2023 Financial Markets Conference - Old Challenges in New Clothes: Outfitting Finance, Technology, and Regulation for the Mid-2020s

Policy Session 1: Monetary Policy amid Higher Inflation

Inflation began a dramatic rise in the spring of 2021 and has persisted at extremely high levels, causing the Federal Reserve and many central banks around the world to abruptly shift monetary policy stances from highly expansionary toward potentially quite restrictive. Moderator William English and panelists Troy Davig, Kristin Forbes, Jan Hatzius, and John Taylor discuss why many central bankers miscalculated the risk of persistently high inflation coming out of the COVID-19 pandemic and the tools and strategies that will be required to regain price stability.

Transcript

William English: My name is Bill English. I'm a professor in the practice at Yale and very pleased to be here. As with everyone else, I thank the organizers for including me in this interesting panel. I thought in my role as moderator, since I'm not going to be able to control this mob later on, I could start by providing a little bit of background as sort of a conversation starter for this panel.

Let me go back to when the pandemic hit in the first half of 2020. There were huge disruptions to the economy. These were a surprise. We hadn't seen anything like this for a century, and output fell sharply, unemployment soared—and perhaps reflecting the experience with the financial crisis, in the post-financial crisis period, policymakers responded because they saw a recovery that they feared would be very weak for a very long time. Growth would be slow, unemployment would be high, inflation would be low for a long time. If you go back and you look at the summary of economic projections at the time, that was what the committee thought.

They responded by cutting rates to the lower bound and by beginning a large asset purchase program and using some very strong forward guidance. I'm sure we'll talk about these things later on. That's all in addition to the emergency programs that were put in place to try to soften the blow in the spring of 2020. But the recovery came faster, sooner, stronger than had been expected—a big rebound in output and employment in the second half. That was welcome, but an unwelcome spike higher in inflation in the first half of 2021 as the recovery in demand ran up against supply that was constrained by the pandemic in a variety of ways, and significant labor market frictions as a result of the pandemic as well.

Initially, that high inflation seemed like it was fairly narrowly focused. We all remember worrying about chips, and auto production, and used cars, and so on. Trimmed mean measures of inflation didn't rise for a while, for example, and the Fed initially looked through the higher inflation. But that inflation then broadened out, and with the continuing robust recovery policymakers began to turn towards policy normalization.

They slowed the pace of purchases in the fall, they moved up the anticipated timing of liftoff, and then in early 2022 the Fed wound down asset purchases, began raising rates in March, began balance sheet normalization in June. But inflation had been given an additional kick by the war in Ukraine, and policymakers responded to conditions by tightening policy rapidly. The target range for the funds rate went up 5 percentage points in about a year, the most rapid increase we've seen since the Volcker disinflation in the early '80s.

Despite that rapid tightening, the economy has continued to expand, and employment gains have remained solid. Inflation has fallen some, but it remains worryingly high. It's expected to fall some more because of the lags in the calculation of shelter inflation, but still the persisting strength in demand and inflation raises the issue of whether policy effects are still to come (we're just seeing long lags), or there's more to be done. Indeed, had it not been for the banking problems this spring, I suspect the Fed would have almost surely had more to do with monetary policy.

But with a lot of banks pulling back from lending and the outlook for the economy and inflation highly uncertain, the Fed has suggested it might pause at some point before too long to gather more information and decide what's the appropriate course of policy going forward. Given all of this background, our panel this morning seems particularly apt. Our title is How to Conduct Monetary Policy amid Higher Inflation and a Nonbinding Zero (Effective) Lower Bound?

The broad questions that we're looking at this morning are, first: why did so many central bankers miscalculate the risks of higher inflation? This wasn't just an American problem, as Lord King said in his remarks. Many central bankers around the world had similar problems, and many central bankers around the world have turned to a fairly rapid tightening of policy over the past year or so.

Second, what are the tools and strategies that will be needed to regain price stability? What do we do now? How do we reverse course and get the inflation that's happened back in the box, back to target? To discuss these questions, the organizers have put together a terrific panel. There's me, but also four economists who don't really need an introduction, so I'm going to be very brief.

First, we have Troy Davig, who I knew as director of research at the Kansas City Fed for a long time when I was at the Board, but who subsequently became chief US economist at Rokos Capital Management and is now chief US economist at Symmetry Investments.

Second, Kristin Forbes, the Jerome and Dorothy Lemelson Professor of Management and Global Economics at the Sloan School of Management at MIT, a former external member of the monetary policy committee at the Bank of England, and a former member of the White House's Council of Economic Advisers.

Third, Jan Hatzius, the chief economist and head of Global Investment Research at Goldman Sachs, where he has had a long and illustrious career including as the number one ranked global economist in the annual Institutional Investor Global Fixed-Income rankings.

Finally, John Taylor, the Mary and Robert Raymond Professor of Economics at Stanford University and the George P. Shultz Senior Fellow in Economics at the Hoover Institution. John was also a member of the White House's Council of Economic Advisers and served as Undersecretary of the Treasury for International Affairs, among a host of other positions and awards.

Thanks to all four of you for your participation today. The rules of the game are clear, I hope. You each get about 10 minutes for some opening remarks. I realize that means that you get 11 or 12 minutes; but after about 12 minutes, I will drag you from the stage. [laughter] Please, do keep things brief. We want to have plenty of time for discussion.

After those introductory remarks, there'll be some time for discussion among the panelists. I'll ask each of the panelists to comment, if they choose, on what the other panelists said. I'll probe, I'll poke with a stick, to try to get some interesting discussion going on. Then I'll open to the floor, so please use the app, send in your questions, and I will, I'm sure, get too many questions to cover. But I will use the list of questions I get from the app, and then if there's time I'll give the panelists a few moments at the end to summarize what they think they've learned.

I'm going to go in alphabetical order: Troy, Kristin, Jan, John. Troy, you're up, please.

Troy Davig: Great; thanks, Bill. I'd also like to thank President Bostic and the organizers at the Federal Reserve Bank of Atlanta for the invitation, and for putting together what looks like a great program. Also, just a quick disclaimer: these are, of course, my views, and do not necessarily reflect those of my employer, Symmetry Investments.

I'm going to start here talking about the banking situation, and the real starting point is whether there's potential now for this to move to a more severe financial crisis that pulls in a larger set of banks. If we look at this first chart, it shows the share of uninsured deposits to total deposits on the vertical axis, plotted against the size of the loan book and security portfolio for US banks with over $50 billion in assets. Basically, as you move to the upper right quadrant, it's more risky; lower left, less risky. The three red dots in the upper right—those are the banks we all know; they don't need to be named. As you can see, they were outliers, both in terms of the amounts of uninsured deposits and the size of their loan and bond portfolios.

In general, what you see here is you do see this negative relationship between these variables, and you would expect that under a well-regulated and sound banking system. What is a little bit concerning is, there's a few banks between the upper-right failure zone and the banks that lie on that dotted line. There is risk that there could be a few more dominoes to fall, but we can assume that we're not going to go...I don't want to be too optimistic, but we will most likely—famous last words, but—avoid a broad-based, systemic crisis.

I don't think that means, though, that the banking system won't face still a number of challenges, whether it's a few more isolated bank failures or, more broadly, ongoing deposit outflow from the banking sector. This deposit outflow is going to put pressure on credit standards and loan growth. If you look at last year, for example, it's pretty remarkable. There was about $1.3 trillion in new loans issued by the banking sector last year. On a nominal basis, that's the most ever. We know that banks more likely service smaller firms, and that's a big reason we did see such a surge in small business activity.

These are job openings broken out by firm size, and you can see firms with fewer than 50 employees account for the vast majority of new job openings. With some of the challenges that the banks will be facing, just in terms of the ability to extend credit, we're going to see more of what we're seeing here, which is a decline in job openings from smaller firms.

In addition, for banks there's a lot of concern about CRE, but I want to pivot a bit about a more immediate concern, which is that I worry a bit about competition for liquidity to banks from the US Treasury and the Fed's overnight reverse repo facility. The ON RRP [Overnight Reverse Repo Facility], shown here in orange—it was designed after the GFC [global financial crisis]—it's been very effective, working as intended. The key here, though, is that its effectiveness can vary in different states of the world. It pays an attractive rate, it's not sensitive to demand, and it's highly liquid, all by design.

During times of stress, or rapidly rising rates, it is and can be an attractive place for investors to park their cash instead of banks, or T-bills. Of course, there's not many T-bills out there now, because of the debt ceiling situation. Of course, the designers of this facility understood this very well. There's a really excellent 2015 paper where they go through a number of these issues and talk about its design. When I think about the ON RRP, the Treasury cash balance, which is in green, and bank reserves, which is in blue—when you think about the debt ceiling situation, and when it is lifted and it's going to be messy, but it will be lifted, Treasury is going to issue say at least $500 billion in net new bills.

Most of that is going to come from the ON RRP facility, as it should. That's the very strong base case. The risk is that it comes from bank deposits, even just a bit, and that's going to put more pressure on the banking sector, potentially. It is prudent, from a monetary policy standpoint, to take some steps such as widening the spread between the ON RRP facility and IORB [interest rate on reserve balances]. This makes bank deposits marginally more attractive than money market mutual funds, and redirects liquidity a bit towards banks. It doesn't mean that liquidity necessarily flows to the banks that need it most, but it's a small step going into this debt ceiling lift that should be considered.

Along those lines, another aspect of the liquification of the financial system that is important to consider is the outlook for QT [quantitative tightening]. One thing that we've seen is that balance sheet policies have a tendency to gain a lot of momentum, and I'll point to QT in 2019, which you could reasonably argue went a bit too far, and then post-pandemic QE [quantitative easing], which also went too far.

Now it's a time when QT has gone smoothly, but it's worth considering enhancements to the Fed's communication strategy around the size of the balance sheet. The best resource on this is the New York Fed's review of open market operations. They lay out a very clear path for QT, but they have QT going for several years, not ending until mid-2025. I hope that's the case, but that's really the only source of clear communication from the Fed on this.

The market pays so much attention to the summary of economic projections, funds rate projections, inflation, unemployment, real GDP, to crystallize thinking, focus minds both on the FOMC and market. It seems potentially adding something like the size of the balance sheet, or of the security holdings on the balance sheet, to the SEP [Summary of Economic Projections] would help clarify communication and maybe focus minds around that a bit more.

At the end of the day, getting things like ON RRP, QT, of course the funds rate—I know we're going to talk a lot about that—but getting these other settings for policy is going to be very important as the hiking cycle draws to a close. Because a path to avoid recession still exists, and here I see the Fed's thinking on this as painting a pretty reasonable picture, and that would be reducing job openings, especially among the smaller firms, as I mentioned earlier, and that cooling of the labor market.

The challenge here, always, is that history shows the process can be difficult to manage smoothly. Labor markets can move from fewer job postings, which is what we're seeing, to a slowdown in hiring, to layoffs in a way that can be difficult for monetary policy to manage in a timely way. That comes to hard landing risk. I do see a hard landing risk as quite elevated. One reason is consumer excess saving is getting quite close to depleted. There's been a lot of borrowing; credit card revolving debt has risen rapidly.

These factors are offset a bit and counterbalanced by lower inflation and higher nominal wage growth. But higher rates, tighter credit, signs of a cooling job market, and in my mind, the risk of a hard landing appears well above the unconditional probability. Of course, the timing and severity of such events is very hard to predict, and on this point, I hear over and over and over, "a hard landing is likely to be mild." That's probably accurate, if we benchmark against the GFC and COVID recessions.

One thing, though, looking back at the last at least three recessions is that there was some element of countercyclical fiscal policy. I haven't talked a lot about fiscal policy at all, but fiscal policy had a very large role to play in the inflation that we did see. When you look at the current political configuration, I don't think we can really count on much countercyclical fiscal policy, at least in a timely and sizable way, if there is a hard landing. That's something that just keeps me a little bit cautious about declaring "any hard landing is likely to be mild."

Just to finish up on a few points on inflation, this chart just shows the disinflation from business cycle peaks from 1973 to today. You can see the red line is the cautionary tale in 1973. The Fed has communicated very strong commitment to not repeating that. Otherwise, though, it shows the disinflation you get; it does happen. It's a bit modest, and it probably will take a hard landing to get inflation from where it is today all the way anchored back down to 2 percent.

It's worth thinking about how to communicate about the price stability leg of the dual mandate once inflation's back at around 2 percent. Here, I think about the next framework review, and this might be too soon, but I do think it's worth thinking about doing something like adding a tolerance band around 2 percent, because I don't think the Fed wants to be in a position of trying to fine-tune the economy if inflation is running 2.2 percent or 1.8 percent, with critics saying they're failing on their mandate if they're at 2.2, concerns about the ELB [effective lower bound] if they're at 1.8.

I've done some work with the coauthor at the San Francisco Fed, Andrew Foerster, and one thing about a tolerance band that I find very attractive is that if a central bank gives a forecast and a tolerance band, you can objectively assess if the central bank is succeeding and hitting what it says it's set to achieve. It's a measurable way to evaluate central banks, and something that down the line, once 2 percent is back where it needs to be, in line with the target, is something that would be valuable to consider. I'll stop there. Thanks.

English: Perfect. Thanks so much, Troy; that was perfect. Kristin, you're up.

Kristin Forbes: Thank you very much. It's a pleasure to be here today. Could you put up my slides? When they come up, I wanted to start off with some numbers. I'm from MIT, so I like numbers, but we need the numbers on the screen. Here we go. Okay, thank you very much. Okay, you're all warmed up now. We've already had some nice speakers to get you thinking. Let's just start off with some numbers. Policy interest rates of some of the major central banks or countries in advanced economies, countries with independent central banks, flexible exchange rates, some sort of inflation targeting regime.

Take a look at those numbers closely. See if anything jumps out at you. Just to be clear, I'm going to show you a number of numbers today, and in each of these, when I report the US, I'm going to just—the US obviously reports a band now when they adjust interest rates. I'm just going to report the top of the band. The US just raised interest rates to the band of 5 to 5 1/4; I'll just report 5 1/4.

For all of you who follow markets closely, if you look at that you might think something's off a bit. Are the MIT RAs not up to their old usual standards? If Mervyn King was here, he'd probably point out right away that the UK rate isn't right. The UK just raised rates to 4.5 percent, but markets are expecting UK will probably get up to about 5 percent. Maybe this is actually a chart of where policy rates are expected to end this cycle, which would fit; the US is expected to end right about at 5 1/4, UK at 5.

But there are a couple others that might be off a bit. Stein, for example, you might be looking at Switzerland and notice that's off a bit. But you don't need to follow markets closely to notice what's really off: I didn't put a date. Take a look again. We all jump to the conclusion that this is today—it looks pretty much like today—but actually, these are policy interest rates from the end of 2006. Not surprised if you missed that, because if you do follow markets closely and know where rates are, today they're pretty close to where they were in 2006.

The average of the set of eight countries is basically the same as where it was in 2006, so this raises the question which I'd like to focus on for my comments, which is: Do we really need new clothes—when thinking about central banking, at least? For some of the other topics of this conference, yes. Electronic currency, some of the shifts in the web, we need new thinking, new tools, new strategies.

When thinking about interest rates and monetary policy, maybe what we shouldn't do is focus on getting new clothes. Of course, when the emperor got new clothes, that didn't work out so well for him in the famous fable. Instead, maybe we should just go back to how central banking used to be done, and rethink some of the lessons from traditional central banking before what was probably an unusual era after the global financial crisis, when interest rates were near lower bounds in many advanced economies and inflation risks were minimal.

Again, the title for the session is How to Conduct Monetary Policy amid Higher Inflation and a Nonbinding Zero (Effective) Lower Bound? If you take a step back and think about it, this is how monetary policy is traditionally done. This isn't necessarily a new thing. This is how these regimes were designed. Before this conference, I had an email exchange with Pablo Garcia from the Central Bank of Chile, who was supposed to be here (had to cancel last minute), and he was very succinct. He sent me this note and said something along the lines of, "What are you going to talk about? Isn't that how inflation targeting works?"

So maybe we don't need to look for something terribly new here. Instead, what we should do is go back to more traditional monetary policy, and maybe unlearn some of the shifts in monetary policy we adopted in response to the unusual era of the 2010s, with low interest rates and low inflation.

More specifically, for the remainder of my comments I want to focus on three old lessons for monetary policy that we should bring back today, and maybe rethink some of the shifts that took place. That's in terms of guidance versus optionality; second, anchoring inflation expectations; and third, strategies of how to adjust rates (gradualism versus front-loading). Let's jump in the first area, where there has been a notable shift in response to the global financial crisis, and that was in the use of forward guidance.

After the global financial crisis, as interest rates went to their lower bounds in many advanced economies, it made sense to shift to relying more on guidance as a monetary policy tool. You couldn't lower short-term interest rates anymore; they were at their lower bounds, so it made sense to rely more on guidance to bring down medium-term interest rates.

It made sense to do that by also using other tools, like QE, as a way to commit the central banks: "We will not be raising interest rates soon, because we're also doing QE. We can't raise rates until after we've unwound QE, stopped QE, and then we could get around to raising interest rates." So that was a way to really pre-commit to keeping rates low for longer. Again, that made sense when inflation was low and there was little risk of inflation. You could be slow about any adjustments. That made sense also when COVID hit, in the initial stages. These are expectations for inflation in the US, and it was believed that inflation would remain low, below 2 percent, potentially even lower.

So again, it made sense to follow the playbook from the 2010s. Relying on forward guidance is a way to commit to keeping rates low, provide additional accommodation, because inflation was not a risk. Part of that was committing to unended QE in the US, or long programs of QE, and saying, "We will not adjust interest rates until after we taper QE, end QE, and then we will start to raise rates."

It was really a forward commitment. This changed when inflation picked up faster than expected. Here's what happened to inflation swaps in the US. Suddenly, it became clear that adjustments had to be made sooner than had been expected, and this is where the dangers of guidance started to kick in. If you're pre-committed not to raise rates for a period, and you've really locked yourself in by having a QE program in place that will take time to end, it became harder to pivot as the economic situation changed.

One of the lessons learned is you don't want to rely too heavily on guidance with a commitment of a QE program in place if you may need to pivot quickly. It is partly because then it took a while for the Fed to talk about talking about tapering QE, and then tapering QE, and then ending. By the time they got through all that and could finally raise interest rates, it was after inflation expectations had picked up quite a bit, and that made the task harder. We can certainly argue about exactly when it would have made sense to raise without all that, but the overreliance on guidance made it harder to adjust when the situation changed.

That's one lesson learned: Keep optionality, especially during uncertain times. Don't lock yourself in too much with guidance.

A second lesson we have learned, or we need to relearn that we used to know that we may have forgotten in the era of the 2010s, is the role of making sure that inflation expectations remain anchored. In the 2010s, there was a lot of work, especially academic work, on the importance of anchored inflation expectations. There was a confidence that central banks had been successful in bringing inflation down, inflation expectations were anchored, and as a result central banks did not need to respond aggressively when inflation did veer from target because anchored inflation expectations would mean that as long as the shocks were transitory, however that was defined, that inflation would come back, and that was hardwired into all of the models that central banks relied on.

That did work pretty well for the era of the 2000s and 2010s. This shows that there was substantial variation in inflation. Inflation went up, inflation went down, but inflation expectations remained pretty well anchored despite these movements. Again, that strategy made sense when many of the deviations of inflation from 2 percent were short-term effects, often global shocks as inflation was more determined by global factors. Many of the shocks driving inflation from 2 percent were commodity shocks, energy shocks against supply shocks that faded after a year, so the strategy made sense in the 2010s.

The lesson that we just learned over the past few years is that we can't take inflation expectations being anchored for granted. If you get large shocks that push inflation away from 2 percent, even if they are transitory, you can't count on those inflation expectations remaining anchored at 2 percent, especially if those shocks are repeated in the same direction.

We saw that, for example, in the UK, where the series of shocks were driving inflation away from 2 percent. Inflation in the UK has now been above 10 percent; it still is there, although it should come down soon, and that has caused companies to adjust how they set prices. They set prices more often not on regular time intervals, but just based on current inflation, much more than in the past.

Workers are much more aggressive in demanding higher wages; this has meant that some measures of inflation expectations have gone up quite a bit in the UK. Short-term inflation expectations always bounce around with current inflation, but some of the medium- and longer-term inflation expectation measures in the UK bounced up to around 4.5, even getting close to 5 percent. That was a big move and was a poignant reminder we cannot take anchored inflation expectations for granted, that could involve responding to temporary transitory shocks if those shocks are big or repeated in the same direction.

The third lesson that we need to go back and relearn from history is different strategies for adjusting interest rates. A big shift in monetary policy, since the global financial crisis, was a focus on gradualism when adjusting interest rates. Here's, for example, how the US and UK adjusted their interest rates over the cycle after the global financial crisis: very, very slowly. The US raised interest rates for its first time in December 2015, and then it waited about a year before raising rates again.

After that, it raised a bit more frequently, but not every meeting and only by 25 basis points. At the end of the cycle, it got raised up to about 250, so that meant it averaged about 50 basis points of hiking a year during the cycle. The UK was even slower, started later, raised 25 basis points, waited about five meetings, raised 25 more, and that was it. Some other central banks didn't even raise rates at all.

Gradualism was the preferred strategy, and that made sense because after a long period of very low rates, there was real concern about what would happen when you did start to raise rates: would something break? Where were the fragilities that could emerge? This also was feasible because inflation risk was low, so there was no urgency of raising rates a lot, and gradualism got a lot of support.

The lesson we've learned over the past couple of years is that this is not necessarily the best strategy when you've got a lot of ground to cover. There are times when front-loading, going in large increments, makes sense even if it does increase financial risks and risks breaking something or causing some financial pressures.

Let me give you a concrete example of that, to make that point. The US and UK is a really nice contrast. These are two central banks that lowered interest rates to basically zero when COVID hit; they're both expected to end their hiking cycles at about the same point (around 5 percent). Both are independent central banks, large financial markets, et cetera.

What makes this interesting is they followed very different strategies to get from roughly the same point A to the same point B over this hiking cycle. The UK started raising interest rates a bit sooner, and their first rate hike was in December of 2021. The Fed waited about four months, until March, to do its first interest rate hike. UK started earlier, but the UK followed the gradualism approach. It raised very slowly, going first 15 basis points, then 25, then 25, and not moving to bigger hikes until the later part of 2022.

The US, on the other hand, started later. Again, it was harder to get started because it had this unlimited QE program that took longer to get out of. US started later, but then it made up for that by aggressively raising rates. It quickly went from 25 to 50 to 75 basis points in hikes. The US very quickly overtook the UK, and it's going to end its hiking cycle earlier than the UK. Again, the dash lines are where markets expect things to go, although we don't know for sure.

The bottom line: UK was a slow and gradual approach—gradualism. US was a more aggressive approach of front-loading rate hikes. What was the effect? To see the effects, let's look at what happened to inflation expectations. I'm going to focus on the window from before either of them started hiking—December 1, 2021, until beginning of September, because after that was when the UK had the mini-budget and LDI issues, and there were other factors driving markets. Let's just focus on this window, and this is the window when you really saw the different strategies for raising rates, gradualism versus front-loading.

In the UK during this window, short-term and medium-term inflation expectations increased quite a bit more. You might say, "Well, some of that was also different shocks" which is fair; the UK was hit by a bigger energy shock as Russia invaded Ukraine. But let's focus on longer-term inflation expectations. Those shouldn't be affected by short-term energy prices and short-term shocks. During this window, when the UK was following its more gradualistic approach, inflation expectations, longer-term measures, went up quite a bit more than in the US.

This is just one example, one comparison, but it is a reminder that the strategy which has again become popular, a more gradualistic approach to raising interest rates, isn't necessarily the best one. We may need to go back to what we did in the '90s and 2000s, of sometimes you need to raise more aggressively. You can't go slowly, or you could risk having inflation expectations become unwound.

Just to sum up: back to the emperor's new clothes. In this case it's time to bring back some of the old styles rather than looking for new clothes or new styles. In particular, when thinking about strategies for monetary policy, not relying overly on guidance. "Maintain optionality" is a key lesson we've learned over the past couple of years—a lesson which we followed back before the 2010s, but we need to keep optionality when there's a lot of uncertainty. Beware of constraining yourself too much with guidance in an era where interest rates are higher and inflation is a risk again.

A second lesson is we need to take anchoring inflation expectations seriously. We can't take that for granted. That could mean more aggressive actions in responses to transitory shocks because we can't let those get unanchored.

A final lesson is on strategies for adjusting policy. The strategy that worked well in the 2010s may not work well going forward. We may need to go back to an era where you front-load, you move in larger increments, and you really need to think about the costs and benefits of different strategies.

It reminds me of, if you keep your old clothes, if you keep an old pair of jeans for about 10 or 20 years, the style often comes back in again, and that can be said for monetary policy strategy, too. The strategies we followed 15-20 years ago are the ones we should be returning to. Some of the shifts that occurred in the 2010s in response to the era of low rates/low inflation made sense then, given the environment, but they don't make sense necessarily going forward the next few years. Thank you.

English: Thanks very much. Jan, you're up.

Jan Hatzius: All right; thank you so much. Thank you to Raphael and the organizers for inviting me. It's wonderful to be here; lots to talk about. In thinking about the inflation upturn, obviously there are unavoidable elements to it and there are avoidable elements to it. How big the unavoidable versus the avoidable elements are really depends on which measure you look at, but clearly there was some of both.

One really important point to consider is that how harshly history will judge monetary policymakers around the world for the avoidable elements is going to depend a lot on what happens from here. If we manage to bring down inflation without a deep recession, then a lot of the monetary policy errors are going to be viewed as relatively minor in the grand scheme of things, of the pandemic and its aftermath. If inflation gets entrenched at high levels, say in the 3 to 4 percent range, or we have a deep recession, then the historical judgment is going to be quite a bit more harsh.

Now, our view is on the optimistic side. Our forecast for how the economy can unwind the very high inflation rates has been on the optimistic side. The consensus forecast for the US, if you look at the private sector forecasters, has been for a recession since about August/September 2022. Right now, more than three-quarters of the forecasters in the Wall Street Journal survey expect a recession in the next 12 months. The board staff expects a recession, though they have characterized it as a mild recession. We have been in the "soft landing" camp and continue to be in the "soft landing" camp, in terms of the baseline forecast.

In terms of near-term growth outlook, our expectation is that we will see a drag from tighter credit availability. We've used a working estimate of about 0.4 percentage points on the Q4-to-Q4 growth rate from a tightening of credit availability, over and above the drag from tighter financial conditions that one would expect given the adjustment in monetary policy and the past cyclical environments on an extra 4/10 of a percentage point.

But we also see some important areas of support for activity, in particular the sizable rebound in real disposable income, and the adjustment in the housing market where the biggest drag from the increase in mortgage rates from 3 percent to 7 percent appears to be behind us, and the drag from housing in 2023 is likely to be smaller than in the second half of 2022.

Then, the question is: if there is no near-term recession, can inflation come back down to the neighborhood of 2 percent? There are certainly still reasons to be concerned. The labor market is quite clearly still out of balance. A reasonable summary of labor market tightness is the gap between job openings and unemployed workers, or between total number of jobs and total number of workers, or the ratio of job openings to unemployed workers. Here is the longest series that we can construct for that ratio using an extrapolation of job openings before 2000, using the newspaper help wanted index. Clearly the labor market is still very tight. The ratio is still at 1.6, which has only been seen a few times in post-war history and, obviously, in periods that were often either at the start of or in a very inflationary period (early 1950s and late 1960s).

However, there is a significant amount of progress that we have managed to achieve, even in an ongoing positive growth environment. The job openings to unemployed workers ratio have come down from 2 to 1.6, halfway back to the level that we saw on the eve of the pandemic when inflation was at target. The way that the adjustment has occurred has been exceptionally benign so far. Obviously, we can see a reduction of the gap through declines in job openings or increases in unemployment. So far, the entire adjustment, actually a little bit more than the entire adjustment, has occurred via declines in job openings, without any increases in the unemployment rate which has been a very benign, if you will, shifting back of the Beveridge curve towards the pre-pandemic position.

How unusual this is, is shown in this chart, which again uses an extrapolation of job openings via the newspaper help wanted index, back to the early 1950s. What I'm showing here is the cumulative increase in the unemployment rate from the cyclical low to date for each business cycle, and the cumulative decline in the job openings rate from the cyclical high to date for that business cycle. As you can see, prior to the last couple of years the increase in the unemployment rate and the decline in the job openings rate were largely mirror images of one another. If you look at this very long-term period, we really have not seen examples of job openings coming down without significant increases in the unemployment rate, until now.

Of course, we're still going through this period, but it is actually unprecedented for the job openings rate to be down by more than 1.5 percentage points without any increases in the unemployment rate. We've made a significant amount of progress.

Other indicators of labor market overheating have also adjusted in a so far very benign way. The quits rate is obviously an alternative helpful indicator of labor market overheating. We're back a little bit more than halfway towards the pre-pandemic level from the cyclical high of 3, and the right-hand side shows a measure that we have put together on the basis of earnings calls for Russell 3000 companies, counting each management team that mentioned labor shortage or worker shortage by quarter, going back to the mid-2010s.

There was a huge increase in 2021: 16.5 percent of all management teams in the Russell 3000 were talking about labor shortages in the third quarter of 2021. That has also come back down to about 5 percent in the fourth quarter of last year; we don't have the first quarter numbers for this measure yet. The labor market indicators, they're all showing significant progress. There is still clearly work to do, but we appear to be on the right path.

A similar takeaway from the wage numbers, where we've also seen clear deceleration. Some measures are showing more deceleration than others. The left-hand chart here is our sequential wage tracker, which is effectively an employment cost index/average hourly earnings combination that was running close to 6 percent in late 2021, early 2022. It's now running at 4.5 percent.

The right-hand side shows a composite measure of wage surveys, where we take the NFIB [National Federation of Independent Business] actual and expected wage changes, the Dallas Fed, Richmond Fed, and a number of other Federal Reserve Banks that have basically diffusion indices for wage growth, and we then scale that to average hourly earnings growth. There's been about a 1 to 1.5 percentage point deceleration from the peak. I would say the details are always going to depend on which indicator you look at, but most indicators are pointing to a move in the right direction, but still above the 3.5 percent or so growth rate of wages that may be consistent with 2 percent inflation in the longer term.

Lastly, if we look at the inflation numbers themselves, we're starting to see clearer signs of improvement. There are different ways, of course, of estimating the underlying trend of inflation. Here, I have focused on the trimmed mean CPI and trimmed mean PCE, and the dots here are our forecasts for the six-month growth rate through the end of 2024.

Now, here I've taken our forecasts for the core (excluding energy) measure, but it's a similar story so we're on the right path. There is still a ways to go. We think the economy is solid enough for the funds rate to remain at restrictive levels, roughly consistent with the March dot plot. The market is obviously priced for much lower rates, large declines in the funds rate. Some of that deviation can be explained by the fact that market pricing is a probability-weighted average of sorts, whereas our baseline forecast, or the median dot plot, is a baseline case. Even if we roughly try to adjust for the entire probability distribution, our view is that market pricing is too aggressive in bringing down and expecting a lower funds rate over time, under our forecast that the economy continues to expand even with inflation gradually subsiding.

That concludes my remarks. Thank you very much.

English: Thanks very much. John, you're up.

John Taylor: Thank you. It's an honor to be here, in this audience and to be on this panel with these people. Thank you so much for all this. I'm going to talk about the topic that was assigned. I'll see what I can do. There it is. I just repeated it. "Policy session one."

Let me just begin with a little background, if you like, which is important, to put this all into context. Just last Friday, we had a conference out at Stanford-Hoover; we called it How Monetary Policy got Behind the Curve, and How to Get it Back. This is actually from a year ago). At that point, monetary policy was behind; there was debate about that. This year we went over that a little bit; but more important, we decided we're going to have "how to get back on track."

That's, in some sense, what we're all about today: how to get back on track in the least damaging way, maybe no damage whatsoever. We'll see if this is the title of our conference that you just had. But let me just mention: it wasn't that long ago, we had a headline of "Fed Boosts Rates to a 16-Year High," and there's a little graph on the front page of the Wall Street Journal. You can see it over on the right there, it's a little peak which is getting a little over 5 percent.

This is not the news that's happening. The news is looking at the overall picture. This is a picture that goes way back in time. You can see in some sense, this is still low compared to where it was in the '70s and the '80s; it came down, and it is a 16-year high, but it's not that high given what we've seen in history.

We need to put this in the context, I believe, of how monetary policy is formulated. What I'm going to add a little bit is how monetary policy is formulated, how it should be formulated, and maybe it should be formulated differently going forward.

Let's also just remember, this rate of increase we've had recently is there. You can see it from the FOMC plot (this is from FRED, obviously), and you can see the last move on the right was just over 5 percent. There's been a lot that's happened with respect to the federal funds rate and the target federal funds rate by the Fed over this period of time, but this is not the whole story because it depends on what is going on that's driving this. What I'd like to do is just review what I think the basic story has been from the Fed.

Not that long ago, the Fed chair, Jay Powell, was saying, "I find these rule prescriptions helpful." Mario Draghi, who was the president of the ECB: "We would all clearly benefit from" such policies. Raghu Rajan, who was the governor of the Reserve Bank of India: "What we need are monetary policy rules."

As you all know, this approach was interrupted when the pandemic began in early 2020. Rules were out, they weren't in the report. Then by February 2021, rules were back in the Fed's monetary policy report. It looked like we were, from my perspective, back in business again. But then rules were out again in the February 2022 report. Jay Powell, under quiz from the Congress, said he would put rules back in. Indeed, in the monetary policy report released on June 7th of 2022, rules were back in, including the so-called Taylor Rule, which was listed number one on the list, and things were changed. In some sense, it's still before the delivery of the change occurred.

Now, as the third tick from the bottom says, many changes were seen in actual monetary policy. You just saw some of those. I'm going to argue a gap still exists between the rules-based policy and the policy actions. It's also important to remind ourselves that the Fed is not the only central bank in the world. At our conference at Stanford-Hoover last Friday, we had the former governor of the Bank of Japan, who's been there for 10 years, speak a little bit; and we also had a characterization of Latin America. It's a global thing, so don't forget other central banks.

We're going to argue, we're still in this era of high inflation. We saw some of the charts already, and it's important to put this in the right context. Also, there's lots of other stuff going on around the world. Ukraine recently raised inflation, but that's not the basic story that I'm going to stress in these short remarks.

The basic story maybe goes back to what is behind the scenes a lot, which is asset purchases or money growth. These are asset purchases, if you can see this in the chart very well. I labeled January 5, 2022. They have come down a little bit.

This is one measure of the impact of policy, is your total assets. You can see the big increase, and then the smaller increase, and now it's actually declined somewhat. This is the picture of M2. Obviously, the same pattern, it's very clear but M2 has also leveled off quite a bit. Don't say "those who are interested in interest rate rules: forget money supply;" it's very important to keep this straight. We always want to match this. In fact, policy rules with the interest rate came out of policy rules for money growth. To put this in context: it's not clear exactly how to do it, but it obviously was a big increase and we're having a decrease recently.

This is the most recent chart from the monetary policy report from the Fed back in March '23, and you can see it's got various rules there. This is actually, whether you like the rules or not, this is constructive because it's a way that the Fed, different members of the FOMC, different staff of the FOMC from where they are, can talk about rules. Now, you can't see this too well in the chart, but it's actually a replica of the table on page 43. These are all variants on the so-called Taylor rule: adjusted Taylor rule, first-difference rule. The balanced-approach rule is a little new, but they're very similar. That's important to keeping track; they're all similar, and what I want to do in my few minutes is just remark on that.

If you go to the so-called Taylor rule, which, this is notation I originally used. R is the federal funds rate, P is the rate of inflation of the previous four quarters, Y is percent deviation of real GDP from a target, and 2 is the normal rate. Then there's little descriptions down below. This is 30 years ago, by the way. There are the descriptions below. If you have an inflation target of 2, then the interest rate should be 4.

There's a notion, which is very popular, that the equilibrium rate is no longer 2, it's 1, so 1+2=3. However, if inflation starts to rise above the target of 2—which is there in the picture, P minus 2—then if it's 3, for example, then the interest rate should be 4.5. Just if it's 3; if it's 4—which is maybe where it is now, maybe it's a little above that—then you have to add from 3 to 4 throughout this equation. That gives you a 6. In some sense, people ask me, what do I think the rate should be? I think it should be around 6. That's a little higher than it is now, but I've been saying 6 for a long time.

The text below that simply describes if you plug in the numbers, inflation rate over the past four quarters of 4 percent, maybe higher than that, target inflation rate of 2 percent, there's talk about changing that; I hope not. The equilibrium interest rate at 1 percent—that's what seems to have come out of the recent research—and the gap between GDP and potential is zero. There's maybe some debate about that, but that's how you get to 6 percent. It's not too far from where we are now, but it's a lot farther from the 0.25 percent we were not so long ago.

This is the algebra. It really says the same thing as the chart. It says 6 percent or 5 percent, depending on how you measure the inflation rate. Let me mention a couple of other things before I stop this. This is a picture of the inflation rate in the US, and you can see it's gone up quite a bit. It's almost as high as it was in the bad old days, when many of us started on this subject.

It's quite remarkable, and that's the thing that got us worried about this a couple of years ago or so, and it's only with a lag that there has been an adjustment of monetary policy. That lag, we will be studying this for a long time, that lag, that slow adjustment, is one of the reasons inflation got so high, and why it is there's debate right now amongst us about how to get it down, and whether that will cause some contraction. I hope it doesn't.

That's the picture. Also, let's think about the inflation rate in other parts of the world. This is the inflation rate in Europe, and you can see it also got very high; it's not just the US. I think those are related. My study of monetary policy over many years has indicated central banks, even if they don't admit it, there's some relationship between the two. Many of you know, European central banks seem to be behind the curve by the same measures that we have used if you look at the Fed.

I put these up just to think about the rest of the world. I also want to mention the situation in Latin America. We had a focus on Latin America in our meeting on Friday. It's very important, because there's a sense in which this whole phenomenon has spread around the world. You can see Brazil, Colombia, Chile, and Peru, et cetera, Mexico, and they're all very high. This is not the most recent data, but it's close, and you can see that, it's about the same time. It's obviously related, and what are, in retrospect, the extraordinary low interest rates in the US led to other countries, including Europe, including Japan, including Latin America, to have extra low interest rates.

In a sense, this is a global phenomenon. If the Fed is able to get back to normal, get back on track, so to speak, in a sensible way. What's most important to remember here is it's implicit in the remarks that we've already heard. If you can do this in a way that's clear, that's announced, that's not a surprise, then our system works well. It should have the inflation rate coming down, without the big impact on GDP that many people are concerned about.

This is just a chart that indicates I'm not the only person talking about this. This is from Grant's Interest Rate Observer. It's a direct chart. You can see it's very similar to mine, and you can see the implicit interest rate in the Taylor rule is still high relative to that.

I have 45 seconds. Let me just conclude. The key question which is behind our thinking: Are we entering a new era of high inflation (inflation has picked up dramatically)? I'd say yes, we are, unless policymakers at the Fed take policies which are consistent with the 2 percent inflation goal, and there's some discussion about that. There are more reasons than ever for central bankers to be guided by some kind of rules-based policy. We would not have gotten so far behind the curve if that had occurred.

In fact, what I've tried to do in these remarks is outline the method to do so. Central banks should use rules that markets understand—the simpler, the better. It can't be too simple; that's why we have models, where we have the description of models.

By the way, the interest rate would increase, just as it has, if inflation starts to rise. Maybe there's a question of whether it's enough. I said, "well, maybe 6 percent would be better" but it's in the direction we've called for, just a little bit late. Of course, all these would be contingency plans. You'd always be adjusting the rate, as we always think about interest rates, I hope, in a contingent way.

That's the key here. You don't just say "it's going to be 25 basis points," you don't say "it could be 5 percent;" you don't say "it will be X in different countries," but you have some strategy, some rule, some procedure ... I don't necessarily use the word "rule," but some strategy ... that you have in place. I believe that if we had this approach, and there's more talk about it to some extent that's because of the big deviation in the US and other countries, it would greatly reduce the chances of large, damaging changes later.

The hope now going forward is that central bankers will gradually ... it's harder where they are now ... but will gradually reduce the inflation rate by keeping the interest rate from coming down too sharply. Well, it depends. It's a contingency plan, so we don't know. But that's the check, that's the thing we're all searching for.

Thank you.

English: Thanks so much. As I forecasted, I was right. All of our panelists went slightly over and we started late, so we actually don't have a whole lot of time. What I'd like to do is give each of you a chance to comment on the views that other people expressed, maybe with a little bit of a focus on the point that John was making: are we back on track?

John suggested the rate had to go to around 6 percent, and because everybody is interested in it, as Jan talked about, what is the path from here? Do we think that the Fed fell behind? It sounded like almost everybody thought that was true, to a greater or lesser extent; but are we now back on track in a reasonable place, or not so much, with more work to do?

I'll go in the same order. Troy, I'm going to put you in the hot seat by having you go first.

Davig: Okay. Comment on each?

English: No, just your main reaction from the discussion so far.

Davig: I thought Kristin's first slide was very interesting, showing similar nominal rates, today's before the GFC. The thing that jumped out at me, though, was inflation is so different, and it makes me think this debate around, "where is the neutral rate?" Most people think it's either about where it was pre-GFC, or possibly higher. That chart shows us, today at least ex-post real rates not neutral rates, but ex-post real rates, are still much more accommodated than at that time.

I'm just trying to think about what that might tell us about neutral rates. I would need to sit down and work it out, but that's an interesting comparison. Jan, the whole "slowing in the labor market via lower job openings" makes a lot of sense. Netting everything out, I might be slightly less optimistic on the soft landing, but that's just a degree of probabilities, and not necessarily major. But I do think, like I said earlier, the hard landing risk is a bit more elevated.

On Taylor rules, I'm a huge fan of Taylor rules. They have an incredibly important place in setting and making monetary policy. Right now, like literally right now, maybe some credit adjustment, as Jan mentioned, seemed very reasonable, and how that maps into the monetary policy setting might be useful. It's very hard to do, in a sensible way. But I guess that's one thought on that.

I guess one thing we didn't talk too much about, that I had a bullet point on and I kind of chickened out saying it, but is: was FAIT [flexible average inflation targeting] a mistake, or did it serve its purpose? When I looked at that in retrospect, and that really being one reason for abandoning the rules, where I come down is FAIT in some ways did achieve the objective of getting rates off the ELB. Fortunately, longer-term inflation expectations have remained anchored. That's a case that can be made.

It's more of the implementation of FAIT, and it was, "we're not going to lift off until hitting full employment," or whatever. It was the implementation and that forward guidance that really was at odds with implementation of the Taylor rules. It's always hard to second guess this stuff, but that's maybe one thing, if it were to be replayed, it would be done a little bit differently.

English: Okay, thank you. Kristin?

Forbes: Thank you. I'll follow that same framework and do one comment on each of the presentations. Troy first, I'm delighted you brought up issues around balance sheet QT. That's a key issue, we just with 10 minutes couldn't cover it all. I didn't cover it, because it was the one topic that didn't fit in with the theme of my talk about needing new clothes. We need to return to sort of old-fashioned monetary policy in terms of adjusting policy rates, but the one area we do need new strategies, new thinking, is on balance sheet unwind. This is obviously a situation we haven't been in before.

This is where, again, we just don't know what the impact of balance sheet unwind will be. When I was at the Bank of England, I did quite a bit of work trying to sort it out, and the best you could really say is, if you shrink a balance sheet you could have an effect which is basically equivalent to doing QE (you just flip the sign), which meant a pretty meaningful effect on monetary policy conditions, or you could have zero effect if it really is in the background and like paint drying.

That left a huge range of effects, so we just don't know what the impact of QT will be. So far, and this is where central banks do deserve some credit, even if they were slow to raise interest rates, they've been pretty quick to unwind balance sheets, given the uncertainty about what the effects would be.

This is where, in a couple of years we're going to have a lot more information. Different central banks are unwinding their balance sheets in very different ways. Some are just letting purchases roll off automatically when they expire, so it's really in the background. Some are doing like the US; some are doing active sales, like the UK has started some active sales but it's in very small increments, so it's going to be still a very slow shrinking of the balance sheet.

Then there's some, like Canada, that are doing automatic unwind, but since they own more short-dated securities, it's going to mean the balance sheet will shrink quite quickly. We're going to have a lot of information pretty soon on what the effects are, if there are effects. That is where we may need some new thinking, new strategies.

Jan, I agreed largely with what you said, especially your optimism. For people who follow the UK, there was at least the changes in their forecast support, a more optimistic outlook. The UK last year was predicting that they would enter a recession starting at the end of last year, and it would be five quarters of negative GDP growth. A pretty long recession: negative GDP growth for over a year on a quarterly basis. They just upgraded their forecast on Thursday, and growth has been so much stronger than expected, they're now not even expecting one quarter with negative GDP growth. It's the largest upgrade in a growth forecast in the Bank of England's history.

We have been missing something. The consumers are more resilient, people have been able to handle these shocks better than expected. There is reason for optimism, and I hope you're right for the US case.

I also had a question on your presentation I hope you might address. You had this long series on vacancies, and a lot of your case is built on optimism about vacancies coming down without a big hit on the real economy. I've always wondered, how much can we trust these historic series of vacancies? How companies list jobs has changed quite fundamentally. There used to be quite a bit more cost. If you wanted to do a job posting, you had to go to the newspaper, pay a fee, they had to print it. Listing a vacancy was hard and costly. Now you just post something online, it's so easy. I've never felt comfortable with the historic series. Wondering if you could comment on that.

John, finally, the discussion of more formal rules versus more discretion: I just want to sort of link that to my comments. Maintaining optionality is key, but that doesn't mean you should disregard rules. Some people might think it's similar. Optionality is important because the world is constantly hit by major shocks we can't predict, so you need to be able to pivot and adjust quickly, and that could be consistent with rules.

Rules, I found when I was at the Bank of England, were consistent, and I often looked at different variants of Taylor rules, not necessarily to follow them religiously, but as a way to force myself to say, "If I'm not adjusting the way the rule tells me, why not?" and really think through, "What do I think has changed? Why am I veering from the model?" If you can't answer that question, then you probably should take the model more seriously.

But at the same time, I also worry about trusting the models too much, because most of our models have just been so wrong. We've missed that there are shifts in parameters, there's non-linearities, It's hard to capture all that.

Finally, just the question on FAIT, Troy, which you mentioned. Here's how I would think about this in the context of my comments. FAIT was a change of strategy that made sense if you thought monetary policy would always be done in an environment of the 2010s, with low inflation and low rates. That's not the world we live in now; that was an aberration. We are likely in a world where inflation is going to be more of an issue, higher rates—so you need a framework that is symmetric, and FAIT is not symmetric. We do need a real rethinking of that.

Hatzius: All right, thanks. On Troy's suggestion for a tolerance band around inflation: that makes a lot of sense. The cost of that is low, because most people probably already think about a tolerance band, but the communication benefits could be significant because there are too many people who think that the Fed actually expects to be exactly at 2 percent, which, obviously, is unrealistic.

There was a little bit of a difference of view on whether we need more guidance or less guidance, maybe guidance around the size of the balance sheet or the securities portfolio. I guess I would probably lean more towards Kristin's view that there is probably going to be less room for guidance in the environment we are in now than there was for a period before the pandemic. I would probably be a little bit more cautious around making projections or providing guidance for the size of the portfolio, because it's just very, very difficult to have a confident view on it.

On inflation expectations, of course, I agree how important it is to keep inflation expectations anchored. I would say if I looked back on the last two years, I've actually been struck by how remarkably anchored especially longer-term inflation expectations have remained.

It does vary a little bit depending on which economy you look at, but even if you take the Bank of England's success or failure on this, and you look at some of the numbers that, Kristin, you had on your charts, the deviations were still relatively limited. If you had given me all of the information about the economy, you'd given me everything else over the last two years and had asked me to predict inflation expectations, I would have been too pessimistic.

On the Bank of England growth forecast, it was very striking. Last year's deep recession forecast that you referred to, Kristin, it was probably the most negative economic forecast that I have ever seen from any central bank relative to the private sector consensus. It's just remarkably downbeat, and of course that also has a lot to do with the different path for where policy rates have gone in the US versus the UK, just because the BOE was so negative.

Lastly, on vacancies: how reliable are historical series of vacancies? I would say that they can see numbers over the long term. Even these extrapolated vacancy series that use the newspaper help wanted index before 2000, they're remarkably stationary. They look like a pretty good series. That doesn't mean I would take them literally, especially now in the post-pandemic period, given the decline in the response rate for JOLTS on a month-to-month basis, or maybe quarter-to-quarter basis. You have to take these numbers with a significant grain of salt.

We do have, fortunately, on vacancies, private sector measures that can serve as a cross-check—a company named LinkUp, a company named Indeed, and then there are a couple of others. LinkedIn has a type of vacancy measure. ZipRecruiter has a type of vacancy measure. We can cross-check, and the broad picture of what we see in the jobs numbers, at least after the declines of the last couple of months, broadly matches what we see in in other measures.

I would certainly say that the labor market utilization is always worth looking at a range of different indicators. The quits rate is also a conceptually very appealing measure. These mentions of labor shortages have their value, so I would want to look at it broadly, for sure, But the vacancy rate is a much more important measure of labor market tightness than most economists would have said 5 or 10 or 15 years ago. Certainly, in my thinking, the jobs-workers gap, and the vacancy rate, now play a significantly bigger role than they did a while ago.

Thank you.

Taylor: Okay, thanks for all the comments. I'd say one thing: let's not forget there are other kinds of policies; there's fiscal policy, there's regulatory policy, international trade policy which are all very important parts, even if we get monetary policy right. But I'd say the lesson I'm drawing, which you could hear from my presentation, which, I listened very carefully to everyone, is still that there is some value in having a systematic approach to policy. It's getting more popular again. I don't know if it will last, but this deviation, which you've seen in the US and we've seen in the UK, we've seen in Japan, we've seen in Latin America, I gave some examples of that, is an indication.

I don't want to say there's only one rule, that's for sure. We have to look at it. They didn't come from nowhere. I used to have these gigantic models, and simulate and make it smaller, make it reasonable, and the 2 percent inflation target didn't come out of nowhere. The more that we can figure out what the best thing is to do. I don't disagree with the panelists saying we should try different methods, try different approaches.

But let's not forget, we've learned a lot by this episode. That's the thing that we should all keep in mind: that we can't do everything. It's not just monetary policy, but monetary policy is a big part of it. It's getting more complicated, that's for sure; but the more that we can keep it as simple as possible so we can communicate to everyone, to markets, to people out there in general, the better off we'll be. Thank you.

English: Thanks very much. We've been given some special dispensation to run a little bit long, which is good because otherwise we'd be out of time. So let me turn to questions from the floor. One that seems to have gotten a lot of interest from folks is, how do we think about policy lags in the current context? Particularly if we're thinking about the Fed, at some stage, turns to easing, but when, and how do you think about when it's appropriate to move rates?

This may tie in with the point Kristin made about gradualism versus front-loading. How gradual do you want to be on the downside? Do you worry about lags and ending up overdoing things, and ending up with a recession that you don't need? Kristin, do you want to take a shot at that?

Forbes: It's a good point; that's also part of the argument. I sort of did the pros and cons of gradualism versus more front-loading, versus a more aggressive approach. Gradualism, the downside is you may get behind the curve, inflation expectations pick up too much, and then your job is harder, but the advantage of gradualism is then you can wait for some of those lags to play out, so you will be less likely to overdo it and raise too much.

If we meet next year, when this conference meets, maybe gradualism will have been the right approach, going more slowly so that monetary policy did hit and have a bigger effect. Countries like the US that went more aggressively might have done too much. Yes, the lags are an issue; that's one of the main reasons why countries really did support gradualism and would like to be more gradualistic.

That's a fair concern, and that's a good argument for doing it more slowly. Let the lags play out, feel it out when you do not have an inflation problem and when you can trust inflation expectations will remain anchored. But this time around, the risks of going too slowly were too great. It was worth moving faster. Here again, I was sort of critical of the Fed for being a little slow to raise rates, but then they did make up for it. Once they started, they got on with it.

Now, the trick will be: will the central banks that raised more aggressively, went more quickly to where they think they're going to end, will they have the confidence to then hold, and let the lags play out? Or if the data keeps coming in showing inflation surprises, will they feel like they need to keep hiking, so then by the time the lags hit, they've overdone too much?

We here, we just don't know yet. It seems that so far at the Fed there does seem to be some confidence of pausing now, let the lags play out and see. But we just don't know, that's part of the tradeoff.

English: Does anyone else want to comment on lags?

Taylor: Just a quick thing: the fact that simple rules don't seem to have lags is just an indicator that they came out of very complicated models. I mean, incredibly complicated. Even lags, there's a lagged inflation rate, there's a simple way to do it, and it wasn't just mindless, it was simulating gigantic models, international models. A lot of it came out automatically. The lags are there. Maybe not the way you'd like it.

English: Okay, thanks very much. Another question that maybe is an interesting question to think about going forward is: do we think that the narrative from the Fed staff that we're going to have a mild recession—from Jan, that maybe we don't have a recession at all, maybe there's a soft landing—is that dangerous in the sense that it communicates to markets that a recession can be avoided or can be painless, it can be really small? Do we expect that we can actually deliver a small recession, or do we get worried that there are non-linearities of the sort that the Sahm rule or something like that would pick up—that you slow the economy enough, you end up with a big recession?

How do we think about talking about little recessions? Is that something that is really hard to convey? Is it risky to convey because we may well end up with a bigger recession? I don't know, Jan, do you want to talk about that a little bit?

Hatzius: Sure. Obviously, the staff forecast hopefully is driven by the best judgment, as opposed to what would be better to convey or worse to convey. There are strong arguments for the idea that if you do have a recession, contrary to my baseline forecast, that it would be on the milder side. One reason is reduced multiplier effects working through the labor market in an environment where excess labor demand is still quite high. If you see a downturn in output growth, you can show that when the level of job openings is higher and you bucket the regression analysis conditional on high versus low job openings, that the multiplier effects working through employment are smaller, not surprisingly.

That's a reason for a relatively mild recession if there is one. The other reason is that if you are in an environment where you're getting help from other drivers of inflation that are not dependent on opening up a large output gap—inflation is already coming down, in part because of supply-side reasons—then I suspect you could also respond to a downturn by easing monetary policy pretty aggressively if needed. Obviously, it's all conventional easing at this point. There are 500 basis points to go if you really need it, so that probably also reduces the risk of a really serious downturn. That said, forecasting is hard, and of course there is a range of outcomes. If you are in a recession, you should certainly not take the idea that it's going to be deeper off the table.

English: Okay, thank you. We can only take one more question, so maybe for both Troy and for John. There's a question on how should monetary policymakers appropriately take account of the problems in the banking sector in thinking about monetary policy? For John, I guess you have a rule, but maybe you're going to deviate from a rule if you think you have a real banking sector problem. But how do you think about that? How much?

Taylor: Well, we had a discussion earlier today about that, from Mervyn. I agree with his approach, and that's that you don't want to have this automatic bailout. You want to have a policy in place. It's like a rule. You don't have to call it a rule, but it's a process that seems to me has worked well when it's been tried, and we've run into little problems.

It's not going to eliminate all these problems, and to some extent it's separate from monetary policy, and if you get monetary policy right, it'll be easier. The very low rates we had may have had some significant effects on some of the banks.

English: Sorry, but I was thinking about, given that we have a banking problem that could potentially lead banks to pull back a lot from lending, for example, that would have macroeconomic consequences. How do you take account of that if you think about applying a rule to monetary policy, does it cause you to deviate and if so, how do you calibrate the deviation?

Taylor: No, there's not much reason to deviate. You have the term structure of interest rates; it takes from the short rate and expected future short rates, to the long rates. You have your question about the exchange rate; exchange rate is not there. It's maybe more relevant for certain countries, but it's not there. You have a target inflation that is 2 percent, and you react to deviations of inflation from that target by a sufficient amount that you bring it back to the goal.

It has all the elements and you don't know the exact parameters, but that automatically deals with at least that part of it. I'm not saying the financial system doesn't need some other work, some regulatory work, but that's it.

English: Okay. Troy, do you have thoughts on this?

Davig: I would just say, don't do anything to exacerbate the deposit outflow. It is a new world when markets are focused at 4:15 every Friday on the H.8 data, right? I mean, that tells you something. I'll just go back to my comment that any small tweak, small adjustment, is worth doing. Again, ON RRP versus IORB, something like this to make deposits marginally more sticky, may help because as we well know, bank runs are super nonlinear so you don't slowly slip into it, you want to try to do all you can to keep as far away as you can from anything that could trigger another bank run.

It does look like there are, still as in my chart, there are banks under a lot of pressure. It is a time just to think carefully about what can be done to relieve some of the deposit outflow pressure. That is part of the monetary transmission mechanism, so that's to be expected. Right now, given some of the stresses, it's just an important time to be thinking about steps to not exacerbate that pressure.

English: Okay. I'm afraid we're out of time. Lots to talk about, and we'll be talking about it today and tomorrow at this event. Thanks very much, and thanks to the panelists.