and move to something else. But again, that brings me back to my first question to Jeremy. That would be a little bit also writing off all the regulatory effort that we've been having since 2007. So that might be the conclusion eventually, but I would like to be sure, I would like to be sure. So let's give a chance to intermediate finance before we move on. Luke. Yeah, so there's another argument that you hear often actually around here that Olivier did not include in this, otherwise I think exhaustive list, which is that there may be such a thing as a reversal rate. So the idea of now at some point when rates are low for a prolonged period of time, it may actually turn out to be contractionary, monetary policy, because it's basically eats up the net interest margin of banks. So Marcus Brinamire, one of his students, has written down a model of that and it's a bit unfair to ask the question since he's not here. But just more generally, I think this is also very much in the public debate. So this is less the plumbing and the collateral, et cetera, et cetera of what happens once we exit, but it's more sort of some of the impetus created in certain parts of society calling for an earlier exit. So what do you, as a group, kind of make of this type of arguments? I'm open to the idea that the sign of the effect of V interest rate, whatever it is, on aggregate demand changes as you get to very, very low rates, in which case it's not a good idea to do it. So the notion of a reversal rate strikes me as logically consistent and maybe even empirically relevant. Does this lead me to want to execute QE faster or not or increase interest rates? I'm not sure. I would have to look at how, you know, where the adverse effects come from and whether they come from the QE part or from the policy rate part. And I have not thought about that part of the answer, so I should stop here. No, I mean, I forget who mentioned it earlier today, but one of the previous speakers mentioned the point. There's sort of, if it goes negative, and then if it's just very low for very long, that the effect of, you know, the income and substitution effects may play out differently for a very permanent change than for a short. I don't know that we have any evidence on it, but I think it's worth at least entertaining the conjecture that, you know, the stimulative effect becomes weaker over time because people start worrying about saving more effectively. Well, that's clearly something we're looking into, but so far it seems that the reversal rate, and I agree with Olivier that it has to be a reversal rate somewhere. Question is, how low is it? And so far it seems that it has to be either very low or low for very long to materialize because there are cumulative effects, as Jeremy just said. And when we look at the eurozone today, where the deposit rate, which is the main policy rate in the current environment is minus 40, it seems that the, so far, but against so far, the negative impact in particular on bank lending of having a very negative rate has been more than offset by the general equilibrium impact on aggregate demand for loans, et cetera. So this might not be true eternally and this would not be true if we would push the deposit rate much lower, but so far it seems that we still have a comfortable margin. Could we just come in on this one? Benoit, I think we had a very interesting paper earlier in the day from Glenn Shepards on why the, at least for the retail deposit rate, it seemed the zero is pretty hard, low bound. And I think a very interesting case study of where this reversal rate might be is just to give it time and see what the impact of a very prolonged period of negative rates might be. Because I think it will actually have an effect on some of the structure of the intermediation as well. And let me just mention the experience that the Swiss have had recently, which is that what happened when the deposit rate went very negative was that the mortgage rates actually went up because the larger banks were able to mobilize their pricing power to actually raise the mortgage rates. But what we've seen, I think more recently, is that as the period of low deposit rates have persisted, the structure of intermediation itself has changed in the sense that the players that I talked about, the insurance companies have discovered these very long term mortgages have very fixed income like attributes and it's the insurance companies who have also entered the mortgage market. And then the initial pricing power that the banks thought they had may not be all that lasting after all. And so I think where the reversal rate may be depends not only on just a number, but also on how long this period persists. No, that's absolutely correct, but then it depends also a lot on the kind of degree of competition that you have on the retail market, which is different in different countries. It's certainly different in say in the Eurozone in Switzerland, in Denmark. What we see in the Eurozone is we still have too many banks, there's a lot of competition. So that's probably not the most likely negative effect. So of course there is no free lunch. That also implies that the low interest margin means implies low profits, which is less internal capital accumulation so that might be an issue for financial stability over time, but it's probably more of a long-term impact than a short-term impact. Thank you, Bert Sinetal from UBS. I was wondering since the title of the panel is exit from non-standard monetary policy and we've talked, or you've talked a lot about the Fed. And I was wondering, I mean, the Fed has started to exit and it has been going very well. I mean, the curve has been very well behaved. In fact, it has flattened. Long-term yields are low. I was wondering what do the panelists draw? What conclusions do you draw from this? I mean, does it mean that maybe the exit is actually not as scary as many other central banks, I guess, including the ECB, the RICS banks, it's National Bank, particularly in Europe, and in Japan as well, where central banks are still engaged in this non-standard monetary policies. I mean, would it be correct in your view to draw the lesson that maybe all this withdrawal of QE may not have such negative impact in the end? That's a very good question. Let me just say that we're not scared by exit. I mean, that's not part of the job description to be scared by anything. So we just want to do it carefully, prudently, and in the light of our price stability mandate, so in the light of the impact on medium-term inflation. But I would be interested to hear the answers of the other panelists. I mean, I think Hewitt has it exactly right. I mean, you can do the usual stuff and say, you know, all this has been telegraphed to the market, so there are no surprises coming to the market, so if the thing is sort of market expectations, we don't expect much. And if you ask me for a baseline about what's gonna happen, at least with the Fed's wind down, I would probably say my baseline is, you know, it's not gonna be a big deal, but we just don't understand markets all that well. We are in a situation where, you know, some asset markets, if you look at like US credit and other things like that, or at least, you know, valuations are at least moderately stretched, you know, to say what the trigger will, even if you believe that you can forecast returns a little bit based on valuations, we know very little about triggers. But at the same, I mean, I would just be, I would be uncomfortable being overly complacent. And then I think Hewitt made some great points about the interconnectedness of all of this. Even if sort of the narrow hydraulic effects of just the US, the Fed shrinking its balance sheet, even if you think that those are sort of modest, I suspect that a lot of the low-term premium and the narrow credit spreads that are in the US, for example, we've imported from Europe, right? And there are these nice, you guys have probably produced some of these charts where you can see the amount of, for example, US corporate bonds that have been bought by European asset managers, right? And so there's all these kind of floak type of things going on that I think we don't understand super well. So there's gotta be some tail out there. Again, it's different than making a baseline forecast, but. I don't think, there you go. Just one quick remark, which is I think the portfolios of the different central banks are different. And maybe the exit from QE from the Fed is simpler than it is for the ECB, the extent that you have the implications for fiscal positions in various countries of the Eurozone are more complex than the implications of for the fiscal position of the US government. So yes, so far so good and clearly a clear, at least initial path, not final path has made an enormous difference and things have been priced in nicely. But the challenge is still there, I think for the ECB and surely even more for the BLJ. You know, I think it's very tempting to reach for the textbook model where we have a single individual investor and you can sit down with the investor and say, look, let me tell you, it's gonna be very gentle, it's gonna last several years, so there's nothing to be afraid of. And I think that's the kind of picture that we have in mind. And I think that's certainly a possible scenario. And I think what we should bear in mind is that the market is a very complicated place with different players, with different motives. And I think it's worth thinking about how those interactions will actually play out. You know, one thing that we do know from the academic literature is that bond markets do overreact relative to the expectation through the yield curve. And there's a very interesting paper by Jeremy's colleague, Sam Hansen and David Luca and Jonathan Wright, which have looked at the recent evidence on bond market overreactions. And what they find is that the frequency and the extent of that overreaction has increased markedly since the year 2000. Now, I mean, you could say, well, we shouldn't really worry about this because it's just a short term, some short term effect on the market, which will die down. And that was certainly the experience we had in 2013 with the taper tantrum. I think the difference perhaps this time around with the taper tantrum was that in 2013, the backdrop was still very accommodative. I mean, there was the central bank asset program in the US just backing up. This time, if the broader backdrop is less accommodative, then I feel like the final destination may be somewhat more uncertain. And I just go back to my previous point during my introductory comments where in textbook comparative statics exercises, if you have a demand curve that slopes upwards, as well as a supply curve that slopes upwards, depending on how you draw those curves, you could certainly have multiple equilibria. And that may be one kind of scenario where everything makes sense, but it's very difficult to foresee where you land up. And I think this is another instance of the old adage that in finance, it's much easier to talk about consistency across assets at a moment in time than to think about the inter-temporal consistency. Because the inter-temporal consistency may be much harder to achieve because there is much less arbitrage going on across time than you have across any single date. And so I think this is something that is food for thought. And I know that, and we are sitting in front of people who have really contributed a lot to this literature. So maybe we could ask them at the end of this panel. Bonnick, I think we're looking at you. They don't seem so eager to answer. Do you want to answer? No. No, two things. I mean, first, I think the point that Hume just made and made earlier in our discussion on granularity is very important. If you want to understand the impact of whatever we're doing on market prices, you've got to be very granular. It's not always about different political structures, which Olivier mentioned, and that's important. It's also about different differential structures and different people holding the bonds, right? So just to illustrate, Hume, you gave the example of, or you flagged the risk of long rates going higher, insurance companies or pension funds marking down the liabilities and then dumping bonds and having some kind of convexity there, which is clearly something that that's possible. What we've seen so far is exactly the opposite, that us buying bonds from investors who are increasingly buy and hold investors, increasingly regulatory constrained investors, putting a higher price on the bonds because of the regulatory constraint, meaning that that has contributed to the very low level of long-term rates, or we've been able to inject the higher amount of duration for a given quantum of purchases because of the nature of the investors we had against us and on the other side of the market. So you've got to understand all of that to understand the impact and it's very difficult, it's very difficult looking forward. May I ask you one last question because we're coming to the end of the discussion. The discussion has been very much about balance sheets, volumes, not that much about rates, which is of course an important part of what we're doing, forward guidance and all the like. And there is one striking difference or striking to me between the US and Europe, which is about the end points, the terminal rate. So Fed has communicated quite a lot on their expectations or FOMC members have to communicate on their expectations of long-term interest rates, or the long-term value of interest rates easily. So you have a fairly good notion of the distribution of the terminal rate in the US and that's not something we're doing. So is that important as a non-core or how much this contributed to the, to stabilizing the exit process? I think it's, can I, maybe I could start with this and hand over to the other panelists. The famous dot plots, the summary of the projections from the FOMC members, I think is something which is very widely, which is very closely followed. And yet it's also an instance of one comment that I made earlier, which is that there does seem to be something of a gap between what market prices are signaling as to what the future conditions will be and what we know is coming up in that the market implied path of interest rates seem to be somewhat different from the summary of economic projections. And I think it's, I think it will certainly have a value, but I think it's also, I think illustrating that if it's a forecast and it's filtered through the market process, so it's not simply a statement about one's intentions, but rather it's a forecast that's filtered through whatever economic model one has. The signaling value of that may be less than full. And I think it's quite telling that we do see this wedge between the market implied path and the dots. You've been disclosing your... I've been a dot. Your dots. You've been a dotter. Yeah, you know, I have to say, you know, obviously I understood the theory of forward guidance, you know, that there's sort of strong form forward guidance at the zero lower bound, where it's more than just, you know, this is my best guess, where it was at least in our case, it was a quasi-commitment. So that, I think, you know, there was a textbook and I think it was the right thing. Away from the zero lower bound, I understand a little bit less well what the role is. And I have to say, I have a little bit of discomfort because I worry, I mean, I worry that you could get into a situation where it basically confines your ability to move when you need to. Now obviously the Fed has not had a problem not meeting its guidance when that, not meeting it has been going slower than the market expects. I wonder if there comes a time when they need to be more aggressive than the market expects, will they feel sort of somewhat constrained or reluctant because they kind of put a number out there. It's like a corporation that puts out earnings guidance and then feels it has to meet or exceed its guidance. So a little, you know, away from the zero lower bound, if you could kind of make the dots go away, I think it's a good, it's a debate worth having. And this is a long discussion what the terminal rate is likely to be. My own view is that my own distribution has a much larger up a tail than the market. I don't know where I would put the dot if I was allowed to put a dot. Dot might be in the same place as the others, but the distribution around it would be very large. Because I mean, the more I, that's something on which I've worked, which is where do the low rates come from? And if it comes from a decrease in the rates of return in general, and then if you think about the reasons we give from the global savings dot and all, it's not clear it will remain, or it comes from a safety discount. And there you can tell stories where it will also largely go away. So my sense is people in the market seems to be expecting very low rates. And I expect, well, at least, I think there's a risk of much higher ones. But I've said this for three years now and I've been wrong, so. But that, by the way, that's exactly where I was coming from. Not that my point estimate is all that different, but I think there is that tail. And you wanna be, if that tail manifests, you wanna be able to do what you need to do. And I just don't know enough about how the market will behave. But I worry a little bit that once you put this marker down, it's a little harder to react if that turns out to be the case. Thank you very much. I'm not going to attempt at summarizing the whole discussion. The conference reconvins tomorrow at 9 a.m. to discuss globalization, inequality, and welfare.