Banner image for 2019 Financial Markets Conference: Mapping the Financial Frontier

Policy Session 4: As We Move on from a Fed Balance Sheet Enlarged by QE, How Will Monetary Policy Be Implemented?

For the last decade, the Fed has relied on large-scale asset purchases (such as QE) and interest on reserves to provide the appropriate level of monetary accommodation. As the need for that accommodation is relaxed, how will policy be implemented, and what does that implementation mean for markets?


Julia Coronado: So this is the last panel of this wonderful conference. I want to thank the conference organizers, it's been an amazing set of panels. I've learned a lot, and I'm very honored to be a part of this. We are going to end it up with a discussion of monetary policy. We've talked about cryptocurrencies and mortgage finance, and now we're going to talk about monetary policy (because it's a Fed conference). And the title of our session is, As We Move on from a Fed Balance Sheet Enlarged by QE, How Will Monetary Policy Be Implemented?

And it's a timely time to be talking about these issues because the Fed is engaged in a review of its policy framework, in which everything is under review. And they're thinking deep thoughts and going out to the public and gathering information and input. There's a policy conference at the Chicago Fed in June, where there will be lots of academic papers and discussions about monetary policy tools and strategies and how they interact.

So this is a very timely panel, and we have just a wonderful balance of people to discuss these issues: Charlie Evans, president of the Chicago Fed, needs no introduction; and Ethan Harris, managing director and head of North American economics at Bank of America Merrill Lynch (BAML); Kevin Warsh, Shepard Family Distinguished Visiting Fellow in Economics at the Hoover Institution, and former Federal Reserve Board governor.

So we've discussed some basic questions to kind of set the stage. We'll go through some of these questions, and then we will open it up, we'll take some input from the audience as well, so start thinking about your questions.

I'm going to start with the title, and now I'm going to sit down over here, because we're supposed to do this nice and conversational style. So the title for the session, "moving on from an enlarged balance sheet," is a little bit misleading because we've actually just recently seen the Fed decide to stay with a large balance sheet. And Chair [Jerome] Powell was—or I guess we'll call him "Jay" or "Jay Pow," I've seen that on Twitter [laughter]—that he kind of answered that question, "Why did the Fed just make that decision?" in a more kind of technical way: that it's demand for reserves, that's why it's out of our control, and so in the interest of liquidity in the system we made that decision to stay with a large balance sheet.

But I'm going to ask these guys a harder question, and that is: how would you explain this to your mother [laughter]? What is the significance of the Fed having a multitrillion-dollar balance sheet forever, and a system of excess reserves? What is the significance of that? Why did you make that decision, and why does it matter?

Let's start with you, president Evans.

Charles Evans: Thank you, Julia, it's nice to be here, I appreciate that. That's a very interesting question, and I kind of thought about this a little bit ahead of time. My mother? We're talking about my mother? My mother was a dental receptionist, and I remember we would always go to the dentist that she was working for, because it was free—that type of thing. And back in the '70s, I remember sitting in the waiting room while she was picking up the phone, and she would call people—you know, "Hey, you've got an appointment tomorrow at 10:30, are you going to be there?" and that type of thing.

Now, number one: that job doesn't exist anymore, right? Nobody calls around, and does things like that, but in the '70s, I remember at some point kind of going, "Well, this must be a pretty stable type of activity, everybody goes and has their teeth cleaned." And she goes, "You know, it's actually amazing, but when the economy goes down, people stop going to the dentist on time," and things like that.

And so I don't think that my mother cares so much about the technical aspects of how the Fed provides accommodation or doesn't, but it really comes down to focusing on what are the outcomes that you're trying to achieve? And maximum employment and price stability, I think, is the most important thing. It just so happens that in a world where monetary policy is highly likely to hit the zero lower bound more often than any of us would like, we need to think about, how is that going to play out? And having a balance sheet, after you've done asset purchases, is going to be part of the proposition.

Coronado: All right, thank you. Ethan, what about you?

Ethan Harris: Well, first of all, I'd like to say to Mom: I love you, and I am eating my vegetables [laughter]. So, I think the important thing to understand for the general public is it feels like the Fed has been doing a lot of pretty crazy stuff in recent years, but ultimately it's an institution of technocrats. These big balance sheets, the zero interest-rate policy, a lot of unusual policy steps, I think, are the sign of the times. And as Charlie said, it's going to happen again, they're going to have to face another episode down the road. And I think it's important that the Fed—my main point to people is that the Fed needs to keep all its options open, because having only 240 basis points of room to cut interest rates is not anywhere near adequate for dealing with a crisis. Other central banks are also in even worse condition, so if we're going to deal with the next crisis the Fed has to have flexibility to use its balance sheet even though that carries some political risks with it as well.

So, the main thing I want to say to my mom is that everything's okay [laughter].

Kevin Warsh: So my mom was a pretty tough disciplinarian, and so what my mom would say is, "Whatever your story is, stick with it, and if you're going to change your story, change your story and explain why your story has changed." So just for the short history of QE—Charlie will remember this—so we were back in the depths of the financial crisis, we cut interest rates to zero, and we were trying to come up with ways to provide liquidity to markets. But really, the truth is we had a Sunday night coming up, Chairman [Ben] Bernanke wisely decided (after we'd rolled out four or five products) that we needed to convince markets through signaling that we're not quitting. We are going to get to the bottom of this panic, we are going to force these markets to realize we are not stopping, whether it's the fifth product that will work or the 15th.

And so we were running against the clock that week, and we needed a product before Asian markets opened on Sunday, and the idea had come that what we then called "credit easing" might not be so bad. So if secretary [Henry] Paulson is going to issue debt on Monday and Tuesday, and we bought it on Wednesday and Thursday, maybe that would be some signaling mechanism. You can imagine, at the highest levels (without naming names), there was some skepticism. Someone said, "Well, how could anyone think that that's any different? This is just two parts of the government working in concert."

But QE broadly began as an emergency response to bring liquidity to markets that did not have it. You'll recall there was a big spread between on-the-run and off-the-run Treasuries, so we were going to become a price maker—try to bring liquidity from the private sector in—and that's what turned out to be what quantitative easing was all about.

Then the crisis is ending some, and we're debating what to do. The crisis is over, and yet we continue to do successive quantitative easing—and again, then that explanation was not so much monetary policy in a crisis, that was monetary policy in normal times. How did we provide incremental monetary policy through a series of transmission mechanisms, but 10 years later we now find quantitative easing—which is largely a permanent feature—in the world's central banks? The most recent explanation is that we're keeping a large balance sheet, Julia, not for monetary policy purposes, but for purposes of regulatory policy, operational purposes. And so I think my mom would be wondering why the story keeps changing.

Coronado: Okay, fair enough. So let's talk about some of the details of this balance sheet policy. And I'm going to start with you again, Charlie. You've spoken more explicitly than many on the Committee about trade-offs between interest rate and balance sheet policy. And particularly, the remarks came in the context of an issue that is on the table right now about whether the Fed should sort of "reverse twist" its portfolio, and as securities mature off the balance sheet, reinvest in short-term bills instead of a more balanced, weighted average maturity portfolio. And the idea tends to be, the argument in favor of such a policy, is that you're going to store up duration or reduce your duration so you can extend it later when you need to ease.

So this is the whole concept of tightening to ease later, which I personally have a problem with, and you seem to, too, given your remarks. You say, "You know, there is no free lunch," you didn't use those words, I used those words, but you pointed out that if you're going to shorten the duration of your portfolio, that has implications for longer-term interest rates, and that therefore implies something about what you're going to be able to do with shorter-term interest rates. In other words, these two things work together, and so there are trade-offs that you're facing. Monetary policy is a whole package now, it's not just short-term interest rates, it's the yield curve.

Okay, so I would like to ask you, how are you assessing the costs and benefits of that particular proposal now? And then more broadly, in this review of policy, is that trade-off becoming more explicit and something that the Fed, that the Committee, discusses as it makes these decisions—these trade-offs between balance sheet and interest rate policy?

Evans: Okay, that's a very good description of where I think we are right now. Let me step back for a minute and take up something that Kevin said, which I think is important, which is, has the story really changed a lot over time, or have we learned a lot more, or do we have a new assessment? So I absolutely agree: March 2009, a huge amount of asset purchases announced, really important stress tests, fiscal policy, and we found the floor for where the economy was headed, and it was recovery after that.

So that was very important, and after that, we struggled with the fact that the recovery was slow and inflation was slow, and we found that we needed more accommodation. I think something that we've all had to come to grips with over the last 10 years is that the zero lower bound is much more likely now, and it starts from recognizing that the economy is different than it was in the mid-2000s. Trend growth rates, I would say—my best assessment—1.75 percent GDP growth. Would I like 3? I'd like 3, let's talk about policies to get there.

But if you look at expected growth in labor input, sustainable over time, half a percent—demographics are a big part of that. Productivity, is productivity going to be as low as '73–'95? I don't think so. Is it going to be as high as '95–2005, the glory years? I don't think so. Put it at like 1.25, which is pretty good and moderate, and you get 1.75 percent trend growth. That necessarily will lead to lower real interest rates, sustainably, and we've seen this around the world, they've also seen lower trend growth rates. Put that together with the 2 percent inflation objective, and you get 2.75 percent.

So, like Ethan was saying, we don't have a lot of capacity at 240 basis points, and so we find ourselves having to do more things to provide accommodation, and so asset purchases end up being a more than normal part of what you're doing. That's the story line. Is it different? If you say so, but that is the appropriate response, I think, short of other policies, like fiscal policy and other things being more stimulative (or productivity growth accelerating in ways that are unexplained).

Now, if you're going to do asset purchases going forward, then you have to deal with the fact that our balance sheet is larger—no doubt about it—the economy is larger. Cash is in demand, it's overseas, and there's sort of a natural lower bound on what the balance sheet should be, and it's way north of $80 billion (where we started in 2007). And so it's going to be higher, but does it have to have the cushion of ample reserves, plus a buffer to make sure that the New York markets' people don't have to constantly add and subtract reserves in order to hit the funds rate?

That's a choice, you could come out in different places on that, but I don't think it's a very large balance sheet. But if you don't like going up and explaining to Congress, if you don't like being criticized and all of that, you might think about, can we reserve a little more capacity for maturity extension programs like we did in 2011? And so that would be, "Let's change the duration of the balance sheet toward shorter maturities," and that way you can buy longer duration when you need it and do that.

So I think the challenge with that is, it means longer-term interest rates are going to be higher, term premia are going to be higher, because market participants have to be induced to hold that additional duration risk, and so that's going to go up. That's going to be more restrictive, that must mean that the short end—unless something else totally outside the analysis happens—is going to have to be lower. It's not obvious that they don't completely offset, so what you think you're reserving for the future you're really going to hit the zero lower bound that much more often and actually have to do this.

And so I think those are honest discussions that we have to do, do the analysis around that, and come up with a choice, and I think that's sort of where we are. But there's no free lunch, in my opinion, I agree with that.

Coronado: So there's no free lunch, you're seeing a trade-off there, different people on the Committee assess that trade-off differently, and then you'll make a decision. Do you have a time frame for thinking about that decision?

Evans: I think we have to talk about it. I think we have to have enough analysis that comes with this, the kind of analysis that we both sort of threw off as sort of "back of the envelope." Is there something about analytics dynamics that are going to be more important? Something else which, as I talk to more people about this, it's kind of like, "Something about the way you describe this suggests that monetary policy setting is going to have a more permanent effect on the economy, and as a monetary economist trained in graduate school, the idea of monetary policy having permanent effects makes me uncomfortable, so there's something about this story that I'm not quite appreciating fully." So I think there's a lot to talk about.

But having said that, to get our balance sheet down through redemptions, it's going to take a while, and so none of this is really in front of us over the next year, except that, "where do we go in a single..."

Coronado: There's no deadline to make this decision.

Evans: It's just not going to're not going to be at that size on the balance sheet for some time.

Coronado: So Kevin, let me follow on with you. How do you see the balance sheet as a policy tool? How does that, in your mind, interact with interest rate policy, and do you want to sort of counter that? Ethan's story to his mom is, "Don't worry, everything's going to be okay, and we're going to come up with the tools that we have to come up with and make the decisions we have to make to make that true." And your mom's a little bit more demanding, I guess, and wants to know, "Justify exactly why you've done this." Do you see that this is a significant decision? And again, what do you see, to stay with a large balance sheet, to stay with QE as a permanent tool? And how do you see QE and interest rate policy interacting going forward?

Warsh: Sure, so I'd say first, Charlie was a great colleague in the crisis, and we're all thankful that he's here. We're also thankful for Raphael [Bostic] for convening this session.

I might be too scarred from the crisis. I think that the Fed's job, importantly, is to mitigate against tail risks that are on the come, so that we don't find ourselves back in that period. I think the Fed has a long and proud history of interest rate policy, such that we understand how it transmits to the economy, not perfectly, but reasonably well. I think households, businesses, and financial markets understand that tool, and monetary policy, we can fight on the margins about where it should be, but I don't have huge degrees of uncertainty around it. We invented quantitative easing in the crisis, as Chairman Bernanke rightly said several years ago: the thing about QE is it works in practice, it just doesn't work in theory.

So I think we should be very clear about exactly how this tool works. There are theories from my friends at the Fed about the portfolio balance channel—which is, I'd say, a nice euphemism for the "wealth effect." The wealth effect tends to work best for the wealthy, so there are distributional effects about that. My judgment is, because we are one government, the signaling effects end up being dominant, and so if that's true, when we're in extreme circumstances and you want to trot out something that is radical—as I think we had to in the financial crisis, much to the chagrin of some of my old professors—then you want your credibility of that to be special and real, so that you can get out of the box in which you find yourself. If you're using that tool in normal operating times, I wonder whether that signaling effect is less effective.

And just let me make a final comment. We should be careful about thinking of the Federal Reserve as a general purpose agency of economic policy. We should be very clear that we're talking about QE as monetary policy, central bankers like Charlie and Raphael really get the benefit of independence of calling it the way they see it. But if we then say, "Well, QE is really about operational policy and regulatory policy, and banks need a bunch of reserves (and for some reason they can't own T-bills)"—which I'm not sure exactly why that would be—then the Fed has moved from a narrow, powerful, independent agency to something like a lot of other government agencies, and that comes with significant risks.

So in conclusion, I guess what I would say is, we do not know what trend growth is going to be. We don't know what labor force participation is going to be, nor productivity, our estimates of that, we missed on the downside for about a generation. Coming out of the crisis we thought the economy would boom, and productivity and labor would come back; that didn't happen. So it's not at all obvious to me where trend growth is, and the economy is being hit right now by a series of supply and demand shocks, so I wouldn't bet on any of that. What I would bet on is the Federal Reserve needs to be very clear when it's making monetary policy decisions and calling it the way they see it, and separating that from everything else.

Coronado: So is it fair to say then that what you're saying is, you are more skeptical of using QE as a regular, ongoing tool of monetary policy, and all else equal you would prefer to lean on the better understood tool of interest rate policy?

Warsh: Yes. I'd say first is, if we're going to use QE in the 10th year of the economic cycle, we cannot say it's not about monetary policy. That was our story for eight years, so if you have a monetary theory for QE, let's have that discussion. "And there's a good discussion," you might say. "We need more accommodation, we need it to find its way into foreign exchange markets and asset prices." But what I'm really saying, Julie, is we can't hide from the monetary policy piece. If you step back, I do think extraordinary tools are meant for extraordinary circumstances.

Coronado: So you're taking issue with the characterization of it as a technical decision?

Evans: Well, you're saying it's an extraordinary tool, and that's the part that I unfortunately think isn't the case anymore. I could only wish that it was extraordinary, it would be extraordinary if the probability of hitting the zero lower bound was about 2 percent or something, and that might be the case if short-run nominal interest rates could be 5 percent, but they're 2.75. And so I think that's what really turns this around, and it ends up being a much more common tool than we would like.

Warsh: So if you believe what Charlie said, that's a monetary policy explanation, that's fine, and then we can have a monetary policy debate. What I don't like is when we say, "This isn't actually about monetary policy, this is about regulators and the liquidity coverage ratio."

Coronado: Yes, so Ethan's struggling over there—he's got things to say [laughs].

Harris: Somehow I feel left out, here at the children's table [laughter]. But I agree that the Fed should not be actively using QE during normal times—but the Fed isn't actively using QE, it's shrinking its balance sheet. It has decided to have a big balance sheet, but as long as that's well understood, it's not—other than maybe having a small impact on the level of interest rates on an ongoing basis—it's not active monetary policy.

So I think the critical thing here for the Fed is, they need to keep as many tools as possible for the next crisis, and everything they're doing now should be aimed at in those terms, so they should be open to the idea of yield curve control in a crisis, they should be open to Operation Twist, to even buying mortgages, frankly. I agree that in normal times you don't want to be buying mortgages, you don't want to be intervening in a particular market, but if the next crisis is another housing crisis, maybe you should intervene in the mortgage market.

Coronado: Or even if it's not, what if it's just a really bad recession?

Harris: So I think that it's a little bit scary, and you know, I told my mom "everything's okay." It's okay in the sense that I think we know what we're doing here, but we also understand that we don't really have good tools for the next Great Recession. And I think it's really important to think of the international...

Coronado: Don't tell your mom that.

Harris: I won't tell her [laughter]. By the way, she has no interest in economics. But, anyways...

I think that the fact that other central banks have virtually no ammunition is extremely important. How is the Fed, on its own, going to fight the next crisis?

And finally, I would say that one of those tools the Fed needs is, I think it needs to overshoot its inflation target and get the funds rate up higher by the end of this business cycle.

Coronado: Okay, we're going to get to inflation, you're running ahead, a little bit fast.

Harris: We'll get to that later.

Evans: But I do think it's a fascinating debate about what's normal for monetary policy: what are the tools? And I keep coming back to my own realization that over the last 10 years, it is very difficult to escape the inference—I would say conclusion—that trend growth rates are likely to be less than 2 percent, not above 2 percent (and certainly not 3 percent).

Now, it is a very different story if you believe, "Yes, I think trend growth could be higher than that," and life would be a lot easier—and it would be extraordinarily constructive—if we had more commentary about, why is it that you think it's 2.5, or why is it you think it's going to be 3? And how do you think labor input growth is going to be more than that measly half a percent, given the demographics and aging? And the productivity growth that we would like, the glory years were '95 to 2005, and to think that it's going to be more than that is sort of outside our experience, and so it's just not something I would put a lot of money on. I can hope for it, for sure, but I think once you take into account where we more likely are—where the rest of the world is—it takes you into this probability that hitting the zero lower bound is much higher, and we need to really plan for that.

Coronado: So Kevin, you sort of had touched on the channels—and that's a part of the ongoing research in understanding how is it that QE works, so if it is here to stay, let's understand it the way we understand our interest rate policy tool a little bit better. There are channels we think that it works through: the portfolio balance you mentioned, signaling. So put that prism, put that hat on, those channels, and tell me, do you think that the market turmoil we saw in Q4 was related at all to the balance sheet policy, and through which channel? And I want each of you to opine on that, but Kevin, why don't you kick us off?

Warsh: So we've been running this experiment with different explanations for QE, really since the invention a decade ago, and there are a lot of lurking variables in this, because again, in monetary policy we have two predominant tools, they have two predominant tools at the Federal Reserve. It's very hard to distinguish the effect on the real side and on markets, so, with that caveat, let me say this. My judgment is that quantitative easing has a more direct effect on asset prices, my judgment is that interest rates, in the realm that we're discussing, have a more direct effect on the real economy.

This isn't a perfect separation—there's no great separation principle there—but the transmission mechanisms appear to be different. And financial markets—perhaps by coincidence, perhaps to tell us one of the transmission mechanisms—seem to be quite responsive in the U.S. and abroad, as QEs have been trotted out by our brethren central banks around the world, that financial asset prices seem to be quite interested in the overall trajectory.

So when Charlie and Raphael and their colleagues talk about QE and QT [quantitative tightening], it is hard to see that financial markets aren't following very carefully. Beyond that, to say how much of market performance is being driven by the signals they're being sent, I don't know, but markets believe, however—even though the Fed might not have a model to prove it—markets believe that when the Fed is growing the size of its balance sheet, that is signaling that risk assets are going to have some support; the Fed will be there. And we don't have to use words like "Fed put," because I don't think Charlie, Raphael, Chairman Powell are trying to put puts under markets.

But the intuition has not been lost on financial market participants, and so when they find and hear that the Federal Reserve will stop shrinking its balance sheet around September of this year, it isn't a coincidence that they consider that to be joyous news.

Coronado: Ethan, do you want to jump in and opine on Q4 developments?

Harris: So I think it's important to have the right diagnosis for what happened in the financial markets in the fourth quarter. We wrote a piece in August called "The Law of Large Numbers," and what we argued in that piece was that the markets would be overwhelmed by policy uncertainty in the fourth quarter. You had the Fed sounding hawkish with the balance sheet shrinkage, you had a couple comments from Powell I think somewhat misinterpreted as being hawkish, but you also had the threat of $100 oil due to oil sanctions, you had the Brexit crisis escalating, you had the Italian debt crisis escalating, you had the trade war escalating, and you had the government shutdown looming.

And so what happened to financial markets is I think they got overwhelmed with shocks, and there's a tendency to talk about quantitative tightening as being the dominant story of the day—I think it was a pretty small part of it. I mean, let's face it, it wasn't like this came out of the blue, every Wall Street economist, every house that does central bank balance sheet forecasting, knew exactly what was going to happen to the Fed and the central bank balance sheets. This was not a surprise at all, and the shift you had globally was from a growth of two billion in September in central bank balance sheet, to a shrinkage of two billion in October.

And so to me, while there's certainly a psychological story here, where people in the bond market put much more stock in the flow view of the world than economists and Fed people do, it's still really a pretty small part of the equity market correction. So I think we need to...quantitative tightening has risk to it, but that was a pretty small part of the story.

Coronado: So you think it was part of the larger mosaic of risks that the market was staging.

Harris: Absolutely, this was an unbelievable confluence of events. We looked at our calendar and said, "This is not good."

Coronado: I'm sure you have some views you'd like to share with us, president Evans.

Evans: I don't have a lot to add to what Ethan...Ethan described the market determination of interest rates and prices as influenced by a whole bunch of things, and certainly I think that the quantitative easing helped in reducing term premia, no doubt about it, but, gosh, we've seen a 10-year Treasury go up above 3, we've seen it come down way below that, virtually independent of what we've been doing at times.

And so there are just so many other things at work, so I tend to agree with Kevin. I put more emphasis on the signaling value from asset purchases. It's more of the "we have a job to do, we're to provide monetary financial conditions to promote maximum employment and price stability," and when we've been underrunning inflation for a long time—even if sometimes not so much, but it's been a long time—we should continue to provide the accommodation.

So I think it's signaling that we're still on the job, and doing that is very important, but all these other factors come into play, and the markets do like to fixate on one thing at a time. And the second thing that's important, we'll think about a little bit later.

Warsh: So Julie, if I could, just for a second here. So just to unite the three of us—which probably makes for a less interesting debate for you and the audience—the way I would describe this is quantitative tightening, and what markets perceived to be the Fed's rather aggressive quantitative tightening schedule—amid the nine other things that Ethan mentioned—signaled to market participants, maybe they don't get it, maybe they don't see these risks and challenges.

So the trajectory of quantitative tightening was a communication from the central bank to the world, and if it was a perfectly benign environment, maybe markets would have thought it was fine. But when markets read the rest of Ethan's report and saw those things, and the Fed acted as though things were on autopilot to markets, they're rightly saying there's a communications challenge, of which quantitative tightening and the language around it is just one means of communicating, which markets perceived to be off topic.

Coronado: Well, you very nicely framed the question I was going to pose to Charlie—no, it's kind of amazing, that was great—which is, if you do believe the signaling, is what Kevin's saying correct—that the Fed appeared to be overly rigid or unresponsive to these risks, and that contributed to the market reaction?

Evans: I suppose it's almost by definition that must be the case, because the market did what the market did, and I thought that Jay Powell, in his December press conference when it came to the question on the balance sheet, and I watched him—once I got home, that is—and I go, "Yes, that's the answer that we've heard a lot, I've given an answer just like that," and it wasn't what the markets wanted to hear.

Coronado: The standard talking points.

Evans: I think it really was the way Ethan described: there's a lot going on, and then you go back to the taper tantrum—it began so innocently, right? Chairman Bernanke's testimony before...

Coronado: Where he said, "We're going to stop this soon" [laughter].

Evans: Well, he gets the question, are you guys ever going to stop doing this? And my memory is Ben said, "You know, I can imagine the day when I've got a forecast, and if the forecast looks like the labor market's improving, one can imagine in September..." and it was a perfectly sensible answer, but that wasn't what they wanted to hear.

Coronado: Right, but even if you believe in the stock channel—so I saw you react a little bit to Ethan's stock and flow...

Evans: Well, he had it exactly right: the theory is about stocks, but the markets like flows, and...

Coronado: Let's put it through the stock prism, because the Fed ultimately recalibrated for technical reasons (or otherwise) to a balance sheet that's about a trillion—more than a trillion—dollars larger than your previous plan, so even if you believe the stock channel, that's a meaningful reconfiguration of monetary policy, or reassessment of the stance. That's an easing, and that's why markets reacted—even if you believe in the stock channel.

Evans: Right, that's right.

Coronado: So that's a fair assessment, so it could be all of these things, actually.

Evans: It undoubtedly must be, it must be.

Harris: So while we're tarring and feathering the Fed for bad communication... [laughter].

Evans: No, no, no.

Coronado: We're not doing that.

Harris: The market misinterpretations of the Fed.

Warsh: If you think it's possible for the weatherperson to get the forecast right every single day without error, then okay, I might accept this, but...

Harris: But let me finish my point. So I think that what made the Fed so much under the microscope during this period is that the Fed was the only institution in Washington that cared about the economy at the time. I mean, this're in the middle of a meltdown in the equity markets, and you're escalating a trade war, you're closing the government down. I think investors were on their knees saying, "Somebody's got to care about what's happening to us." And I think it put a lot of onus on the Fed, and so small statements kind of got blown out of proportion during that period.

Warsh: You were so close, until you said somebody has got to care about "us," the markets. It's caring about everybody in the entire economy. Yes, I agree [laughter].

Coronado: Let me add, since I think I can do that, in this financially savvy audience, I'm going to ask a more technical question about this episode, which is, we know from the minutes that you got a rather detailed briefing in December about these technical issues—basically, that demand for currency was running much stronger than you had originally projected, and that the estimated equilibrium reserves that banks needed, to just grease the wheels of the system, was higher than you previously thought.

And so you had gotten this briefing, so even if you wanted to go the technical route it was a little surprising that the Fed didn't say in that December press conference, "You know, we're facing some technical issues that may lead us to recalibrate our policy," rather than the usual "It doesn't matter," and "It's on autopilot."

How do you...I'm going to stop tarring and feathering you any minute now, but I just want to hear your thoughts on that.

Evans: So just to repeat my endorsement of Chair Powell's original answer: the Committee talked, and perhaps this is the problem of...and under a calm, tabletop discussion maybe you've got a chalkboard and you can draw up what you're trying to do, and you kind of go, "Our balance sheet, sustainable over time, should be south of where it is right now, it's going to take, still, way over a year to get there—so we're just talking about something that's going to take place in a year, or something like that."

And it could be the case that with everything going the way that it is, this might impart a bit more restrictiveness than what we would like to do, but we've got the federal funds rate up at a target of 2.25 to 2.5, and we know that we can offset that restrictiveness by adjustments in the funds rate target (and if you thought you were going to increase it, you could stop it).

And so that was a tool that was available, and so I always thought that among the most highly compensated people anywhere in the financial markets, that they would understand that [laughter], and this wouldn't be as heavy a lifting as it is.

Coronado: That's some shade that you're throwing there [laughter].

Evans: No, it's just really difficult, apparently. And so I do have quite a lot of sympathy to the idea that, would I like to be out of the asset purchasing game? I certainly would, but I think we have to have enough ammunition in the short-term policy rate—of course, we're still subject to criticism as to whether or not we're going to raise rates, so it's just part of what monetary policy criticism is about, I guess.

Coronado: You're always a lightning rod, always. So let's move on to inflation. Ethan brought it up, he made a bold assertion that the Fed needs to get inflation above 2 percent in this expansion to be credible on its target. And that is a topic in the policy review about whether the Fed should communicate a policy of what Vice Chair [Richard] Clarida has called "makeup strategies:" that could be a price level target, that could be an average inflation (expressing your target in terms of average over time).

So there are various strategies under review. You've always been somebody who's been sympathetic to getting true symmetry around the inflation target, so I guess one of the questions that comes up is, this is on the table, but what are the challenges to these strategies? What are the challenges you see to these types of strategies? Is Ethan right that you need to do this as an absolute imperative? Do you need to do it before the next recession, or is it a longer-term objective over the next business cycle? How are you thinking about the pros and cons of these strategies?

Evans: So most of the way I approach this is that, when you hit the zero lower bound you've run out of ammunition with the interest rate mechanism, and so, how can we reserve as much capacity for our normal policies that people understand better than the asset purchases and things like that. And so if you look at price level targeting, temporary price level targeting—I called it "state-contingent price level targeting" back in 2010, and all of that—some element of this involves, if we've been underrunning our inflation objective for a period of time, and if we're going to make up for those underruns, then it's just a matter of arithmetic that you're going to have to have short-term price level increases that you would calculate as inflation well above 2 percent, and it could easily be 3 percent in a year—or 4 percent, if you've got a big gap and you want to get there pretty quickly.

Now I can see by your facial expression, you kind of go, "That's a big number!" But that would be what this mechanism includes, and the reason why you would want that is so that you could have real interest rates—after you subtract off that higher inflation rate, they're negative. And then they induce lower borrowing—it's not just the asset purchase side of it, but it's lower borrowing costs, it's easier to get a mortgage at a rate that people can buy a house and things like that, and that's the monetary transmission mechanism.

But, again, you had the same kind of facial expression that I'm sure I have had, too—which is, "Well, I don't know if they're really going to believe that we're going to follow through on this." If we say we're going to pursue this policy, we doggone better actually do it, and if we are timid about that, we're going to lose credibility. And that takes me back to our existing framework, where we've said that we're here to deliver maximum employment, but symmetric inflation around 2 percent, and we've been undershooting 2 percent for so long. And there's always a story, and it's sort of temporary, until it's not (the Verizon data issue ran away after a year in the data).

But over a longer period of time, it's been under 2 percent, and if we get productivity growth that's stronger, that's going to have a depressing effect on pricing as well, and so I think it just means you really need to follow through with accommodative policies and be willing to go above 2 percent. So I take it as necessary to establish credibility on the symmetric inflation front in order to be presumed to have credibility for any other framework that we might choose as a good one.

Coronado: So last night, Chair Powell was asked about the transitory characterization that you just alluded to, and specifically, he said, "We're well aware that there are these other forces—technology, globalization, a service sector economy—these are the sorts of forces that are, in a secular fashion, depressing the run rate, if you will, of underlying inflation." And the question is, well, okay, let's take that as given that there are...above and beyond the business cycle, we've got these forces that are kind of depressing, that are headwinds to inflation, and I'll stop picking on you, I'll pick on Kevin. What is the appropriate response of monetary policy, if we take that as given that the underlying run rate has gone from maybe 2.5 to 1.5? What should the appropriate policy be in that circumstance?

Warsh: So I'm the "has been" government bureaucrat here, so these comments will sound very old-fashioned. One, to connect the prior discussion to this question, Julia—the central bank shouldn't be weathermen, we should be in the climate business, not the weather business. And so I'm not that interested in what the Fed's view is of the economy in the last six weeks. I'm quite interested in the central bank's view, what are the trends of the last year, and what are the trends of the next five years? Monetary policy works with these long, invariable lags, and the urgency to come out and opine on the latest data report from the Bureau of Labor Statistics strikes me as not the right way for the Fed to be framing things.

So that's one. Two, the debate over this inflation framework strikes me as a debate that is well to the right of the decimal point: where exactly, to the 10th or 100th, should we be on the Fed's inflation objectives, when the world is moving to the left of the decimal point? The kind of structural technological changes...our ability to measure what underlying inflation is in the economy is getting harder, not easier, and so I was really comforted by what I heard Chairman Powell say last night about there was a reasonably high bar to changing this regime. If you believe what Charlie said at the beginning, which I believe, which is, when we have the next recession, when we have the next shock, we will see—much as Ethan suggested—many fewer tools, much less conventional, unconventional ammunition to deal with it. Well, if that's true, what the central bank should be building up now is its credibility, and changing its inflation framework to something that works better in theory than what we have now strikes me as a risk to that credibility.

The truth is, back in my time we had this debate inside the Federal Reserve about price level targeting: nominal GDP targeting, hardening up the inflation target to 2.0 percent, having a range around it. And my recollection—Charlie can weigh in on this—my recollection is, we were quite divided. Chairman Bernanke really wanted this 2 percent inflation objective, and the Committee and the Board rallied around it. We very easily could have come up with an inflation objective at that time that was 1.5 percent, that was a very close second in the debate.

We can't run a perfect controls test—we can't run a counterfactual, but let's just say the Fed had adopted a 1.5 percent inflation target. Would we say today, "The Fed has hit its objectives, there's no reason to have such monetary policy accommodation"? Well, I hope not, so I wouldn't get too preoccupied on this discussion that's happening in the ivory tower, and instead I would ask ourselves the simple question that Congress assigned to the Federal Reserve, to put guys like Charlie and Raphael in the price stability business: is the change in prices negatively influencing households' and businesses' decisions about the future? Over the last 10 years, my answer to that is, emphatically no.

Because of the good work of the Federal Reserve, and some other structural factors, inflation has not interfered, the change in price level has not interfered with households' and businesses' decisions—that would be very low on the list of what they've worried about. Mission accomplished! So if you want to have more monetary policy accommodation, there might be good reasons to do it, but I certainly hope it's not because inflation isn't running at 1.65 by some governmental measure, instead of 2.00. That really speaks to a degree of precision that we do not have in statistics, and that we do not have in any kind of inflation framework.

Evans: So I don't think that's a six-week kind of discussion, or even a six-month discussion. I think that's really a five-year ahead discussion, because I come back to what you said about tail risk, which I kind of agree with. Monetary policy, a lot of it is about trying to be there to address the tail risk when it happens, that nobody else can really address very well. And during the [Alan] Greenspan years, and when things were going really pretty well, I think it's a lot easier to have the viewpoint, monetary policy is supposed to step back, get out of the way of businesses doing the right thing, pursuing good profit opportunities, and workers doing their job and things like that. But tail risks come along, and you do that. I think the real issue about not living up to what you've said is symmetric 2 percent inflation is, five years from now or whenever, when you're at the zero lower bound and you go, "I really need to be Ben Bernanke or Mario Draghi, not Jean-Claude Trichet;" they're going to wonder about where you really are. Are you the kind of person who tightened at the wrong moment?

And so I really think that it's of more importance to actually get that 2 percent—or 2.5 percent—because it's symmetric, it's not just...are you just trying to get to 2? That's why 1.6 is okay if it's temporary, because getting to 2 is okay, or is it sort of above 2? That's the longer-term perspective I have.

Harris: So I'd make two points. First, that debate over 1.5 or 2 percent as the target occurred at a time when the focus was much more on defining price stability. It was about measurement error in the way you construct the CPI [consumer price index]. And I think that over time, the debate has shifted to policy space and whether we have adequate room to deal with recessions in a world of low equilibrium real interest rates. So I think the debate is very different. If I was to go back in time, and they were foolish enough to put me on the FOMC [Federal Open Market Committee], I might have picked a bigger number than 2 percent, back in the day.

Coronado: You would've picked a bigger number?

Harris: I would've picked a bigger number, in hindsight. I want to quote a pretty well-known dude named Jay Powell here, from earlier this year. He said, "We're trying to think of ways of making that inflation target highly credible, so that inflation averages around 2 percent rather than only averaging 2 percent in good times and then averaging way less than that in bad times."

The reason this quote caught my eye when he said this to Congress is, this is the first time I've seen a pretty clear statement of the Fed's inflation goals within the context of the business cycle. And there's an explicit recognition here that recessions are nonlinear events that create very low inflation, and you can't have an average inflation target of 2 percent if you don't have an offsetting period of strength and heat in the economy. And I think that that to me, I agree that there shouldn't be a big change at the Fed to go to a higher number. I think the idea of staying with 2 makes a lot of sense.

But to actually put into the statement a recognition that there's a business cycle, and that averaging 2 percent requires that there be a period where you're above 2 percent—and I think there also needs to be a number about what's too high, because we don't want to throw the baby out with the bathwater here. So to me, this is the right direction to go. Let's start talking about inflation in the realistic world where there's a business cycle, rather than this idea that we're just randomly bouncing around 2 percent.

Coronado: Which is Charlie's five-year horizon, essentially.

Harris: Yes.

Coronado: But what is your answer, Ethan, to the question of, what if there are these secular forces that have shifted the underlying run rate a percentage point lower, which arguably is the case. What's the right policy response? A), push harder on the cyclical forces? Or B), lower your inflation target to reflect those secular forces?

Harris: Yes, I think that giving up is not an answer. If monetary authorities can't control inflation, then what business are they in? And so, yes, there are structural forces pushing down inflation. One of them is one where the Fed hopefully has some control, which is the beginnings of a dip in inflation expectations in the last three years. And so I think the Fed can't change the globalization or productivity growth or the other forces, but I think that the Fed can try to engineer...I think that actually what the Fed's doing now is just right: engineer strong markets, strong labor market, steady acceleration of wages, and hope that that last step in this phase finally kicks in at some point.

And so, yes, there's a risk in all of this of just pouring liquidity in the economy and creating a bunch of bubbles—but I don't think that...

Coronado: Which Chair Powell talked about last night, one such risk.

Harris: Yes, and it is a risk, and I think that there's got to be a little bit of balancing of that because you don't have perfect macroprudential tools. But I think this current strategy of having a sustained hot labor market, just the way I described the Fed policy, makes a lot of sense, and I'll keep fingers crossed it works, and we end up achieving average inflation in the long run here.

Coronado: All right, so before I turn to the audience questions, I have one last question, and I'm going to exempt president Evans from this one.

Evans: I'm going to the penalty box [laughter].

Coronado: Is the Fed's independence at risk, and if so, what would be the consequences of a central bank that answers more directly to, and is more responsive to, political pressures? Kevin, do you want to kick us off?

Warsh: Sure. So the Fed's independence is up to the Fed. Politicians on all sides are unafraid to offer their judgments about what Charlie and Raphael and their colleagues should do. If the Fed doesn't let in the building, it doesn't come in the building, the Fed's job is to call it the way they see it. So the Fed's independence is this incredible virtue that was earned by Chairman [Paul] Volcker and has been then passed to successive generations of central bankers, including to the Powell Fed; that's an asset that needs to accrue in value. So I've got a lot of confidence in Chairman Powell, and Raphael and Charlie, to call it the way they see it. If they get it wrong, they get it wrong, and those of us who think they get it wrong can call them out on it. But I don't doubt for a minute they're calling it the way they see it.

People come from different parties, different presidents. When they get in there, that sanctity is awfully important. What I worry about, however, is when the Fed conflates monetary policy with regulatory policy and operational policy, look around Washington, say, "Well, they can't handle this, so we're going to step into the void"—that's when the Fed goes and puts its independence at risk. If it's just another general purpose agency of the government, then I totally understand why politicians would say, "Well, if it's regulatory policy, you don't deserve any special privileges there. If what you're doing on QE is really about bank reserves, then we are in the right position to question and criticize."

So this is a view that the Fed is powerful, absent exigent circumstances, it needs to operate reasonably narrowly. It needs its independence to be confined to monetary policy, so when there is a shock like there was in 2008—and there will be another shock—the Fed has the credibility in monetary policy to be as aggressive as necessary. When that aggressive policy bleeds into other periods of time, and when the Fed asks for independence beyond monetary policy, we shouldn't be surprised that politicians of all stripes decide that they want to get their hands on the printing press.

Harris: So I have mixed feelings here. The current construction of the Fed, the people there, I have 100 percent faith in that they're not going to be politically compromised. I do worry when you've got things like GAO [Government Accountability Office] audit of actual policy decisions, as opposed to the financials of the Fed. I worry about picking appointees that clearly have no interest in serious debate on where we're going—I'm not going to mention any names, because I haven't seen anyone like that so far. But I do think we need to keep a little bit of perspective here. The history of the Fed is that it did allow political influence to affect it in the late '60s and the '70s.Part of that was because of probably misunderstandings around how inflation works and unwillingness to suffer the pain necessary to get inflation down. But there also seemed to be some successful political pressure during that period.

And then we had Paul Volcker. And when I talk to people at the Fed, they don't talk in reverence about [William McChesney] Martin and [Arthur F.] Burns and [G. William] Miller, they talk about Paul Volcker. And Paul Volcker did what had to be done, which is, if you've made a series of political mistakes and then you have to fix it, you then face the most massive political pressure in the modern history of the Fed—and he delivered. And so the simple historical message to the Fed is, don't go down that path, or else you're going to end up...the political pressure that Volcker faced was many times worse than the Fed today, and it was in part because the Fed had let itself go down a slippery slope there.

So I think that we're a long way from a compromised Fed here.

Coronado: Okay, all right. So let's go to some audience questions. We've got some exciting ones, so let's see, is that coming up on the screen? There we go. How does the decision to not unwind QE compare to MMT? Let's go there [laughter]. Doesn't it give credibility to the MMT advocates? So I'm going to presume the three of you are familiar with modern monetary theory.

Evans: Do you have a citation for that [laughter]?

Coronado: I have a few, actually. So it is a growing theory, which we could summarize that I think it gets mischaracterized a lot, but basically that there is a printing press and the government should take advantage of it as long as there's no inflation. I think that's a fair summary, so does anybody want to jump in and tackle that?

Warsh: So the only thing I'd say is, MMT is a new name for a very old temptation, and it's a temptation to conflate fiscal and monetary policy. It's a temptation to seize the Fed's independence for other purposes beyond that which was given to the Fed. And you can see the attacks on the Federal Reserve in 2019 coming from all quarters—it isn't unique to party or president—and the Fed needs to do its best job to call it the way it sees it.

The part of MMT that I find particularly intriguing is this idea that somehow "g" is greater than "r," and always will be, that is, the growth rate will always be higher than the interest rate, so we can keep outrunning the bear—that's how I interpret MMT. Well, I don't know why that is. R-star, and current views of "r," are much lower than the Fed forecasts even several years ago. I wouldn't be complacent that we think that these are the new low levels that will stand forever; and "g": my goodness, our forecast for "g," for growth rates, both outside economists and the Fed, I think have been wrong by a tremendous amount for decades. I haven't run the research perfectly, but I think in 2009 the Fed forecast that in 2010 the economy would grow at 4.5 percent. In 2010, the Fed forecast in 2011 the economy would grow at 3.9 percent. I could go on and on. I'm neither confident that I know what "g" or "r" are going to be, and if you don't really know what they are, then it's further reason to be somewhat skeptical of new adherence of MMT.

Coronado: Charlie, you have thoughts on MMT that you want to share with us?

Evans: It's hard to know how to talk about MMT because there's not a really nicely written down version of this. I've seen a lot of people describe it. Nearest I can tell it sort of goes back to the old Alvin Hansen arguments in the '30s and the '40s about inadequate aggregate demand, and there are other ways to do this—and of course, fiscal policy is certainly one way to do it.

Now if you think about fiscal policy running very strongly with a central bank that's not going to be addressing inflation, then if you counted on Congress and the president to run the deficits in order to get inflation to where it is, that would be MMT. Now, it just so happens that we're now running $1.5 trillion of deficits over the next 10 years, but nobody is calling it that—in part, that's because you've got the Fed, and we're looking at inflation and all of that, but it does turn out that inflation's underrunning this.

So I don't take it very seriously because it hasn't been delineated in a careful way, and it runs into obvious problems. And at the extreme, like Kevin is saying, you end up with some incentives that aren't very good. But I don't think of that as linked with the unwinding of QE.

Coronado: "Aren't very good," so you don't draw a connection between QE and MMT?

Evans: We've got this on our balance sheet, and it's the size of our balance sheet, but it's not necessarily delivering deficits.

Coronado: Correct, so there is no explicit, fiscal monetary coordination—QE is a decision that has been made independently of fiscal policy.

Harris: So we titled our weekly, "Mainly Muddled Thinking" [laughter].

Coronado: That sounds dismissive [laughter].

Harris: No, I think it's a great idea. The basic problem with it, is it seems to me—again, getting back to what Charlie said, you're not actually sure what it is, but it feels to me like it's a policy that makes sense in extraordinary times, like 2009 in the Great Depression when aggregate demand was massively low and so you just throw open the spigots to get things back on track. But the idea of implementing it now, at a hot phase of the business cycle, is like throwing gasoline on the fire.

So the other thing I would point out, though, in fairness to the advocates of it—and I will offend one side of the room with one comment, and the other side with the other comment—when politicians get a hold of these theories, like MMT (which is kind of like Keynesianism on steroids or supply-side economics where tax cuts pay for themselves), it becomes an excuse to do irresponsible policy, and the rest of us sit around trying to figure out, "Well, in the margin, what's correct about that theory? And, in what circumstances might it make good sense?"

But basically, MMT is the liberal version of exaggerated supply-side economics, they both give you carte blanche to do whatever you want.

Coronado: Interesting, thank you.

Harris: So how many people did I offend, there [laughter]?

Coronado: Are there any MMT diehards in the room? It got a lot of votes, it got the most votes of any of the questions, so...I'm just saying.

Harris: Julie, you should admit that you are.

Coronado: I'm what?

Harris: You're an MMTer.

Coronado: I am not an MMTer! No, I am not! I think it's important to understand theories that get the kind of political traction MMT is getting, and where I come to it is, it does acknowledge that the availability of money printing in times of weak aggregate demand more explicitly than our macro models do, and that's fair.

But to me, one of the biggest risks is when—as Charlie said—you're relying on the political authorities to monitor inflation, which seems like a bad idea in this political environment, and two, it does not fully, in my view, incorporate the risk that financial markets might ultimately call you on it (and I think you kind of alluded to that). It's not new, lots of governments have tried to use the printing press, and if you're not a reserve currency, that often runs into big troubles—you can look at Turkey or Argentina now, and so that's the risk, obviously.

Harris: Okay, sorry, I take it back. You're a supply-sider [laughter].

Coronado: I'm not a supply-sider! All right, we're moving on. This is a great question, and brings with it other questions, which is, will the Fed consider using negative interest rate policy; if not, why not? And we can throw in other tools that other central banks have used that the Fed has chosen not to, such as yield curve targeting—Ethan, you mentioned that earlier, the Bank of Japan has done that as a shift away from QE to yield curve targeting.

But let's start with negative interest rate policy, and then you can add any other things that you think are intriguing. Will the Fed do negative interest rates, Charlie?

Evans: Do you always start here? Why don't you start don't here [laughter]? You should mix it up a little bit.

Coronado: I have been!

Harris: And are you about to do it now? That would be market-moving comment there.

Evans: From me [laughter]? I mean, if I said that would it really have any effect? Come on.

Harris: Okay, you're right.

Evans: You know, I don't really sense much of any discussion of this particular policy, till now. I have not seen the papers that are going to be presented and discussed at the Fed conference, and I'm pretty sure there is one on different tools. And it could be that the authors considered this and show some evidence, and how they work, but other than that there clearly wasn't anything that was considered during your day, right, Kevin? During the crisis period, with all the liquidity programs [laughter]?

Coronado: Nice way to hand that off.

Warsh: So Charlie just phoned a friend [laughter].

Evans: Happy to do it.

Warsh: So it was considered. I haven't read about it in most of the memoirs that have been written of the period. But you'll recall that the language we used in the darkest days was the "zero lower bound." We didn't call it then the "effective lower bound," we said the "zero lower bound." Because my recollection is, we did ask ourselves—this is before QE had become the obvious thing for central bankers to do at the zero lower bound, and we didn't know whether QE would be perceived as a bit of an "in one pocket, and out the other" with the Treasury and the central bank in coordination, so it is a question, I've got four votes today in this panel—I think it got more than four votes back in the day as something we needed to pursue as a possible option.

I thought then it was a bad idea, I think now it is a bad idea, for the U.S. and for the Federal Reserve. Remember, we want monetary policy to transmit itself through the financial markets—and more importantly, through the traditional banking system—so it gets to the real economy. The real economy is not a subject we've talked about most in the last hour, but it should be. And I think that negative interest rate policy does significant harm to that transmission mechanism, does significant harm to the banks that stand between the central bank and the real economy—so it's among the reasons why I think we were right to dismiss it more than a decade ago, and we'd be right to dismiss it now.

I'll just say one other thing. Had the Federal Reserve followed its original exit principles when it adopted QE, I don't think these sorts of questions would be so much in the news, either in academia or among financial markets.

Coronado: What do you mean?

Warsh: Our original exit principles, which the Fed changed in June 2011, but before that, this was our theory. We would cut interest rates to zero—and we cut them pretty quickly, but Ethan and others might say we didn't cut them quickly enough; I think that's a fair criticism. We cut them to zero, we are going to grow the balance sheet. The original exit principles were, "And then when we can, what we will do is shrink the balance sheet and then raise interest rates."

The Fed, I think several times between those darkest days when we announced our extraordinary measures and announced the exit principles, soon thereafter have changed those. Had we adopted that, I think we would have gotten to a smaller balance sheet. Interest rates might not have gone up at the period in which they did, but I think then these sorts of questions wouldn't be germane and the broad conduct of monetary policy would be different.

But to be fair to Charlie, the Fed did change these exit principles after my departure.

Evans: That's a counterfactual that runs against the data developments, the reason why in June 2011, when the Fed published a strategy for exit principles. And then a month later, the data showed us that the recovery was not on track and inflation was lower and we ended up not wanting to do more asset purchases in 2011 (we did Operation Twist instead, in 2012). I don't think it was really according to the data, which you've got to look at, that it was ever really on the table to shrink the balance sheet as quickly as might have been—probably was—contemplated in 2009. It just wasn't really in anybody's thinking that we would have been going through this for as long as we were.

Warsh: Right. My only point, though, Charlie, is that the use of quantitative easing in normal times begs these sorts of questions. When the tail risks happen, what will the Fed do then? What will be the surprise? What will the Fed do to pull a rabbit out of its hat?

Coronado: Are you saying that there's a perceived cap on how much QE can be done? Is that how you're drawing the link between that?

Warsh: So I think that QE has been normalized, QE is not just those things you do in exigent circumstances, QE is now part of a more permanent tool kit. So if you find yourself in a period of panic, will using normal tools in extraordinary times be sufficient? It wasn't in 2008, I don't know if it will be the next time.

Coronado: So in a sense, there's like a psychological element you're suggesting, like if QE is perceived to be normal and you need to send a strong signal that you're doing extra-normal things, you need another tool.

Warsh: So if you believe what I do, which is signaling is really important, that signaling was essential to Mario Draghi's "Whatever it takes," signaling was really important to fight back the panic that we had now. I think these questions are germane in ways they wouldn't otherwise be.

Evans: But I think we've been pretty clear in the way we've talked about it, if not written down exactly, which is there is a distinct preference for using the interest rate mechanism to provide accommodation, so the virtue of the December rate hike—because the times allowed it, the economy's been doing pretty well, and since that time that's been shown to be a better decision than not. We're at 2.25 to 2.5, and so we can cut; we don't have to do QE. QE, I think, is very much "as you're getting closer to the zero lower bound, you might be thinking about warming that up."

Now, it is the case that we are running down the size of the balance sheet, and so if you tried apparently to continue to do that adjustment while you're lowering rates we have learned that the markets are kind of going, "Please explain this a lot more." So I can imagine that aspect of the balance sheet coming into play. I mean, if you want to call that QE, okay, but I think that's just sort of...I think that's where we are.

Coronado: Well, so let's transition to the next one, which is the next tool in the tool kit. What's your view on yield curve targeting? Is that something that's under review or will be included in the review of tools? And you want me to start with Kevin, I assume?

Evans: Ethan's the expert on this [laughter].

Coronado: We'll start with Ethan.

Harris: Well, I think on the list of things this is higher on the list than negative rates. Negative rates, there's a definite cost-benefit concern, and so the results have been very mixed—partly maybe the way they were introduced, but it's less clear. I don't think we should be ruling things out. I think that these things should be on the back burner.

Yield curve targeting, to me, makes sense. I think it's worked for Japan, I think that it helped insulate them from movements and global bond yields. It hasn't worked in the sense that Japan hasn't been able to end its low inflation problem, but they've got a red-hot labor market, they're in pretty desperate shape, and this seemed like a logical next step. And so I would think that the Fed and the ECB [European Central Bank] should both be considering this as part of the tool kit.

Coronado: Kevin, maybe you can speak to the concerns that were tossed around in the room when QE was initiated. Certainly one of them was that you didn't know how big the balance sheet would have to get, right?

Warsh: Right, so this might sound like a new question, in light of what the Bank of Japan has done with, arguably, some success in recent years, it's not a new question for those of us that were scarred by the last crisis. Again, as we're thinking—and amid great panic in the real economy and financial markets—we might not have called it "yield curve control" or "yield curve targeting." But the question we asked in 2008 and '09 was, if we want to set, suppress, keep down long-term interest rates, and we said, "We expect that the 10-year Treasury yield shall be X"—we had a healthy debate in Chairman Bernanke's office—does that mean we'd have to buy everything, or nothing? Would our balance sheet not need to grow by $1 because we would state that that's what we're solving for, and we will come into markets and buy anything it takes in order to achieve that? Or will our announcement alone get that result?

In that case, there would be no increase—or virtually no increase—in the central bank balance sheet, it would be set by financial markets. The other side of the debate, again, as I recall more than a decade ago, is another one of our colleagues said, "But really, this is open-ended QE, what if they don't believe us?" The world is being shook by its foundations, we don't know what our next move would be if we were to commit to X hundreds of billions of dollars of purchases—at least we know, around the Federal Reserve Board and FOMC, what policies we've agreed to—would this suggest that QE would effectively be unbounded, because markets would say, "Well, let's just see about that"?

So again, it might seem like a new question. It doesn't for those of us that were in the war, and we decided then that that was suboptimal—but like Ethan, I would say I would rank this, for a future crisis, considerably higher than I would negative interest rates in the context of U.S. policy.

Coronado: Can I just follow on to that, to your answer. So in the event—and actually, for both of you—in the event you did the QE, you did the fixed size (because, as you say, that felt like the right balance of uncertainties you were facing). But you did repeated, and you ended up doing more than one QE, because the fixed size wasn't enough—and then we ended up having these market rumbles around the ends of QE, because there was an anticipation of that going away and so there was some market tightening. And then the Fed, the third iteration was open-ended until we achieve our objectives—which is more along the lines of yield curve targeting.

So what is best practices on QE? If you say, "Not yield curve targeting is best practices on QE the next time it's used," open-ended QE or fixed-size QE?

Warsh: Right, so I actually wasn't there for all the QEs, coincidentally enough [laughter]. I was there for the first one, I think the right strategy is to first, to announce it. It is best to say how you're going to deal with tail risks when you are in peace times, and you can have the robust deliberation that I'm sure the Federal Reserve has under Chairman Powell, and ask yourself these hard questions—do the work, so that you're not making it up during the course of the week as we did last time. Be very rigorous and self-critical about what's worked in QE and what hasn't, who have been the beneficiaries and who haven't, and how does that differ from our models? Instead of somehow suggesting, "Well, it's all worked perfectly according to plan"—because I don't remember any of these plans.

So again, if you're a climatologist and not a weatherman, these are the questions that you should devote considerable time to, as opposed to opining on how the economy looks six weeks from now.

Coronado: So Charlie, do you have a view on yield curve targeting, and/or best practices of QE?

Evans: I do think that the open-ended QE3 was a very good way to go. I think it was successful, I think combined with the threshold forward guidance that we had, it also tamped down any thoughts that we were going to raise interest rates too quickly.

Coronado: Some might even call that the "Evans rule."

Evans: Thank you, you're kind [laughter]. There are a lot of details in all of these, negative interest rates, I should have said, "You know, the mechanism makes sense, it's that banks ought to be incented to get rid of their reserves and lend things out at attractive rates of interest so that borrowers can do more." And so the mechanism is sound, but I don't know that it plays out so well. I have a lot of questions about that.

But we can talk about a lot of different issues, but I keep coming back to what we were talking about: portfolio balance versus signaling, is it the term premia or is it the getting up and having the credibility when the Fed chair or the ECB president says, "We're going to be here, we're going to do what it takes, we're going to keep at it, because we're focusing on the outcomes. Our outcomes are to deliver maximum employment and price stability." And if we're perceived as kind of going, "Yes, but I just don't like doing that, so..." and things like that.

So I think however you can convey the most confidence—and you have the most confidence in the tools that you're using—is most important, and so I don't know enough of the particulars on the yield curve targeting. Kevin laid out a number of issues that make me nervous, for sure, but once you make a decision, you know if you're going to storm the beaches of Normandy, I think you have to go—and keep at it.

Coronado: So it's something that's worth considering, and looking at experiences and pros and cons.

Evans: Need a plan.

Coronado: Yes, okay. Excellent. Ethan, do you want to opine on best practices on QE, do you think "best practices" is open-ended?

Harris: No, I one hundred percent agree, I think it's about always having another step. And the exact sequencing is less important than the determination and the...because I think the big part of...

Coronado: The message of "by any means necessary."

Harris: Yes, "by any means necessary." And there is a big difference between the Fed's success—which I thought was the first major central bank to really go down that route—and what was going on with Trichet, and then you get Draghi and [Haruhiko] Kuroda come in, there's a good reason why they had more success in moving the markets, because they presented a "do what it takes" approach. And I think that's the key, and that's why I don't like the idea of preannouncing that these three tools are not in the tool kit. I'd rather figure that out, if I ever have to get to that awful moment where I have to decide whether to do negative rates or not, let's at least have it as an option even if we end up not using it.

Coronado: So you're actually...that's a little bit in line with what Kevin was suggesting, that there's a sort of psychological element that comes in here, that's important that you have a surprise tool to come in with and use. Is that what you're saying, in a way?

Harris: Well, and you saw that, in some of the announcements out of the Fed, the ECB, and the BOJ [Bank of Japan], some of them worked great and others didn't. Their announcements were, you basically told the, there are several...I forget the exact instance, but in Europe and in Japan, where basically the message was, "We're not going to do this, we're not going to do this," and then they did it. So when they first did QE in Japan, "We're not going to do it, it won't work; we're not going to do it, it won't work"—and then they did it! And you're like, "Well, what's the message you're sending to people there?" You're saying that this is a desperate measure, and this is a sign of...

Coronado: And we don't know what we're doing [laughter].

Harris: Right. So you have to...I agree with this idea that you have to have thought this out in advance, and if you're going to go for it, go full force.

Coronado: This is a more technical question, but I'm going to put it up here because it is one that is asked a lot, and I'll start with you, Ethan. Why is the interest on excess reserves, why does it continue to hit the upper bound of the effective fed funds rate—or, I guess that's vice versa, right? The effective fed funds rate keeps hitting the upper bound on the interest on excess reserves. Is that giving us a signal about liquidity that's too tight? Is that a monetary policy signal? What do you think, Ethan, is going on in funding markets?

Harris: So don't tell our rate strategist, but these kinds of questions make my eyes glaze over [laughter], because I've been avoiding dealing with the weird stuff going on in the money market. I don't think there's any big story here, I think it's technical issues in the markets, and maybe the Fed needs to add a new facility to deal with this.

Coronado: A standing repo facility, perhaps?

Harris: Yes, but it's not something I worry a lot about, to be honest with you. But don't tell our rate strategist that.

Coronado: I won't tell your rate strategist, are we being televised [laughs]?

Do you want to add to that, Charlie, at all, do you want to opine on this? Do you think that this is telling you something about the sort of equilibrium size of the balance sheet, or is it more of just a technical issue as funding markets kind of evolve to a new postcrisis structure?

Evans: I think there is—part of that is the evolution. Over a longer period of time we expect IOER [interest on excess reserves] rates to be below the funds rate, and so we seem to be playing through that here. Obviously, if we saw more volatility in keeping the federal funds rate within our target range then that could be a symptom that we're approaching the upward-sloping part of the reserve demand. So that would be the balance sheet size it would be, as efficient as we could achieve, and then there's a question as to how much volatility do you want to show through into short-term money markets? That would take the size of the balance sheet away from just, "Are you making life easier for financial players, or political discussions, or whatnot?" But if you're willing to entertain 25 basis point of volatility or something—that's a choice that somebody might make, but we're kind of looking at that and going, "Well, we don't like that."

Now, maybe a repo facility takes care of that, I guess if you don't have to add to the balance sheet day after day after day. I mean, that would just be a different way to have a bigger balance sheet, or whatever.

And so those are some of the issues, and I'm a little more like Ethan in terms of the eyes glazing over, so I take great comfort in the fact that we have a lot of experts in the Federal Reserve System who pay attention to things like this.

Coronado: But in your answer you are pointing out that there is both a plumbing aspect to this, and a bit of a policy aspect—not in terms of the current setting of monetary policy, but what is the appropriate role of the Fed to play in funding markets, right? So if you do a standing repo facility, there is both a sort of technical influence of that, in terms of what it does to funding conditions, but it's also, the Fed has decided to play a certain role in funding markets on a sort of permanent basis.

Evans: But I think there's a monetary policy reason for that. If through our decision on the size of the balance sheet we impart more volatility on short-term money markets, that's going to find its way into the pricing for auto loans, mortgages, and things like that. So if we're shooting ourselves in the foot there—I mean, there ought to be a pretty clear value proposition for why that's a good thing to do, if that's the choice that's made.

So it's still monetary policy, but it gets close to other issues, I certainly take that.

Warsh: So just a couple things. First, good corporate finance theory, in practice, says there should be one risk-free asset—there shouldn't be two competing with each other. The Fed has created a new risk-free asset that competes imperfectly with Treasury bills. If banks want to own risk-less, short-term assets, there are T-bills out there and the Treasury can solve for that demand, but now, again, with fiscal and monetary policy sort of—how do we say this kindly?—at the edge of one another, you have two risk-free assets that compete. You also have a very complicated industry structure that we've built up.

If the Fed is in the business of having its credibility announcing a target, my recommendation is they should hit the target. If the U.S. government is in the business of preserving the dollar as the world's reserve currency, having a single risk-free asset, then this line is getting quite uncomfortable, so the technical changes the Fed has had to make I think are indicative of these broader problems.

Coronado: So you might be a little bit more skeptical of a standing repo facility, just in terms of complicating the message of markets and the role of Treasuries.

Warsh: Credibility here matters a lot, and the more complex these facilities, the more this is a little world in which people are trying to justify Ptolemy instead of going to Copernicus.

Evans: In the old days, the New York desk, there's morning call, and they'd have to talk about how many reserves have to be added in order to make the Who light dim [laughter]. That's right, but this is the kind of issue that we're talking about, what are the actions that the New York desk ought to take in order to limit the volatility or hit the funds rate target? And so we're just caught up at the policy level in these details, and it's not really that different except that the market is so much bigger now, it's a choice.

Coronado: So we're going to end with a fun question.

Evans: Okay, good, finally [laughter].

Coronado: Do you ever wish that the dot plots would just go away [laughter]?

Harris: What's this about pot, here [laughter]?

Coronado: Dot plots, dot plot, dot plot [laughter]. With the market's sometimes extreme response to the dot plots, at times, do you ever wish the dot plots would just go away?

Ethan, do you wish the dot plots would just go away?

Harris: No, because it's a jobs creator in Wall Street [laughter]. Somebody has got to sit down and put a name on top of each dot, so...

Coronado: Oh, yes, guilty as charged!

Harris: I think that it kind of feels like we're in between two potential regimes: one is to not be as detailed in the information we're giving, and the other would be to actually connect the dots to the forecasts—and call Charlie Evans number one and Jay Powell number two, or whatever [laughter]. So it feels like the current—that's the way Charlie wants it—but it feels like the current system's kind of a bit of a hybrid, it's kind of halfway into communicating what they really are thinking.

But alternatively, the Fed has a problem in that putting out all this information seems to create weird sunspot developments in the markets, where people get obsessed about some little detail.

Coronado: How many "zero dotters" [laughs]?

Harris: Yes.

Warsh: I'll just add, full disclosure, I didn't want the dot plots when I was there, it was among the battles I lost. The dot plots, to say they've outstayed their usefulness suggests that they once had this utility. But listen, my judgment is, when the Fed is putting dots on a piece of paper, that has two very bad implications. One, the central bank needs the world's businesses and economists and market makers to have their own independent judgments of what's going on. When the world sees the series of dots from Fed participants, FOMC participants, they've had a tendency to supplant the independent judgment of others. What will GDP be next year? What will interest rates do? What will the unemployment rate be?

Monetary policy works best when there's an independent source of information that confronts Charlie and Raphael and Chairman Powell, and they say, "Huh, I wonder what they're seeing that we're not?" I don't like it that the dots are supplanting those that should otherwise be independent views, and I also don't like it for another reason: the Fed meets quite frequently now, and speaks even more frequently, and the first question they get asked and have to justify is, "Tell me about your dot, tell me about your forecast for GDP this year."

Monetary policy works with long and variable lags, and I don't really care what their dot is. I think it leads to some anchoring, which is a human condition—"That was my dot, I said it last time, I better not change it." So I want people to be able to respond to incoming information, to think broadly about the future and not be preoccupied with their batting average going into the All-Star break.

Coronado: Do you want to launch a defense of the dot plots?

Evans: Well, I guess sometimes it sort of feels like I'm downstairs as a little kid, and mom and dad are upstairs arguing, and I wish somebody would get them to stop.

Coronado: Aww [laughter]...

Evans: It's just too loud, and all of that. The dots are mommy and daddy, so when there's a dispersion of the dots, then there's a disagreement over what the appropriate monetary policy is over the next year or two years. And I've heard people say, "Why do you guys have such a dispersion?" It's because we're having a serious discussion about what's the nature of the economy, how is it evolving, why do you think trend growth is 1.75, I'm way more bullish and optimistic—and that has implications for your forecasts.

And it just makes no sense, in my mind, my judgment, to display forecasts for GDP growth, unemployment, and inflation, and not say, "Oh, and by the way, I think I really needed accommodative policy to get my inflation forecast to look like my good friend and former colleague, Charlie Plosser, who had the same inflation forecast but, boy, he had something a lot tighter"—it just isn't coherent. So the way that we chose the Committee to display this is with the dots, it's not perfect, but it shows a lot of mommy and daddy disagreeing [laughter]—ultimately for a good cause, I hope.

Coronado: So then it sounds like you might be sympathetic to one of the options on the table, which is connecting the dots to the forecast so we can distinguish between, say, a Charlie...

Evans: I don't have a problem with that. I don't, but a lot of other people are less comfortable revealing their forecast and whatnot. But I kind of get that.

Coronado: All right, okay. Well, we are just about on time here, so, does anybody want to add any concluding thoughts?

Warsh: Raphael, thank you very much, it's been a great conference.

Evans: It has been a very good conference.

Coronado: Yes, likewise, join me in thanking these guys.