Wednesday, April 22, 2015

Was Monetary Policy Loose During the Housing Boom?

Did the Fed's set its policy interest rate rate below the market-clearing or 'natural' interest rate level in the early-to-mid 2000s? Or did it simply lower its policy interest rate down to a depressed natural interest rate level during this time? The answers to these questions determine whether U.S. monetary policy was loose during the housing boom. 

John Taylor believes the Fed pushed interest rates below their natural interest rate level. He views this departure from a neutral stance as a key contributor to the housing boom. Ben Bernanke and Larry Summers believe otherwise. They see the Fed simply doing its job back then by adjusting its policy rate down to a low natural interest rate level. Bernanke believes the natural interest rate level was low because of a saving glut while Summers holds that its was depressed because of secular stagnation. Either way, both individuals do not blame the Fed for any role the low interest rates played in fostering the housing boom. The Fed''s lowering of interest rates was simply an endogenous response.

George Selgin, Berrak Bahadir, and I recently published an article that lends support to John Taylor's view of Fed policy during this time. It received some pushback from Scott Sumner who is sympathetic to both the saving glut and secular stagnation views. At the same time, Tony Yates provided a critique of John Taylor's argument on the financial crisis that was heartily endorsed by Paul Krugman. So the debate over the Fed policy during this period continues.

What I want to do here is to step back from this debate and review what I see as the key economic developments that affected U.S. interest rates at this time. Then, given these considerations, I will jump back into the debate and ask whether Fed policy pushed interest rates in the same direction as that implied by these developments.

The key developments as I see them are threefold: a falling term premiums, a spate of large positive supply shocks, and the emergence of a monetary superpower. Let us consider each one in turn.
I. Falling Term Premiums on Long-Term Treasuries
The term premium is the extra compensations investors require for the risk of holding a long-term treasury bond versus a sequence of short-term treasury bills over the same period. The term premium had been declining since the early 1980s and therefore put downward pressure on long-term interest rates. This development can be seen in the figure below which is created using the Adrian, Crump, and Moench (2013) data. (For more on this data see here.)


The decline has been attributed to several factors. First, there was a decline in inflation volatility and an overall improvement in macroeconomic stability during this time that made investors less risk averse to holding long-term bonds. They therefore demanded less compensation. Second, regulatory and accounting changes for certain firms increased their demand for treasury securities relative to their supply. This further reduced the term premium. Third, globalization was taking off, but without a concurrent deepening of financial markets in many of the affected countries. That meant that global income was growing faster than the world's ability to produce safe assets. Consequently, many developing countries started turning to the United States for safe assets. This further depressed the term premium and is the basis for Bernanke's saving glut theory.

A close look at the above figure shows this term premium decline intensified in 2003, falling about 1.4 percentage points over the next two years. This happened right during the time the Fed pushed its policy rate to record-low levels. This, then, appears to support the endogenous view of the low policy rates argued by Bernanke and Summers.

However, this conclusion needs to be tempered. For the next two developments discussed below suggest that a sizable portion of the declining term premium at this time may have been an endogenous response to the Fed's low interest rates policy during that time. 

II. A Spate of Large Positive Supply Shocks
The second important development is that the global economy got buffeted with a series of large positive supply shocks from the  opening up of Asia--especially China and India--and the rapid technology innovations that reached a crescendo in the early-to-mid 2000s. Global growth accelerated because of these developments as seen in the figure below:



The opening up of Asia significantly increased the world's labor supply while the technology gains increased productivity growth. The uptick in productivity growth, which peaked between 2002 and 2004, was widely discussed in the early 2000s and raised long-run expected productivity growth at the time. This can be seen in the figure below which shows a consensus forecast of annual average productivity growth over a ten year horizon:


Note that the rise in the labor force and productivity growth rates both raised the expected return to capital. The faster productivity growth also implied higher expected household incomes. These developments, in turn, should have lead to less saving and more borrowing by firms and households and put upward pressure on the natural interest rate. Interest rates, in short, should have been rising given these large positive supply shocks during this time. 

III. Emergence of a Muscle-Flexing Monetary Superpower 
The third development is that in the decade leading up to the financial crisis that the Fed became a monetary superpower that could flex its muscles. It  controlled the world's main reserve currency and many emerging markets were formally or informally pegged to dollar. Thus, its monetary policy got exported across much of the globe, a point acknowledged by Fed chair Janet Yellen. This meant that the other two monetary powers, the ECB and the Bank of Japan, were mindful of U.S. monetary policy lest their currencies became too expensive relative to the dollar and all the other currencies pegged to the dollar. As as result, the Fed's monetary policy got exported to some degree to Japan and the Euro area as well. Chris Crowe and I provide formal evidence for this view here as does Colin Gray here.
Now let's tie all these points together and see what it says about the Fed's role in the housing boom. Let's begin by noting that when the large positive supply shocks buffeted the global economy they created disinflationary pressures that bothered Fed officials. They did not like the falling inflation. So Fed officials responded by easing monetary policy. Recall, though, that the supply shocks were raising the return to capital and expected income growth and therefore putting upward pressure on the natural interest rate. The Fed, consequently, was pushing down its policy rate at the very time the natural interest rate was rising. Monetary policy was inadvertently being loosened.

This error was compounded by the fact that the Fed was a monetary superpower. The Fed's easing in the early-to-mid 2000s meant the dollar-pegging countries were forced to buy more dollars. These economies then used the dollars to buy up U.S. treasuries and GSE securities. This increased the demand for safe assets and ostensibly reinforced the push to transform risky private assets into AAA assets. To the extent  the ECB and the Bank of Japan also responded to U.S. monetary policy, they too were acquiring foreign reserves and channeling  credit back to the U.S. economy.  Thus, the easier U.S. monetary policy became the greater the demand for safe assets and the greater the amount of recycled credit coming back to the U.S. economy.  The 2003-2005 decline in the term premium, in other words, was to some extent an endogenous response to the easing of Fed policy during this time.

The figure below highlights this relationship for the period 1997-2006. It comes from my work with Chris Crowe and shows that almost 50% of the foreign reserve buildup was tied to deviations of the federal funds rate from the Taylor Rule. Colin Gray estimates several regression models on this relationship and finds that for every 1% point deviation of the federal funds rate below the Taylor Rule, foreign reserves grew by $11.5 billion. The Fed, therefore, was putting downward pressure on interest rates not only directly via the setting of its federal funds rate target, but also by raising the amount of credit channeled into the long-term U.S. securities.


Given these points, I think it is reasonable to conclude the Fed contributed to the housing boom. I hope they give Scott, Tony, and Paul something to think about.

Let me be clear about my views Even though the Fed kept its policy rate below the natural rate for a good part of the housing boom period, the opposite happened after the crash due to the ZLB. This is a point Ramesh Ponnuru and I made in a recent National Review article. So unlike some observes who see the Fed as being eternally loose, I take a different view: the Fed was too loose during the boom and too tight during the bust. 

Update: There are multiple measures of the output gap that show the U.S. economy overheating during this time. Below is a figure from this article that compares the real-time and final measures of the U.S. output gap. Everyone shows ex-post an overheating economy during the housing boom:

Friday, April 17, 2015

It Takes A Regime Shift to Raise an Economy

This week we learned that Ben Bernanke does not view NGDP level targeting, price level targeting, or a higher inflation target as the best way to deal with the zero lower bound (ZLB) problem. Now I believe the "what to do at the ZLB" debate is becoming moot as the U.S. economy improves, but it is interesting to consider Bernanke's thoughts on these alternative approaches to monetary policy. Here he is questioning their usefulness at the ZLB relative to his preferred approach:
The second possible direction of change for the monetary policy framework would be to keep the targets-based approach that I favor, but to change the target. Suggestions that have been made include raising the inflation target, targeting the price level, or targeting some function of nominal GDP... a principal motivation that proponents offer for changing the monetary policy target is to deal more effectively with the zero lower bound on interest rates. But economically, it would be preferable to have more proactive fiscal policies and a more balanced monetary-fiscal mix when interest rates are close to zero. Greater reliance on fiscal policy would probably give better results, and would certainly be easier to explain, than changing the target for monetary policy.
So Bernanke wants the Fed to keep its inflation target of 2% and complement it with more aggressive use of fiscal policy when up against the ZLB. It sounds reasonable, a more balanced mix of monetary and fiscal policy. What could possibly go wrong? A lot, actually, if you believe this approach would have generated substantially greater aggregate demand growth over the past six years.

Here is why. Fiscal policy can only create aggregate demand growth up to the point it pushes inflation to the Fed's 2% target. The Fed's preferred inflation measure, the PCE deflator, averaged 1.4% since 2009. That means fiscal policy would have had 60 basis points on average with which to work over the past six years in closing the output gap. Do we really think that would be enough for the level of aggregate demand shortfall experienced over this time?

What was needed was a monetary policy regime shift, one that would have allowed, if needed1, a permanent increases in the non-sterilized portion of the monetary base to spur rapid growth in total dollar spending. It was never going to happen with a 2% inflation target. Imagine, for example,  the U.S. Treasury Department sent $5000 checks to every household. As individuals began to spend their new money inflation would start rising, but it could only rise by 60 basis points before the Fed started tightening policy. The 'helicopter drop' would be stillborn.

As a counterexample, consider a regime shift to a price level target. This is not my preferred approach, but it illustrates well that the type of monetary regime needed to restore full employment over the past six years was far from a 2% inflation target. Here is how I describe it from an earlier post:
One [such] monetary regime change would be a price level target that returns the PCE to its pre-crisis trend path. To return the PCE to its targeted path would require a temporary burst of higher-than-normal inflation. The expectation of and realization of this inflation burst would be the catalyst that spurred robust aggregate demand growth.
Now let us pretend the Fed actually implemented a price level target back in 2010 when it began QE2. Specifically, imagine the Fed had made QE2 conditional on the PCE returning to its 2002-2008 trend path. The figure below shows this scenario with three different paths back to the price level target. Note that each path represents differing rates--5%, 4%, and 3%--of 'catch-up' inflation and for each path there is a significant amount of time--16 months, 26 months, and 49 plus months--involved to catch up to trend.


What this illustrates is that to get the kind of robust aggregate demand growth needed to close the output gap back in 2010, there needed to be a sustained (but ultimately temporary) period of higher-than-normal inflation. Doing more fiscal policy to squeeze out the last 60 basis points of the Fed's 2% inflation target would not cut it...it would be tinkering on the margins. If fiscal policy really wanted to close a large output gap at the ZLB it too needs the support of monetary regime change.
Again, I think the ZLB debate is becoming moot for the U.S. economy. But this discussion highlights why I believe greater use of fiscal policy over the past few years would not have heralded a stronger recovery. The U.S. economy has been chained to a monetary regime that while effective in anchoring inflation expectations also served to stymie a full-throttle recovery. If you want fiscal policy to work, you have to have a monetary regime that works. This was the core issue, not the degree of monetary-fiscal policy mix.

So yes, a regime shift was needed over the past six years. I would like to have seen a shift to a NGDP level target. Not only would it addressed the ZLB problem, but it also deals with the knowledge problem, financial stability, and supply shocks better than a price level or higher inflation target. For an accessible discussion of these points see my piece with Ramesh Ponnuru on the right goal for central banks. For a NGDP proposal that utilizes a monetary-fiscal mix see this post.
1A permanent increase in the monetary base may not be necessary if the the regime shift causes the velocity of money to sufficiently change. The actual expansion may not be needed or be very small if this commitment is credible. To see this, imagine the Fed targets the growth path of nominal GDP (i.e. a NGDP level target). If the public believes the Fed will permanently expand the monetary base if NGDP is below its targeted growth path, then the public would have little reason to increase their holdings of liquid assets when shocks hit the economy. That is, if the public believes the Fed will prevent the shock from derailing total dollar spending they would not feel the need to rebalance their portfolios toward safe, liquid assets. This, in turn, would keep velocity from dropping and therefore require minimal permanent monetary injections by the Fed to hit its NGDP level target.

Tuesday, March 31, 2015

It's Already Priced Into the Market

Former Fed chair Ben Bernanke blogged about secular stagnation today and he's not buying it:
Does the U.S. economy face secular stagnation? I am skeptical, and the sources of my skepticism go beyond the fact that the U.S. economy looks to be well on the way to full employment today. First... at real interest rates persistently as low as minus 2 percent it’s hard to imagine that there would be a permanent dearth of profitable investment projects. As Larry’s uncle Paul Samuelson taught me in graduate school at MIT, if the real interest rate were expected to be negative indefinitely, almost any investment is profitable. For example, at a negative (or even zero) interest rate, it would pay to level the Rocky Mountains to save even the small amount of fuel expended by trains and cars that currently must climb steep grades. It’s therefore questionable that the economy’s equilibrium real rate can really be negative for an extended period.
His successor, Janet Yellen, is also skeptical. She acknowledged in a speech the possibility of secular stagnation, but believes it is an unlikely outcome. Her baseline scenario is for the U.S. economy to continue to recover and, as a consequence, to continue to pull up the real equilibrium interest rate:
[T] economy's underlying strength has been gradually improving, and the equilibrium real federal funds rate has been gradually rising. Although the recent appreciation of the dollar is likely to weigh on U.S. exports over time, I nonetheless anticipate further diminution of the headwinds just noted over the next couple of years, and as the equilibrium real funds rate continues to rise, it will accordingly be appropriate to raise the actual level of the real federal funds rate in tandem, all else being equal.
If we take the New York Fed's estimate of the 10-year treasury term premium and risk-neutral nominal rate as given, then the market is already pricing in a non-secular stagnation future as seen in the 10-year real risk-free treasury yield below:


The red line is a risk-adjusted measure of the expected average real short-term interest rate over the next ten years. Put differently, this measure reveals the expected path of real short-term interest rates and it is pointing up. Since it is risk free, it should only be reflecting the expected fundamentals of the real economy over the next 10 years (see here for further explanation). It started rising in early 2014 and now has been positive for about five months. That bodes well for the economy.

Now the real risk-free 10 year yield seems to have stalled a bit in the past three months, but overall it appears to be on it way to its pre-crisis trend that was discussed in my previous post. Dare I say history is repeating itself when it comes to secular stagnation?

Monday, March 30, 2015

Ben Bernanke and the Secular Stagnation Debate

Ben Bernanke is back and mulling over low interest rates. From the Washington Post:
Blogging isn't dead. At least, Ben Bernanke, former chairman of the Federal Reserve, doesn't think so: He's now blogging at the Brookings Institution. In his first post, he says he wants to write about this fascinating chart, which shows the steady decline in interest rates over the past 30 years or so.
The question of why interest rates keep falling is an important one these days. Former Treasury Secretary Larry Summers has argued that declining real rates are a symptom of secular stagnation, meaning the global economy just isn't what it used to be, for some reason, and might never recover its old strength. Here's what Summers told Wonkblog about his theory, and here's a response from Western Kentucky University's David Beckworth, who disagrees with him.
It would be great to see Bernanke engage this secular stagnation debate. Although Larry Summers never directly responded to me, he did reply to Marc Andreessen's tweetstorm where my critique was raised. The crux of my argument was that the long-decline in real interest rates given as evidence of secular stagnation ignores the sustained decline in risk premiums. Once this phenomenon is recognized, there is no long decline in real interest rates. 
 
Larry Summers replied to my critique with the following:
Markets – in the form of 30-year indexed bonds – are now predicting that real rates well below 2 percent will prevail for more than a generation... I think it is quite plausible and consistent with Marc’s picture that equilibrium real rates were roughly constant at around 2 percent until the mid-1990s and have trended downward since that time.
Looking to inflation indexed bonds or TIPs as a guide to the market's prediction of real rates is also misleading. It too fails to account for a liquidity premium priced into TIPs. On his second point, below is an updated version of the picture to which Larry Summers says he sees a downward trend since mid-1990s. Note that the real interest rate is now turning sharply up whereas before it was still flat. The real interest appears poised to return to its previous trend.


That the real interest rate appears to be returning to its previous trend--rather than finding a new lower one as suggested by Summers--makes sense if we plot this interest rate against the CBO's output gap. This is done in the figure below and reveals a striking fit. It also suggests the real culprit behind the sustained low rates is a prolonged business cycle. Now that the economy finally appears to be on a path to full recovery, the output gap is closing and the real interest rate is following it. So much for the smoking gun of secular stagnation.


I really hope Ben Bernanke joins this conversation. He had a great speech back in March 2013 that hinted at some of these topics. Welcome to the blogosphere Ben!

P.S. Although my original secular stagnation critique was made at the Washington Post, I did a more thorough follow-up piece with Ramesh Ponnuru at the National Review. I also did an interview with a Brazilian newspaper on secular stagnation.

Saturday, March 28, 2015

Do Air Conditioners Explain the Rise of the South?

Paul Krugman says yes and yes:
The rise of the US sunbelt can be understood largely as a response to the emergence of widespread air conditioning, which made places that are warm in the winter attractive despite humid, muggy summers. It’s a gradual, long-drawn-out response, because location decisions have a lot of inertia[.]
And this:
[T]here’s a real demographic turning point for the South circa 1960, as a steadily falling share of the total US population shifts to a sustained rise...this turning point coincides with the coming of widespread home air conditioning. So when you ask why Sunbelt states have in general grown faster than those in the Northeasy, don’t credit Art Laffer; credit Willis Carrier.
This is an intuitive explanation, but it is incomplete and cannot explain recent net migration patterns within the United States. On the first point it is important to note that the South's economy actually began to accelerate in the 1930s and 1940s.

 

This takeoff is one of the great stories from 20th century U.S. economic history. From the close of the Civil War up through World War II, the South's economy had been relatively undeveloped and isolated from the rest of the country. This eighty-year period of economic backwardness in the South stood in stark contrast to the economic gains elsewhere in the country that made the United States the leading industrial power of the world by the early 20th century. Something radically changed, though, in the 1930s and 1940s that broke the South free from its poverty trap. From this period on, the South began modernizing and by 1980 it had converged with the rest of the U.S. economy. But why the sudden break in the 1930-1940 period? 

There have been a number of attempts to explain this sudden change in the trajectory of the South. Gavin Wright (1997) attributes it to the opening up of labor markets, Connolly (2004) looks to improved human capital formation, Cobb (1982) points to industrial policy, Beasley et al. (2005) finger increased political competition, Bleakley (2007) sees hookworm eradication as important,  Glaeser and Tobio (2008) look to housing regulations, and Bateman et al (2009) see the large public capital investments in the South during the 1930s and 1940s as facilitating a 'Big Push' for the region. The takeaway is that there were probably many factors that supported the rise of the South, not just air conditioning.

The other problem with looking to air conditioning as the reason for the rise of the South is that there are recent net migration patterns that show movement into the warm South from other warm regions. not just from cooler regions. For example, this Census Bureau figure shows for the 1995-2000 period net migration out of California often went into other warm states, including many in the South. 


To further illustrate this point, he next three figures focus in on three big counties in Texas: Harris Country (Houston), Dallas County (Dallas), and Travis Country (Austin). They all show for the years 2008-2012 net inflows coming from southern California where it is also warm (source): 




So air conditioning cannot be an important story over this time. It may have contributed, along with the other factors listed above, to the the initial takeoff of the American South, but not so in recent decades. The rise of the South is a far more complicated story.

Friday, March 27, 2015

Are Wages Primed for Take Off?

Just a quick note on whether the Fed should raise interest rates later this year. One concern that many observers have with the Fed tightening is that nominal wage growth has yet to show any signs of accelerating. They often point to the employment cost index which only shows modest nominal income growth.

An alternative way to think about this issue is to look at the University of Michigan/Thompson Reuters Survey of Consumers where households are asked how much their dollar (i.e. nominal) family incomes are expected to change over the next 12 months. I have used this series in the past to talk about the stance of monetary  policy. This measure had averaged near 5 percent prior to the crisis, but then crashed and flat lined near 1.5 percent thereafter. The Fed, I argued, should have restored expected household dollar income to its 5 percent trend growth and its failure to do so was a dereliction of duty

How times have changed. Here is the same series updated to the present. Since late 2013 it has accelerated and is now almost back to 5 percent.


This suggests the economy is heating up and getting closer to full employment. If we plot this series against the employment cost index we get the following figure. Note the expected nominal income growth tends to lead the actual income growth. This suggest that wages are primed to start accelerating.


So  the wage growth concerns over the Fed tightening this year may soon become a moot point. This may one reason some folks like James Bullard are calling for a rate hike later in the year.

Thursday, March 19, 2015

Ramesh Ponnuru on the 'test' of Market Monetarism

Ramesh Ponnuru has an article in the National Review where he revisits the 'test' of Market Monetarism put forth by Paul Krugman and Mike Konczal in 2013. Here is Ramesh:

The story begins in late 2012. The Federal Reserve had begun its third round of monetary expansion following the economic crisis of 2008. Keynesian economists were sounding an alarm about the deficit-cutting measures — a combination of tax increases and spending cuts — that were scheduled to take effect at the start of 2013. Rapid deficit reduction, they warned, would harm the economy. A letter from 350 economists referred to “automatic ‘sequestration’ spending cuts everyone agrees should be stopped to prevent a double-dip recession.”
It is worth  noting that these concerns about a double-dip recession were translated into explicit forecasts about the number of jobs that would be lost. Estimates ranged from 660,000 to 1,800,000 jobs would be lost as can be seen be below:

Organization
Estimate of Sequester Impact on Employment
Source
Promoted by
Congressional Budget Office
750,000 to 1,600,000 jobs lost
Here and here
Macroeconomic Advisers
700,000 jobs lost
Bipartisian Policy Center
1,000,000 jobs lost
ThirdWay
1,800,000 jobs lost
Economic Policy Institute
660,000 jobs lost

   
Now back to Ramesh:
David Beckworth, a professor of economics now at Western Kentucky University, and I challenged this view. In an op-ed for The Atlantic’s website, we wrote that the Federal Reserve could offset any negative effect that deficit reduction might have on the economy.

[...]

In April, the liberal economics writer Mike Konczal resurrected an op-ed that Beckworth and I had written for The New Republic in 2011 making the same basic argument about the power of monetary policy, which is associated with a school of thought sometimes called “market monetarism.” He wrote: “We rarely get to see a major, nationwide economic experiment at work, but so far 2013 has been one of those experiments — specifically, an experiment to try and do exactly what Beckworth and Ponnuru proposed... Krugman concurred with Konczal, writing that “we are in effect getting a test of the market monetarist view right now” and “the results aren’t looking good for the monetarists.”

Read the rest to learn how the 'test' turned out.  Here is my own take on how the test turned out.