 Great to be here today. Great to be back at the ECB, and it's great to see so many friends and colleagues and a real honor to be invited to share some remarks for today's keynote. So as a policymaker, I have the opportunity to speak on a wide range of topics, but money markets is where I grew up as a central banker. So it's a particular pleasure to take a deep dive into the technical aspects of policy implementation today and to do so from the perspective of a policymaker rather than a practitioner. But before I go any further, I need to give the standard disclaimer that the views I express are mine and not necessarily of those of my colleagues on the FOMC. So in recent decades, central banks around the world have been shifting toward floor systems for policy implementation. And that shift has been motivated by the actions central banks had to take in response to the global financial crisis and subsequent stress episodes. By growing appreciation of the benefits across a wide range of economic and financial environments, and by post-GFC changes in bank regulations and banks own risk management, which increased the demand for liquidity and the costs of interbank transactions. So among the key benefits typically cited is that they do not put a price on liquidity. In a floor system, banks' demand for reserves is satisfied with market interest rates close to the remuneration on the marginal dollar of reserves. There's no incentive for banks to economize on reserves, the most liquid asset in our financial system. As a result, floor systems satisfy a version of the Friedman rule. The opportunity costs to banks of holding reserves is approximately equal to the central banks' cost of supplying them, which most analysts view as small. I'm speaking here about the Friedman rule in an environment with interest-bearing reserves. When money does not bear interest, the Friedman rule also has implications for the optimal inflation rate, but that's beyond the scope of my discussion today. So in theory, because floor systems remove incentives to economize on liquidity, they should reduce liquidity risk in the financial system. Yet in economies around the world that use floor systems, we've seen serious liquidity strains in recent years, including the pressures on LDI funds in the UK, the banking challenges in the U.S. earlier this year, the repo market pressures in 2019, and of course, the global dash for cash at the onset of the pandemic. So in light of these events, some observers have questioned whether floor systems deliver the promised liquidity benefits. Some have even argued that floor systems make liquidity risk worse, an idea I strongly disagree with. And some have questioned the other side of the Friedman rule equation asking whether central banks can supply large quantities of reserves at minimal cost to the financial system and to society. So today, I'll address both sides of those critiques. I'll argue that floor systems do reduce liquidity risk, but they don't eliminate it. It remains incumbent on all players in the financial system, including banks, other market participants, and central banks in both our roles as regulators and financial institutions to appropriately manage liquidity risk. The appropriate strategies to manage liquidity risk look somewhat different in a floor system, and I think all players have work left to do to fully gain the benefits. On the other side of the equation, the costs of supplying reserves, it's important to distinguish between two types of floor systems. When the central bank's balance sheet is only as large as needed to meet the demand for liabilities, such as reserves and currency, we call the system a liability-driven floor. That's the approach we've selected for the long run in the United States, where our long run operating regime will supply ample reserves. And I don't see large costs of supplying the quantity of reserves needed to establish a liability-driven floor. But there are also circumstances, such as the pandemic in which it is appropriate for central banks to acquire additional assets to provide monetary accommodation or address market dysfunction. And such balance sheet expansions create an asset-driven floor, where reserves are more than ample, and you could say abundant. Asset-driven floors can pose some challenges, but we can mitigate these challenges when circumstances make asset purchases appropriate. But it's better to minimize the challenges when we can. And that's one reason I think it's important to continue normalizing the Fed's balance sheet. Another reason, of course, is to remove monetary accommodation to restore price stability. So in the rest of my remarks, I'll address three topics. I'll first review how floor systems work. I'll describe key policy lessons from the liquidity stress episodes of recent years. And then I'll turn to the distinction between liability-driven and asset-driven floors, and the challenges that very high levels of reserves can create. So let's begin by fixing ideas, what is a floor system, and how does it implement monetary policy. So to answer these questions, we need to consider reserves' role in settling payments, which make them a unique asset. Reserves are always and immediately liquid. Other assets may be viewed as highly liquid if market participants can usually convert them into reserves quickly and at low cost. But no other asset is ever as safe and liquid as reserves. Banks hold reserves for several reasons to settle payments throughout the day, to buffer against unexpected payment outflows above typical daily needs, and to satisfy regulations and supervisory guidance. When a bank holds reserves, it foregoes putting that money into other assets that may have higher risk-adjusted returns. Banks individually and in the aggregate, therefore, have a downward-sloping demand curve for reserves, as shown here on my first slide, represented by the blue line. As reserve levels increase shifting to the right on the graph, the spread between the returns on other assets and reserves must decrease. And when reserves reach levels further to the right that satisfy banks' demand for liquid balances, the demand curve levels off with the market interest rate close to the interest rate the central bank pays on reserves, which is shown in red. So there are two ways for the central bank to implement policy in this environment. As shown here, the central bank can provide relatively limited but potentially still quite large supply of reserves intersecting with the steep part of the demand curve. The central bank must then adjust reserve supply so the market clears at the desired interest rate. Small changes to the supply or demand for reserves can lead to significant changes in the policy rate in this regime. Now alternatively, the central bank can provide more reserves, as shown here, and vary the interest rate on reserves to influence the market interest rate. And such an operating regime is called a floor system because the interest rate on reserves is intended to create a floor below money market rates. The floor provided by interest on reserves can be soft if some money market participants are ineligible to hold interest-bearing reserve accounts or if commercial banks face costs of expanding their balance sheets. And that's why I've drawn the graph with the reserve demand curve dipping below the interest rate on reserves. To establish a firm floor, central banks may need additional tools and in the United States we use the Fed's overnight RRP facility in this regard. Now a point I think is important to make is that the level of reserves in a floor system might not be much larger than the level of reserves needed to implement policy on the steep part of the demand curve. So why is that? That's because post-GFC changes in regulations and banks, liquidity risk management, have substantially increased banks' reserve demand. So for those arguing for scarce reserve framework, we're not going back to the level of reserves of 2007. Even in this type of framework, the level needed to meet in the scarce reserve scheme might not be that far from the level in the ample reserves regime. So as I see it, floor systems have several benefits and I wanted to highlight a few. First, rate control in floor systems is more robust. Since the GFC, both the supply and demand for reserves can be subject to much larger daily shocks. For example, in the United States, the Treasury Department, other official entities and central counterparties hold accounts at the Fed that can drain reserves from the banking system. As this next chart of daily changes in the Treasury General Account illustrates, the supply reserves can swing by tens of billions of dollars in a few days because of factors such as fiscal flows. Now on the demand side, changes in liquidity regulations and risk management since the global financial crisis have made banks' demand for reserves less predictable. Controlling money market rates requires absorbing these shocks to supply in demand. In a floor system, this is a relatively straightforward exercise. By contrast, in a scarce reserve system, the central bank must actively manage the supply reserves to offset those shocks. Now second, floor systems continue to effectively control rates when the central bank expands its balance sheet to provide macroeconomic stimulus or to support financial stability. In the US, we saw this benefit most recently at the onset of the pandemic. The FOMC responded to disruptions in market functioning by purchasing treasuries and MBS at an unprecedented pace and scale. Incumulatively, the Fed's balance sheet expanded more than $2 trillion in a few months, as shown here in this next figure. At no point during this expansion did we have any difficulty keeping the Fed funds rate within the FOMC's target range. And by contrast, as shown here, when we rapidly expanded our balance sheet in 2008 in response to the GFC, the Fed funds rate shown in blue often traded significantly below the FOMC's target, shown in the gray line. So third, floor systems eliminate what would effectively be a penalty for good liquidity risk management. As I said earlier, reserves are the most liquid asset bar none. The banking stresses earlier this year prove the unique value of reserves beyond any doubt. Only reserves are immediately liquid, monetizing any other assets takes time, and sometimes it's difficult to do at all. Counter parties may offer only limited balance sheet space for borrowing against assets, while selling assets can crystallize those losses. There are, of course, good reasons for banks to hold assets other than reserves. By making loans, banks provide credit crucial for economic growth. But banks should not pay a penalty for holding reserves when that is the right choice for their liquidity management. If the interest rate paid on reserves is meaningfully below money market interest rates as occurs in a scarce reserves regime, there's a penalty to banks for managing liquidity well. Floor systems eliminate this penalty and meet banks' demand for reserves at market prices. This is the sense in which floor systems implement a version of the Freedman rule. So let me turn now to lessons from recent liquidity stress episodes. And I want to ask the question, if floor systems do such a good job of reducing cost of liquidity, why do we keep seeing liquidity stresses? My answer is that the floor system reduces liquidity risks but doesn't eliminate it. We've seen concrete evidence that the floor regime does reduce liquidity risk. For example, this next chart shows daily peak daylight overdrafts. Since the Fed moved to a floor system, peak daylight overdrafts are about one-tenth the magnitude seen in the prior regime. And interbank payments are substantially less concentrated at the end of the day because banks are less likely to need to wait to receive incoming payments before making outgoing payments. These are both signs that banks are not seeking to economize on liquidity as much as they used to, reducing the vulnerability of individual firms and the banking system more broadly. Still, recent episodes show shocks may arise that a floor system can't absorb on its own. Large shifts in the aggregate supply of demand for reserves, idiosyncratic shocks to individual banks that exceed the system's capacity to redistribute those funds immediately, and shocks to non-bank intermediaries that spill over to bank funding markets. And we saw examples of each of these in the shocks that I described earlier. So I draw a few lessons from these recent experiences, which I've highlighted here. First, floor systems should include strong backstop ceiling tools. Second, operational readiness is critical. Third, regulations should support strong liquidity risk management. And finally, some shocks are too big for any operating regime to absorb easily. And I'll discuss each of these in turn. So first, on ceiling tools. In September of 2019, as the Fed normalized its balance sheet after the GFC, reserves fell below the ample level. It was necessary to add reserves to continue operating a floor system, and that's what we did. Market data can help signal when reserves are becoming less ample, but given the uncertainties I discussed earlier, I doubt the risk of reserves falling below the ample level can ever be eliminated, especially if the central bank wants to have an efficient balance sheet with ample but not abundant reserves. So it's a practical matter. Central banks that operate floor systems must be prepared to respond to tail events when reserves fall too low, and money market pressures can arise. Now, while discretionary operations such as those we employed in 2019 can effectively address such pressures, a standing ceiling facility as a backstop makes a stronger, more transparent first line of defense. In the U.S., we've significantly strengthened our standing ceiling facilities over time. We've repeatedly revised discount window policies to make that tool more effective. In addition, we established the standing repo facility and the FEMA repo facility. But we should study ways to further strengthen our ceiling tools. Such a study could even examine how we can take on board the lessons of recent years, as well as ongoing changes in the structure of the financial system. For example, the FOMC could further consider the potential benefits of centrally clearing the standing repo facility operations to reduce our counterparties' cost of intermediating funding to the broader market. Other central banks have considered adding counterparties to their liquidity operations. These choices depend on a central bank's charter and financial system. In my mind, the focus for the Fed, given our context, should be to first enhance the effectiveness of our operations with existing counterparties. The standing repo facility being centrally cleared would be a good step in this direction. Second, operational readiness. Now, one criticism that has been leveled against floor systems is that when liquidity is ample, interbank trading may dry up and banks may become less practiced at sourcing funds. I don't see this as a reason to criticize floor systems. That would be like criticizing modern building codes for reducing the risk of fire. If a city adopts a strong building code, fewer buildings will burn. Residents will have less experience escaping a blaze. Firefighters will have less experience putting out fliers. These are good things, and while the lack of experience could have drawbacks, we know what to do about it, and that is to conduct fire drills. Similarly, the lack of daily liquidity challenges doesn't absolve participants in the financial system from remaining ready should liquidity pressures emerge. During the banking stresses earlier this year, we found that some banks had not established access to the discount window, had not pre-positioned collateral so they could borrow against it, or had not tested the borrowing procedures like a fire drill. This is unacceptable in an area when bank runs can start in minutes on social media. Ceiling tools won't work well if financial institutions aren't prepared to use them. The central bank too has a responsibility to maintain operational readiness in a floor system. The New York Fed's trading desk regularly tests open market operations to ensure readiness to respond to potential stress. But we should also consider expanding the hours of critical services such as the discount window. With the launch of instant payment services such as FedNow, liquidity in the U.S. and other markets is increasingly a 24 by 7 by 365 business. Our liquidity backstop should be available whenever banks may need it. A supervisory or regulatory expectation that depository institutions establish and test access to the discount window could help make individual firms and the financial system more resilient. And if we required banks to pre-position some amount of collateral at the window, we could reduce the risk that a bank is unable to borrow because its collateral is in the wrong place. And we should also study ways to make our discount window as strong and effective as possible in this new environment. For example, we could consider the potential benefits lending to legally eligible depository institutions purely on the basis of their collateral. In contrast, the current program of primary and secondary credit varies the terms of lending based on a borrower's financial condition. Collateral based discount lending could strengthen the ceiling by ensuring all eligible institutions have equal access to liquidity against good collateral. It could also improve operational readiness by reducing the need to take time at a critical moment to evaluate a borrower's condition. So I'll now turn to my third lesson, regulations. Strong regulations are fundamental to a financial system that serves society's needs. Regulations must also evolve alongside the structure of the financial system, the policy implementation framework, and the risks that financial institutions face. Regulations that were well suited to a scarce reserves regime and a world of relatively slow moving liquidity demands are not necessarily ideal today. For example, the Basel III leverage ratio counts all of a bank's assets, including reserves and the denominator. This rule can raise banks' costs of holding reserves and intermediating in funding markets. More broadly, it's also important to consider the interactions between regulations, ceiling tools, and liquidity risk, including buffer usability and how to account in liquidity regulations for access to the discount window and the standing repo facility. Or other central bank ceiling tools. Finally, the fourth lesson, some shocks are just too large. At the onset of the pandemic, market participants had to adapt to a dramatically different global risk environment while shifting to remote work. These developments triggered extraordinary financial market volatility, one-sided trading flows and demand for cash. While improvements in ceiling tools, operational readiness and regulations could conceivably have partially mitigated the stresses, I don't think any steps could have eliminated the stresses entirely. While improvements we should work to ensure the financial system is resilient and central bank intervention is really needed, but I think the key point is that rare does not mean never. In fact, the idea that some shocks are too large for a reasonable operating regime to absorb is built into the Freeman Rule framework. The Freeman Rule balances the benefits and costs of liquidity. It calls for providing liquidity up to the point where benefits equal costs, not for going beyond that point. So this brings me to the cost of supplying reserves and the distinction between liability-driven and asset-driven floor systems. Ordinarily, the policy rate is a central bank's monetary policy tool and the balance sheet plays a supporting role. The Freeman Rule describes the optimal quantity of reserves to supply in those normal times, the quantity of reserves that banks demand at market interest rates. Because reserves supply is determined by demand for the central bank's liabilities, this floor system for ordinary times is called a liability-driven floor. But a central bank may also use its balance sheet more actively in response to deep economic downturns or severe market dysfunction, issuing more reserves and creating an asset-driven floor system. In such a system, the amount of reserves is not governed by the Freeman Rule. The tradeoff between liquidity benefits and costs rather reserves supply in an asset-driven floor is a side effect of buying assets in response to severe macroeconomic or financial stress. The Federal Reserve is currently in an asset-driven floor system as a result of our pandemic purchases. We're running off our asset holdings to return to a liability-driven floor. And this is our second such normalization cycle following the post-GFC normalization of our balance sheet in 2018 and 2019. So why normalize our balance sheet? In my view, there are two reasons to return to an ample rather than abundant reserves. In the Freeman Rule framework, two costs of supplying reserves above the ample level. First, abundant reserves can distort the price of liquidity for non-bank market participants. And second, while acquiring assets during a severe downturn or in response to severe market dysfunction can provide much needed support for the financial system and economy, holding those assets too long can undermine the achievement of our monetary policy goals. So the first reason focuses on money markets specifically. Reserves can be held only by banks. Other financial institutions must use other instruments such as Treasury securities as their liquid assets. By acquiring non-reserve assets to back reserves, the central bank adds to the supply of liquidity for banks but may increase the cost of liquidity for non-banks. But in an efficient system, the cost of liquidity should be similar for banks and non-banks. Significant differences in liquidity costs imply that liquidity is too abundant for one type of firm or too scarce for another. So one way we can test whether the costs of liquidity are similar for banks and non-banks is to look at money market rate spreads. For example, if Treasury repo rates are significantly below interest rate on reserves, then Treasury securities, the key liquid asset for non-banks, have greater liquidity value than reserves. The Fed's balance sheet expansion in response to both the GFC and the pandemic pushed repo rates below interest on reserves. As you can see here, these spreads indicated that liquidity conditions were more abundant for banks than for non-banks. In 2018 and 2019, the gap closed as we normalized our balance sheet and repo rates eventually stabilized a few basis points above interest on reserves. Repo rates and interest on reserves shouldn't necessarily be exactly equal, but in an efficient market, the spread should not be large. This argument is the mirror image of the idea that if money market rates are above interest rate on reserves, there's an undesirable penalty on liquidity for banks. So some researchers have taken the analysis of relative liquidity costs a step further by estimating the liquidity value of long-term Treasury securities. And in such models, the liquidity value of Treasuries includes a negative term premium associated with their flight to safety qualities. The central bank's asset holdings influence that term premium, and one can attempt to calculate the level of asset holdings at which the liquidity value of Treasuries measured in this way would equal the liquidity value of reserves. But I'm not persuaded that monetary policy implementation frameworks should be designed around this term premium. To the extent that long-term government debt is scarce, the Treasury can address the scarcity by changing the maturity structure of its debt issuance. And when a central bank considers the costs and benefits of supplying more reserves, it should focus on the effects in money markets where conditions reflect the liquidity value of reserves, a unique asset that only we the central bank can provide. The second reason for balance sheet normalization is outside of money markets. Expanding the central bank balance sheet and acquiring assets makes setting and communicating the stance of monetary policy more complicated. The central bank must explain its use of two tools instead of one. And asset purchases are more difficult to calibrate because their effects are less certain. These challenges are worth surmounting when the policy rate is at the effective lower bound or when asset purchases are the best available response to the threats to financial stability. But in the absence of such needs, the benefits of higher reserves aren't commensurate with the drawbacks. Indeed, a large balance sheet can even be counterproductive to our policy goals. The Fed bought large quantities of assets during the pandemic for a reason to support the flow of credit and economic activity in response to a severe shock. Circumstances are now different. Financial markets are functioning smoothly. And today, our primary economic challenges is not a deep recession, but rather too high inflation. In this environment, it's important to remove economic stimulus by returning our assets to the level needed to supply ample, not abundant reserves. So let me conclude, central banks have vigorously debated monetary policy implementation regime for many years. And I have no doubt we'll continue to do so for many more. The Freeman Rule offers a useful guidepost in that debate. Central banks should equate the marginal benefits and costs to society of supplying additional reserves. And I believe a floor system with ample reserves satisfies that criterion. Thanks very much. Thank you very much, Laurie. Are you fine to take a few questions from the audience and also online? Yeah, so maybe we'll take three questions first from the audience. Yes, please. Alright, Wenxin Du, Columbia Business School. Thank you so much, Laurie. I think that was an excellent speech. So I find one point very interesting where you are saying if markets are efficient, we should expect the repo rates to be quite close to the interest on reserves so that liquidity value of both non-banks and banks are somewhat equalized. I think that makes a lot of sense. In practice, if the repo rate does get too close to the IOR, does that mean we're too close to the kink, right, thinking about the September 2019 experience? So is it safe that the repo rate is still somewhat below if we want to ensure that we remain in the ample and not crossing the bridge? Thank you so much. Yeah, Jeremy. Thanks. So another sort of application of Friedman Rule Logic would say holding fixed the size of the asset side of the balance sheet as to the choice between reserves and RRP, you really ought to be setting the rate on those two very close. In other words, you can almost think about as the Fed should be maximizing its profits because you want to give your liabilities to those who value them the most. So I don't know what your reaction to that is. I mean, the Fed took a step in that direction by reducing the spread between IOR and RRP. Should I think of that as a normatively good thing? Want to go further? Yeah, thanks. And one last question. Over there. Sorry, I won't be able to take all the questions. Ricardo Ries, LSE. You've talked about, very adequately, the size of the liabilities of the central bank. But of course, as we discussed that size, at some point we have to talk about the composition of the assets. Making it completely neutral would be to buy some very short-term government paper on the other side. But in so far, central banks have not done that. Or even that maybe in the new regime would be so large that there wouldn't just be enough T-bills. Or the purchase of those T-bills would lead to scarcity in the use of T-bills for collateral, as we've heard in this conference. I understand why you are talking like that and only spoke about liabilities, but could you comment a little bit on the spillover to the assets and whether there would be the assets to sustain this? Yeah, and I had actually exactly the same question, which is, when you look at the asset side, what are the trade-offs you have in terms of interference with the monetary policy stance? And I think that was also Nicaragua's question. And also the trade-offs regarding market footprint and how the presence of the central bank in financial markets and how we should think about that. And of course, this compositional issue between the lending operations and asset purchases. Well, thank you. And then thanks again for having me here. It's great to be back at the ECB. Those are all great, great questions. Maybe I'll start in reverse order in the first one on the asset side. So the discussion of the asset side of the balance sheet for the FOMC is still an open question. And as we were moving into the normalization period 2018 and 2019, it was a conversation that was never finalized. I think if you go back and read the minutes during that period, you'll see a discussion of how some of the trade-offs were being considered. If you go back to the portfolio that we had 2007 before the global financial crisis, that asset side of the portfolio was over-weighted to Treasury bills. And the reason for that is because if we needed to drain that liquidity, the bills would run off very quickly. That reason doesn't exist any longer in the floor system. So the question is what would be the driving factors to determine the composition of the balance sheet. I think some might argue for a shorter duration balance sheet just to keep the overall effect on term premium lower, where others might argue for a longer duration balance sheet in order to have some stimulative effect if you thought our star was going to be very, very low for a long time. I mean, my own perspective coming into the discussion is that taking a neutral approach to the composition of the balance sheet leaves the ultimate duration of the Treasury holdings to the U.S. Treasury Department to adjust. And over the long term, they would adjust those. And so if we make short-term decisions leaning in one direction or the other, ultimately Treasury could reverse those decisions. So I think about it at a first order of thinking about it in a neutral way. But there could be other market frictions or other situations at the moment based on the market structure that could lead you to lean in a particular direction. But I would start the conversation from a neutral one. In terms of the repo rate and IOR and getting close to the kink, I think if we go back and we look at what we were seeing in 2018 and 2019, when the repo rate got close to IORB, it was pretty stable, you know, just under IORB. And then as it got a little closer, we started to see some volatility and it moved above. That's when the volatility really started to increase. So, you know, to me, I think as we look at normalizing the balance sheet, the price signals I think are a good place to start that conversation and having repo rates close to IORB, but stable suggests to me, you know, a good normal place to be. If we started to see volatility in that repo rate, you know, just below or just at, then I would think you're getting to a position where the market is fragile and vulnerable to those potential shocks. So, I think close is an appropriate place to be. As I said, there could be some spreads, I think, but a starting position of wanting those market rates to be neutral. And then in terms of the question on overnight RRP and IORB, I mean, I think if we have those market rates, you know, close together and around IORB, we are equating banks and non-banks for the rates that they're getting. And I think the tool can be set lower. I don't think about equating the tools so much as I think are the market rates that they're receiving close together as satisfying the Friedman rule. Yeah. So, we're doing extremely well on timing and actually, even if you allow us, we'll take a little more questions because I saw that one or two in the audience that we couldn't take. Yes, please. In thinking about the backstop ceiling tool in an ample research regime, this tool is rarely going to be used. So, there is the issue of stigma. If someone goes there, then this is going to be a problem. Note the very different setup in the ECB where banks regularly have been from the very beginning borrowing from the ECB at sort of standard rates based on collateral. And so, there is no such issue as a stigma. So, do you think that this is a problem in designing your backstop ceiling tool? Well, I'm trying. Maybe Arancha just... I'm just going to ask you about FHLB, you know, because in theory it shouldn't be used as close to a discount window, but it was heavily used during the events in February. I know they put last week the papers, they're trying to caveat that, but I don't know if you have views because it shouldn't be a quasi discount window for the banks. Yeah, so it's about the articulation maybe of, you know, the liquidity that is provided by the FHLB, the central bank tools, yeah. Yes, Oder. Andy Hill, ICMA. One of the arguments you gave for normalization is that it's more difficult to execute monetary policy when you have abundant reserves. Does this raise a question about sequencing, i.e., should we consider QT before raising rates? Then one last question, Sebastian, over there. Thank you very much, Sebastian, from the ECB. To what extent do you think that extending the FETS counterparty framework, especially for the repo operations, would allow you to actually operate with a smaller balance sheet because more parties can actually tap FET funding through that way? Thank you. Just another great round of questions. The first one around stigma, you know, I think it's very important for central banks to continue to focus on the issue of stigma, if the ceiling tools are going to be effective. You know, I think one of the reasons I think it's so important that banks, all of the banks have access to the discount window and test it regularly is to further that process of usage and reducing stigma more broadly, that it is an important tool, an appropriate tool. With respect to the standing repo facility or the FEMA repo facility or the swap lines, I think those tools have been successful at being used. You know, if we look at the swap lines, which we've had in place as a ceiling tool for a longer period of time, they've been a very stabilizing ceiling tool and I don't get the sense that there's a stigma issue and there's stress in the markets. They become used quite actively. I think we should anticipate that the standing repo facility and the FEMA repo facility are seen in similar regard. My expectation is that if market rates do were to move in that direction that they would be used. And over the course of my career, I haven't heard from banks who have been reluctant to use our repo tools as we've used them. So my expectation is that they are strong ceiling tools and should be used and don't have a sense that stigma would affect that. But I do think it's an important issue and something you can think about in terms of how far as a backstop you set those tools and how often you're comfortable for them being used. And if you look at the swap lines, again, that rate's been moved around over time. Men may have some comment on that because you are the one that's running the auctions on terms of the swap lines. I think it's important that you have this history of a period where these tools were used by the entire market. And actually, for the swap line, we have that during the pandemics. You look at the numbers, it's just incredible how much they have been used. And I would say the discount window now with the episode in March, you also created such a history when this window was used widely. And so having these experiences in mind, I think for the future, would hopefully also lower the bar for accessing them in case of stress. In terms of the counter parties and the Fed's counter parties, as I said in my remarks, I think central banks are looking at their counter party frameworks in the context of lessons learned from the pandemic. And where we are. And that discussion will differ based on individual legal frameworks within different jurisdictions or in market environments. I think an important first step for the Fed in our standing repo facility is to have it centrally cleared, which I think would improve its effectiveness by reducing the intermediation costs of it being used. And I think starting with our current counter parties and making sure that tool is effective is a good place to begin before thinking about something that brings other costs and tradeoffs associated with it. In terms of the FHLBs, I know there's a lot of focus on the future of the FHLB system. And I think my comment in the context of the most current experience is just to say as this money market system is structured today, the FHLBs do serve an important part of the intermediation process. And it's a natural place for banks to go as part of that redistribution of liquidity in the system. And I think if there's reform to the FHLB system, I'm sure money markets will continue to evolve. But in the current state, the intermediation process through the FHLB system I think is an important one in allowing those redistribution of reserves. I think those are the questions that I received. There was a question on sequencing. Oh, on the sequencing of questions. So I mean, I guess I think about the experience in the QT processes is that the QT is still a tool in the background and the first tool and the most effective tool that we have the most understanding of is the policy rate. So I think it's appropriate to start with the policy rate, the tool that you know and understand and have the balance sheet in the background. And so it's starting a little bit later seems appropriate to me. But I could imagine other central banks thinking about that in a slightly different way. I think the key point for me is that it's in the background as a tool with a stance really being driven by the policy rate. We have one last question online. So how much of a coincidence do you think is that the failure of SVP happened at exactly the moment when bank reserves at the Fed dipped below $3 trillion? That seems to be a magic number. There's a lot of focus on that number. You know, I think there were specific issues related to bank management at particular firms. I think that's where I would draw my focus to those issues and wouldn't put that to the aggregate level of supplies in the system. Yeah. And of course, there's also the creation of the distribution of this amount across the system. That is also important. Great. Lori, thank you so much. I think you covered an amazing ground. And I would like to thank on behalf of everybody for being so open and also drawing on your extensive past experience to bring it to us today as a policymaker. So thank you so much. Thank you. It's been a great conference.