Monetary Policy Normalization: Why Normalize?

Stephen Williamson, a vice president and economist with the St. Louis Fed, compared the current state of the economy with that of February 2005 (the last time prior to the recession that the unemployment rate was 5.4 percent), and he looked at whether low interest rates translate into high inflation. He also discussed the monetary policy normalization process as laid out by the Federal Open Market Committee and implications of normalization for consumers and businesses.

Presentation (PDF)

• Part 1: Introduction
• Part 2: The Bank of England
• Part 3: Why the Fed Was Founded
• Part 4: The Tools of the Fed
• Part 5: Monetary Policy around the Great Recession
• Part 6: Normalization of Monetary Policy
• Part 7: Audience Q and A

Transcript:

Stephen Williamson: Okay, so the question is, why would you normalize? You might think, well, you don't look at these pictures, and think, well things don't look so bad. Your inflation's a little low, but what the heck? Unemployment rates come down, economy looks okay. So why do anything different? So one argument for doing something different, for normalizing, would be that—so here I just took an example. Here's an example. So go back to the last time we saw 5.4 percent inflation. See, that's what we have now. Previous date, when we had 5.4 percent inflation was February 2005. Look at the state of the economy then. So we had inflation, PCE inflation, that's the raw inflation number. That was above the target. It's above 2 percent, but not by so much. Core inflation, when you strip out food and energy, that's 2.2 percent. Close to 2 percent. Fed funds rate at the time was 2.5 percent, and going up. So it was going up, and a year later, so it was up above 5 percent. So currently, this is the actual Fed funds rate. Again, close to zero. So one argument might be, well if you looked at how the Fed behaved prior to the recession, and if we were consistent with that behavior, interest rates would be higher now. That's one argument for normalizing. So the argument based on the fact that, well, we're doing the right things in the past, we think. And if we're consistent with that, we'd be up above zero now.

Okay, here's another argument. This is a little more idiosyncratic. So this is kind of where my thinking comes into this. So here's another argument that says, okay. So you might think, well, low interest rates might mean high inflation. That might be a traditional way of thinking about it. You know, that low interest rates mean a kind of stimulative monetary policy, and that might make inflation higher. But if we look at the data, you know, so this is reflecting long-term trends. In terms of long-term trends, here's the Fed funds rate here, on this axis. Here's the inflation rate, on that axis. What do you see in the picture? What you see is that a high Fed funds rate tends to be associated with high inflation. So if you're thinking a world where inflation is hovering around 2 percent all the time, you'd be—here's 2 percent inflation. You'd kind of be in here. You know, you'd be like 3-5 percent Fed funds rate, on average. So one argument might be, well if you want to be back to normal, you want to hit your 2 percent inflation target forever, you're going to have to see higher interest rates. So this doesn't come naturally, necessarily.

But if I show you this next picture—here's Japan. So Japan is a country where they've had 20 years of experience with low interest rates. So the blue line here is the—it's not exactly like the Fed funds rate, but this is—the blue line is the—that's the policy interest rate. The interest rate target of the Bank of Japan. So since 1995, when it kind of comes down here, mid-1995, it goes down below 1 percent, and it's been between zero and 1 percent for 20 years. So what kind of inflation do you see? The inflation is this dotted red line. Average inflation over that period is zero. And it kind of fluctuates, right? So sometimes it goes up above zero, goes up 2.5 percent, it's down to -1.5 percent, it's down to -2.5 percent and it moves around. But average is about zero. So what happens in an economy with low interest rates, is this is what we see. What we see is an economy with low-flation. Not necessarily deflation, but low inflation. So if you want to hit a 2 percent inflation target, maybe nominal interest rates have to be up above zero somewhere. That's the second argument. Okay, so what's normalization about? So it's—you can read about it. This is posted on the Board of Governors website, you can find it there. There's a set of—it's under policy normalization principles and plans. So a bunch of bullet points on what happens with normalization.

So part of normalization is getting the Fed funds rate off zero. We call that liftoff. Two is to end reinvestment. What reinvestment is, is that we made all these purchases of assets, long-maturity assets, mortgage-backed securities and Treasury Securities. Reinvestment is as those assets mature, the New York Fed keeps replacing them. So the first step, as laid out in the plans, is that liftoff happens. Second step, sometime later, is that reinvestment ends. And then the balance sheet starts to shrink. And then ultimately, what the goal is, to return the balance sheet to a state that looked like, you know, the picture I showed you earlier, where you're close to zero reserves. Small amount of reserves, relative to now in the system. And how long would that take? Well, some economists at the board have studied this. A couple of times over, they've studied it. And the estimate they come up with is about seven years. It depends on assumptions about various things, which have to do with this chart. So the balance sheet, how's the thing going to shrink? Well, there are two ways it happens. One is these assets are going to mature. You could sell the assets, right? You could sell the assets on the balance sheet, but that's not in the plans.

The plan is that the assets mature. As they mature, there's a mechanism by which this works, that it's just going to happen—it happens that way. The assets mature, and then the reserves just—the reserves, effectively, disappear as the assets mature. So as these asset quantities go down on this side of the balance sheet, reserves go down on that side of the balance sheet. There's this other thing in here, I'm going to talk—I told you what a repo was. A reverse repo is, as you might imagine from the Fed's point of view, is just the opposite. So a repo is when the Fed lends to some participant in the financial market, usually, you know, a big financial institution. And a reverse repo is when somebody lends to the Fed in the market for repos, in the repurchase agreement market. I'll tell you a bit more about that, in the next slide or two.

Okay, and the other thing that happens is, you know, currency is just kind of driven by the demand for the stuff. And as the demand for currency grows over time, and it's still growing. You may find this surprising, probably most of you don't use it as much as you used to, but the world demand for U.S. currency just keeps growing. So as it grows, you know, it's like given the asset side of the balances sheet—as currency grows, the other stuff has to shrink. Because the balance sheet has to balance. But that happens slowly. That happens more slowly than, you know, the maturing of the assets.

Complications—well, one complication is reserves now bear interest. It's just—it's lower now. It's a quarter percent. But that affects how you think about the normalization. In theory, that should make interest rate control easy. It should make interest rate control easy, because the interest rate on reserves is—that should take overnight interest rates, but it actually doesn't. It doesn't. Usually because, as I say here, U.S. financial systems are a complicated animal, and that makes it not so easy to directly control market interest rates, just using the interest rate on reserves. That's where this reverse repo facility comes in. So again, a reverse repo is like—it's like a short-term loan to the Fed, from the financial market. And they're really—they're very much like reserves. You think of it as just another Fed liability. It's like reserves, except other financial institutions can hold those things, that can't hold reserves. That expands the reach of the Fed's liabilities, and kind of makes interest rates easier to control.

Here's the problem. The problem is, that in theory, in a perfect financial market, 0.25 percent, which is the interest rate on reserves, should be the Fed funds rate. But it's not. You can see what happens here. This is going back to, like, 2009 to the present. And the Fed funds rate here, Fed funds trade below the interest rate on reserves. And we're just beginning to understand why, you know, fair amount of thinking going into what goes into explaining the difference between the Fed funds rate, and the interest rate that banks get on reserves. But the idea is that this reverse purchase agreement facility will effectively give the Fed more control. So there's effectively another interest rate in here now. The other interest rate is this reverse repo rate, or ONRRP rate, for people who think about these things. So that interest rate would put a floor under this thing. So what happens then is that the Fed funds rate gets hemmed in by the interest rate on reverse repos and the interest rate on reserves. As all interest rates go up, we can control the Fed funds rate as we want. That's kind of mechanics.

This is reverse repurchase agreement. So you can see what's going on here. It's not a new thing, because the Fed has these, you know, back here, you know. But they begun to use them more. This is like experimentation. This is the New York Fed figuring it out. Figuring out how these things work. The mechanics of how they trade with the market, and what kind of effects—how they set the interest rate on these things matters, et cetera. But that will matter. That will start to matter. Once liftoff happens, that thing will matter.

Okay, how does this matter for anybody, you know? How does normalization matter for anybody? For people trading in financial markets, it does. Because you know, if you were in the repo market, for example, the Fed will be a participant, a big participant you'd want to think about it. You know, there's this idea that in terms of where we are now, interest rates, which is what affect most of us, have nowhere to go but up. Have nowhere to go but up. Sometime, liftoff will happen, and that will matter to everybody. That will require an adjustment, but all the policy statements tell us that the process of normalization will be a gradual thing. And there shouldn't be big surprises. You'll have lots of warning. Normalization is an end of itself. It's not like we just like normal, or something. Our primary concern is the dual mandate. That's what we care about. That's our assignment from Congress. We pay attention to that. So you know, normalization is a vehicle for dealing with the dual mandate effectively.

Okay, so probably—lots of people ask this question. When will liftoff happen? I don't know. We don't know. So it's not like this is a secret. We know it, and we're not telling you. We don't know. And we—and for good reason. So the idea is that policy is always, or should be data dependent. Depends on the state of the economy. You look at the state of the economy you decide what to do. So what happens for the rest of this year, what the FOMC decides, depends on the views of the FOMC participants, and people who go to Washington, make these decisions about what the data tells us, what's the state of the economy, what should we do? And we don't know. You know, we don't know between now, and the fall. We've got forecasts, and you know, we have ideas about what's going to happen, and we've studied it hard, but you get surprises. We can't tell you, because we don't know. So the large Fed balance sheet, I'm convinced, that's no threat to price stability. Not at all. Don't worry about that. So the Fed's actions, like I said, they're determined by the—with the dual mandate in mind, determined by what the FOMC thinks is best, given the data. So that's it. So thank you.