Posts Tagged ‘scott sumner’

How do you do it?

16 November 2011

Count me among the skeptics who believe the Fed has pretty much already done all it can to pull the economy out of the deep hole that it’s in. Zero short-term interest rates, purchases of longer-term bonds to keep long-term rates at historic lows, backstopping various asset markets, emergency loans to banks, etc. It’s helped avert a Second Great Depression, which is nothing to sneeze at. Some economists who I usually agree with are convinced that aggressive new policies could pull us out of the current Little Depression, too. They’re smarter than I am, but they have yet to convince me that these policies could work.

The tonic du jour is nominal GDP targeting, by which the Fed would try to reach a certain level of nominal GDP — say, $16. 3 trillion (the current level of potential GDP assuming that, as I’ve read, current GDP is 7% below its potential. Do the math and that’s a $1.1 trillion gap between current and potential GDP). Christina Romer, Obama’s first head of the Council of Economic Advisors, recently backed this approach in a New York Times op-ed. Scott Sumner has been pushing it all along, and there’s now a whole new school of macroeconomics, “market monetarism,” which revolves around nominal GDP targeting. (Economists: see here for Ed Dolan’s helpful explanation of how nominal GDP targeting is a form of Milton Friedman-style monetarism.)

Now, once the Fed announces this new target, how does it actually get there? Romer provides the clearest answer I’ve seen yet:

‘Though announcing the new framework would help, it probably wouldn’t be enough to close the nominal G.D.P. gap anytime soon. The Fed would need to take additional steps. These might include further quantitative easing, more forceful promises about short-term interest rates, and perhaps moves to lower the exchange rate. . . .’

‘Nominal G.D.P. targeting would make it more likely that the Fed would take these aggressive actions.’

That’s clear, but so is weak tea. None of these actions sound all that different from what the Fed is already doing. Proponents of nominal GDP targeting seem to be counting on a huge “announcement effect,” i.e., that people will hear about the Fed’s commitment to raising GDP and will assume that Fed will make it happen. Yet the Fed’s goals already include maximum sustainable employment, which is the employment rate you’d have at potential GDP, so why should this change the public’s behavior? (Although there is a difference between monetary policy goals, like low unemployment, and targets, which now include interest rates, it’s a rather subtle one. I don’t see why it would move markets.)

Another popular tonic is a higher inflation target. Right now the Fed’s unofficial but almost universally acknowledged inflation target is 2%, and for the past few years the core inflation rate has been below or near 2%. When inflation is very low, real interest rates (nominal interest rates minus inflation) can still be high even when nominal rates are also low. In the U.S. in the early 1930s, for example, nominal rates plunged toward 0%, but deflation was raging, so real interest rates were actually quite high. Economic historian Nick Crafts, in a Financial Times op-ed, says that Britain’s recovery from the Great Depression was greatly aided by a combination of low nominal interest rates and rising inflation rates — i.e., negative real interest rates — which promoted homebuilding. Crafts says targeting a higher inflation rate — say, 4% — could do the trick today.

Again, I just don’t see how you get there. Would I like to see lower real interest rates? Sure. But for 4% inflation to happen, a lot of other things have to happen first. Banks need to loan out their excess reserves, people and businesses need to buy stuff with those loans, the money needs to be redeposited in banks,  more loans need to be made, etc. That’s how monetary policy works — when it works. Right now, the banks have over a trillion dollars in excess reserves that they’re just sitting on. Banks are not eager to lend, and businesses and households are not eager to borrow. Classic liquidity trap.

Nominal GDP targeting and higher inflation targets sound radical, but are they? Chicago Fed President Charles Evans said in a speech this week that he viewed the 2% inflation target as a medium-run target, not a short-run target, saying that as long as inflation averaged out to 2% over a multi-year period, higher inflation rates would be acceptable in the short term. That statement is consistent with either a nominal GDP target (shoot for low inflation when real GDP is high, tolerate higher inflation when real GDP is low) or an inflation target (let inflation rise when unemployment is high), which suggests that neither of those policies is all that new. Both seem to promise much more than they could ever deliver.

Advertisements

Dead man’s curve?

3 August 2011

Scott Sumner’s blog The Money Illusion has two provocative posts today that argue that Federal Reserve policy is currently too tight, despite the near-zero fed funds rate, and cite as evidence the extremely low interest rates on 5-year Treasury notes. I’m going to grapple with his monetary policy argument at some later date (the gist of it seems to be that the Fed should target nominal GDP and therefore combat the current slump with much bigger increases in the money supply than we’ve had). For now I want to focus on those 5-year bond rate data, which are bad news indeed.

The most recent 5-year bond yield, as of yesterday, was 1.26%. That’s close to a historic low. What does it mean? The standard interpretation in a money and banking course is that interest rates on 5-year bonds are the average of current and expected future short-term bond rates over the next 5 years (since a close substitute for a 5-year bond is a series of short-term bonds held over the same span), with some allowance for people’s preference for short-term bonds as more liquid (you get your money back sooner) and less risky (you don’t have to worry about market interest rates shooting up after you’ve bought your bond and then being stuck with a subpar yield for a long time). So, the number suggests that people are expecting very low short-term bond yields over the next five years. Short-term interest rates are a function of two big things: the state of the economy (they’re lower in economic slumps) and the expected inflation rate (when inflation falls, lenders and bondholders will accept lower interest rates; and inflation also tends to be lower in economic slumps). And in the case of T-bonds, the interest rates reflect the jitters of worldwide investors — the more nervous investors are about stocks and corporate bonds, the more likely they  are to flee to the safety of Treasuries.

So, then, a near-record-low 5-year T-bond rate means investors are expecting (1) economic weakness for the next 5 years and/or (2) low inflation for the next 5 years and/or (3) investor anxiety for the next 5 years. I’d vote for all of the above. And (2) and (3) are common symptoms of (1).

To see just how low these recent 5-year Treasury yields are compared with those of the past nine years, see this Dynamic Yield Curve, which shows the interest rates on T-bonds of different maturities.  If you click Animate, yield curves from 2003-present flash by and you’ll see that that the norm for 5-year rates is about 3 to 5%. So the bond market apparently expects the economy to be way below average for the next five years. Even the 10-year bond rate was only 2.66% yesterday (it’s normally above 4%), so a ten-year depression is not only possible but maybe even probable, according to the market.

Fed up with Bernanke?

31 July 2011

Greg Mankiw has a good column in today’s NYT in defense of embattled Fed Chair Ben Bernanke. How embattled is Bernanke? Mankiw notes an (admittedly unscientific) online CNBC poll from June, in which the question was “Do you have confidence in the way Ben Bernanke is handling the economy?” 95% of respondents answered no.

Mankiw says the Fed has done basically all it can to combat the Little Depression (unfortunately “all it can” is not enough), while steering clear of high inflation. The core inflation rate in recent years has been just 2%, widely believed to the Fed’s unofficial target inflation rate. Mankiw suggests making that 2% target official, but otherwise sees no obvious room for improvement in Bernanke’s performance.

I tend to agree that Bernanke’s Fed has done about all that monetary policy can do here, but Scott Sumner, one of the more interesting monetary thinkers I’ve come across lately, says the Fed actually has a lot more ammunition in its arsenal and compares the situation to the early 1930s, when the Fed increased the monetary base but needed to do a lot more to stem the massive tide of bank failures and monetary collapse. Unfortunately, I’ve yet to find the specifics of his argument, but I’ll share them with you when I do.

Sumner, by the way, loves the idea of a 2% inflation target and even suggests that Mankiw be appointed to the Fed’s Board of Governors. Maybe Mitt Romney (to whom Mankiw is an adviser) can do that next year.

Why is the Fed still paying interest on reserves?

12 November 2010

Matt Yglesias, channeling Scott Sumner and Louis Woodhill, makes a good case that the interest rate on bank reserves, which was 0% up until just a couple years ago, should be lowered from its current 0.25%.  He suggests lowering it to 0.15%; I’d go lower, to 0.10% if going back to zero is out of the question.

Paying interest on reserves made some sense back in 2008 when the Fed was flooding the system with reserves in order to prevent a deflationary catastrophe.  The fear then was that when the economy picked up, banks would start loaning those reserves out and unleash a huge inflation; to prevent that, the Fed put an interest rate on reserves that could be raised whenever it became necessary to “soak up” those reserves.  But nothing like that is happening now — instead we have a banking system with about $1 trillion in reserves that they’re not loaning out, and the amount is likely to grow as the Fed makes its monthly $75 billion purchases of longer-term bonds under QE2.  The string the Fed is pushing on ought to move a little more if the interest rate on reserves were closer to zero.  0.25% might not sound like much, but it’s more than the federal funds rate on any given day and more than the short-term Treasury bill rate.  If banks could only earn 0.10% on reserves, I think they’d be more likely to loan them out, i.e., monetary policy would be more likely to work.

When the recovery finally shifts into high gear (and it could be sooner than most of us think, considering all the “green shoots” among leading indicators at present) and banks start loaning out those reserves, then the Fed can raise the interest rate on reserves.  But keeping it this high now gives preemption a bad name.