Wednesday, September 30, 2015

Doubling Down on Abenomics

So it appears the Bank of Japan (BoJ) had already doubled down on Abenomics before the prime minister announced a new NGDP level target. In late 2014, the BoJ said it would increase the growth of the monetary base and by implication the number of assets it would purchase. 

Here is the original BoJ announcement on Abenomics in early 2013 (my bold):
The Bank will achieve the price stability target of 2 percent in terms of the year-on-year rate of change in the consumer price index (CPI) at the earliest possible time, with a time horizon of about two years.In order to do so, it will enter a new phase of monetary easing both in terms of quantity and quality. It will double the monetary base and the amounts outstanding of Japanese government bonds (JGBs) as well as exchange-traded funds (ETFs) in two years, and more than double the average remaining maturity of JGB purchases...The Bank of Japan will conduct money market operations so that the monetary base will increase at an annual pace of about 60-70 trillion yen.1
And here is footnote one:
Under this guideline, the monetary base -- whose amount outstanding was 138 trillion yen at end-2012 -- is expected to reach 200 trillion yen at end-2013 and 270 trillion yen at end-2014.
The BoJ got close. The monetary base hit 267.4 trillion yen at the end of 2014. The BoJ did not, however, hit its inflation target. So it announced in October, 2014 it would increase how fast the monetary base would grow:
[T]he Bank of Japan decided upon the following measures.
(1) Accelerating the pace of increase in the monetary base by a 5-4 majority vote. The Bank will conduct money market operations so that the monetary base will increase at an annual pace of about 80 trillion yen...
So the BoJ decided to double down its bets on Abenomics by growing the monetary base an additional ¥10 trillion a year for a total of  ¥80 trillion per annum. This non-trivial pick up in growth can be seen in the figure below under the 'Abe II' label:

So maybe Japan is more determined than we realized to make this reflation experiment work. Maybe there is more credibility to Shinzo Abe's new NGDP level target than we first imagined. Abenomics has, after all, consistently raised aggregate demand as noted in my last post. The key is not to repeat the mistake made with Japan's first QE program by reversing the monetary base growth as is seen in the figure above. 

P.S. Both the core inflation and rate deflator inflation rate show sustained rises. This is not good for Neo-Fisherism.

Monday, September 28, 2015

How Practical is Japan's New NGDP Target?

So Japan's Prime Minister Shinzo Abe has decided to adopt a NGDP level target:
Formally re-elected head of the ruling party, Prime Minister Shinzo Abe said Thursday he has set out three new goals for “Abenomics” and will target a 20 percent increase in gross domestic product to ¥600 trillion.
The details are murky, but what is clear is that for this plan is to work it will require a credible commitment to permanently expand the monetary base, as shown by Michael Woodford, Paul Krguman, William Buiter, and many others. But was not this already supposed to be happening? Was not Abenomics to permanently double the monetary base? Recall this figure of Japan's monetary base:

Some observers like Hausman and Weisman worry that Abenomic's monetary base expansion is not credible and therefore not permanent. If so, then this latest Abe goal of raising the level of NGDP by 20% may not be credible either. 

Or maybe not. One way to check whether the monetary base expansion is expected to be at least somewhat permanent is to see whether nominal spending has been rising. If Abenomics is just a larger version of Japan's 2001-2006 QE program where the monetary base expansion was temporary and did little to raise aggregate demand, then we should expect to see similar patterns in nominal expenditures under Abenomics. So what has actually happened?

The figure above shows that Abenomics compared to the 2001-2006 QE program has been fairly successful in generating aggregate demand growth. This suggests that the monetary base growth is seen at least as somewhat permanent.

Raising the  level of NGDP from ¥499 to ¥600 trillion is a far bigger task. Below are three different growth paths for NGDP to reach its new target: a three-year path, a five-year path, and a seven year path. Although no official time table has been set for hitting the ¥600 trillion target, many observers are mentioning 2020 as the target date. As Simon James Cox notes, this time frame  seems to be corroborated by Japan's Cabinet Office. This would correspond to the five-year path. If NGDP were to follow the trend growth of NGDP during Abenomics it would follow the seven-year path.

It is worth repeating, as I often do, that the more credible this policy becomes the less need there is for additional growth in the monetary base. If the public perceives the government is firmly committed to permanently expanding the monetary base then its velocity will grow as the public tries to rebalance their portfolios away from it. This increase in monetary base velocity, then, will do much of the heavy lifting in raising aggregate demand growth. 

The question is whether a 20% increase in NGDP is a credible goal. It is one thing for Abenomics to gain some credibility, but a 20% NGDP increase in five years? I want to believe, but I am a bit skeptical.

Most of the NGDP growth completed under Abenomics has resulted in higher inflation rather than in higher real GDP growth. The figures below show the inflation rates for the GDP deflator and the core consumer prices in Japan. These increases in inflation have not been matched by sustained increases in real growth. Real GDP has average 0.58% year-on-year growth every quarter since 2013:Q1 whereas the deflator has average 0.93% growth.

What if the public expected these patterns to continue going forward? That is, what if most of the NGDP growth was expected to result in more inflation? In that case, I suspect the aging population in Japan living off fixed incomes would strongly object to the ¥600 trillion NGDP target. This IMF study lends support to this view. It shows It shows that economies with large numbers of old people tend to experience lower inflation rates. This suggests they  have political influence.

One could argue that this higher level of inflation would reduce real debt burdens, ease excess money demand, lower real rates, and increase real economic activity in a way that has not been seen yet. This may be true, but that these developments have not happened speaks to why a 20% NGDP increase may not be credible. Abenomics may have been tolerated, but would an aging population and others holding government debt tolerate five more years of higher inflation? I am not so sure. 

Another issue that complicates matters is that raising NGDP growth alone will not solve all Japan's problems. It has numerous structural problems that need to be addressed. These were supposed to be tackled in third arrow of Abenomics. Not much has happened here and maybe more robust NGDP growth would make it easier to address the structural problems. 

I do not want to be too pessimistic here. Abenomics successfully raised NGDP growth compared to the 2001-2006 QE program. That is a real accomplishment and speaks to possibility of doing more. I just worry that a 20% NGDP target in five years might be asking too much given the amount of credibility it requires.

Update: One can view this NGDP credibility problem through the lenses of the fiscal theory of the price level. Below is an excerpt from an earlier post:
Paul Krugman [notes] that there could be too much fiscal credibility. If so, it would be creating a drag on aggregate demand growth (i.e. higher expected future surpluses imply lower velocity today and, in turn, mean lower aggregate demand growth). While this argument may apply to the United States, Krugman is certain it applies to Japan. Here is Krugman discussing the proposed tax hikes in Japan:
I see no prospect that Japan will put off the tax hike forever. But even if it were true, this is credibility Japan wants to lose.

After all, suppose investors conclude that Japan will never raise taxes enough to service its debt. What would they think would happen instead? Not default — Japan doesn’t have to default, because its debts are in its own currency. No, what they might fear is monetization: Japan will print lots of yen to cover deficits. And this will lead to inflation. So a loss of fiscal credibility would lead to expectations of future inflation, which is a problem for Abe’s efforts to, um, get people to expect inflation rather than deflation, because … what?

Long ago I argued that what Japan needed was a credible promise to be irresponsible. And deficits that must be monetized are one way to make that happen...
Interestingly, John Cochrane the made the same point in his 2011 paper:
The last time these issues came up was Japanese monetary and fiscal policy in the 1990s... Quantitative easing and huge fiscal deficits were all tried, and did not lead to inflation or much‘‘stimulus’’. Why not? The answer must be that people were simply not convinced that the government would fail to pay off its debts. Critics of the Japanese government essentially point out their statements sounded  pretty lukewarm about commitment to the inflationary project, perhaps wisely. In the end their ‘‘quantitative easing’’ was easily and quickly reversed, showing those expectations at least to have been reasonable.
What Krugman and Cochrane are saying is that there may be too much fiscal credibility in Japan to allow the public to believe current and future monetary base expansions will be permanent. The interesting question, then, is whether the announcement of this NGDP target changes that credibility.

Update II: See John Cochrane's post on Japan. He briefly responds to this post in an update.

Tuesday, September 22, 2015

Ad Hoc Monetary Policy Redux

The Cato Institute for Monetary and Financial Alternatives asked me to write a brief response to the last FOMC decision. It is now up at the Alt-M blog. Here is an excerpt:
Just a few months ago the FOMC was signaling it would almost certainly raise interest rates, but now it has changed its mind.  This change would not be so bad if it were predictable, but it was not so.  No one expects the Fed to perfectly forecast the economy, but we should expect the Fed to make clear how it would respond to differing states of the economy.  This simply has not happened.  From the QE programs to forward guidance to lifting interest rates from zero, Fed policy has been made up on the fly.  This unpredictable behavior has meant that no one, including Fed officials, knows for sure what will happen from one FOMC meeting to the next.

As a result, markets have become more and more obsessed with every word coming from the mouths of Fed officials.  Post-FOMC press conferences like the one last Thursday became must-watch TV for anyone concerned about investments.  Ironically, then, the Fed’s attempt to calm markets through these ad-hoc measures has only made them more fragile.
I am as guilty as anyone about getting worked about FOMC meetings. Nothing like the anticipation  about what the Fed will do in the days leading up to the meeting. I am especially bad about looking forward to being on twitter during a post FOMC Janet Yellen press conference. It is sad that we have come to this state of affairs. Below is a repost of an earlier piece from 2014 that elaborates on this theme.

Friday, September 4, 2015

Revealed Preferences: Fed Inflation Target Edition

Over the past  six years the Fed's preferred measure of the price level, the core PCE deflator, has averaged 1.5 percent growth.  That is well below the Fed's explicit target of 2 percent inflation. Why this consistent shortfall?

Some Fed officials are asking themselves this very question. A recent Wall Street Journal article reporting from the Jackson Hole Fed meetings led with this opening sentence: "central bankers aren't sure they understand how inflation works anymore". The article goes on to highlight some deep soul searching being done by central bankers in the Wyoming mountains. It is good to see our monetary authorities engaged in deep introspection, but let me give them a suggestion. Dust off your revealed preference theory textbooks and see what they can tell you about the low inflation of the past six years. 

To that end, and as a public service to you our beleaguered Fed officials, let me provide some material to consider. First consider your inflation forecasts that go into making the central tendency consensus forecasts at the FOMC meetings. The figures below show the evolution of these forecasts for the current year, one-year ahead, and two-years ahead. There is an interesting pattern that emerges from these figures as you expand the forecast horizon: 2 percent becomes a upper bound.

So the first insight from revealed preferences is that you and your fellow FOMC officials have been consistently looking at an upper bound of 2% on core PCE inflation. Now if we add to this observation the fact that the FOMC has meaningful influence on inflation several years out, then these revealed preference are saying you want and expect to get an inflation upper bound of 2%

Your chair, Janet Yellen, conceded this point in the press conference following the December 2012 FOMC meeting:

But it’s important to point out that the Committee is not anticipating an overshoot of its 2 percent inflation objective (p.13).
Now an upper bound means there can be activity below it. And we see just that that in your inflation projections. Collectively, then, all of this revealed preference evidence suggests that you and your Fed colleagues do not have a 2% inflation target, but rather an inflation corridor target. Based on your above forecasts your corridor target appears to be somewhere between 1% and 2%. 

We can get a better sense of where this inflation corridor target lies with additional revealed preference evidence. This evidence is found in the following figure which shows core PCE inflation and the timing of your QE programs. The figure suggest that you and your fellow FOMC members tend to start QE programs after core inflation had been drifting away from the 2% upper bound and you do so in a manner that prevents it from drifting below 1% for very long. (Based on this reading, you all are likely to change tack soon if core inflation does not stop drifting toward 1%. )

This reading is corroborated by looking at the changes in the Fed's share of treasury securities as a percent of all market treasuries. Ever since the zero lower bound kicked in late 2008, the FOMC has tended to allow its share of treasury holdings to adjust in manner that offsets changes in core PCE inflation. That is, when inflation was falling the Fed started increasing its share of treasury securities and vice versa.

So rest easy dear Fed official. No need for any existential angst. According to revealed preferences, you are still driving core inflation--which ignores supply shocks like changes in oil prices--it is just that you have a roughly 1%-2% core inflation target corridor rather than a 2% target. So even though you may not realize it, you are doing a bang up job keeping core inflation in your target corridor.

Thursday, August 27, 2015

How to Create a Chinese Economic Crisis in Three Easy Steps

First, peg the yuan to another currency that has been rapidly appreciating over the past year...

...this will choke the tradeable sector...

...and really put a damper on nominal spending growth.

Second, respond to the above economic weakening by easing domestic monetary conditions. Over the past year, the People's Bank of China has done so with five cuts to its prime lending rate and several large reductions in the required reserve ratio for banks

Third, allow nervous investors to pull their capital out of China. Investors have come to expect a large yuan devaluation given the above developments. For the easing of domestic monetary conditions has put downward pressure on the yuan while the dollar peg has pushed it up inordinately high, creating an imbalance. The only way China can maintain this imbalance is to defend its dollar peg by burning through its foreign exchange reserves...

...a response that becomes more expensive the the tighter Fed policy becomes...

...but there is a limit to burning through these assets since China still needs some foreign reserves buffer. Ambrose-Evans Pritchard reports some observers are already wondering if China is getting close to its buffer limit. If so, China will be forced to devalue. This is what investors are now expecting and therefore are eager to get out. This puts further pressure on China's stock of foreign reserves.

Note that devaluing the yuan will be costly too. Many Chinese firms, previously expecting the dollar peg to hold, have taken out lots of dollar denominated debt. So either China burns through its reserves or it increases the private sector's real debt burdens. There are no easy choices here.

China, in short, has backed itself into a corner because it has attempted to do all three goals of the impossible trinity...

...something the emerging market crisis of 1997-1998 taught us does not end well. But China is trying anyways and the recent stock market roller coaster ride is just one of many adverse outcomes likely to come out of these efforts. Fortunately, most emerging countries are not making this mistake as noted by Greg Ip. That is good news. The bad news is that it will not be easy for China to undo the three steps it has taken toward doing the impossible trinity.

P.S. For a broader perspective that makes the same point see this post by Lars Christensen.

Friday, August 14, 2015

A Work Around for the Fed's Knowledge Problem

Ramesh Ponnuru has a new article titled Fallible Fed Needs a Few Good Rules. Here is an excerpt:
Some Republicans think that Congress should supply a framework for the Fed, making it more rule-bound. One much-discussed approach would have the Fed follow the “Taylor rule” -- named after John Taylor, a Stanford economist -- which says that the federal funds rate should be set according to a formula involving inflation, the long-run real interest rate, and the gap between the economy's actual output and its potential output. The Fed’s behavior might then be more predictable. It would raise and lower the fed funds rate according to the formula.

The Economist magazine recently criticized the idea of following a formula, arguing that the Taylor rule has drawbacks. It argued that the Fed should continue to exercise broad discretion over monetary policy.

But to say that a central bank acting on its own discretion could perform better than one following the Taylor rule doesn't mean that it is likely to do so. And the Taylor rule isn't the only possible rule. The Fed could instead act to keep the growth in nominal spending -- the total amount of dollars being spent each year on consumption and investment -- at a fixed percentage each year. That approach wouldn't require it to make confident estimates about the output gap or [equilibrium] interest rates. It would also lend itself to greater predictability, and a recent paper suggests that it might work better than either inflation-targeting or the Taylor rule, especially given uncertainty about those values.
Ponnuru alludes to an important point in the paragraph above. Monetary policy suffers from the knowledge problem. Monetary authorities simply do not know enough about the economy in real time to make informed decisions. And this is not just a problem with theoretical Taylor Rules. It is a problem for modern central banking. One area, in particular, where they have a hard time is knowing how to respond to supply shocks. They are challenging since they push output and inflation in opposite directions and have plagued central banks in advanced economies over the past decade. As I noted earlier:
Supply shocks were an issue in 2002-2004 when the much-ballyhooed productivity boom (a positive supply shock) of that time made Fed officials worry about deflation. They consequently kept interest rates low even though the housing boom was taking off. Supply shocks were also an issue in the fall of 2008 when Fed officials were concerned about rising commodity prices (a negative supply shock) and, as a result, decided to do nothing at their September FOMC meeting despite the collapsing economy. 
Across the Atltantic, the ECB has struggled even more with supply shocks. The ECB raised interest rates multiple times in 2008 and 2011 in response to the commodity price shocks (negative supply shocks). Below is a figure from a Robert Hetzel paper on this crisis that shows how misguided the rate hikes were. They occurred even though spending was already falling. It is no wonder the Eurozone has struggled so much since 2008.

One can trace this wrinkle back further. Ben Bernanke, Mark Gertler, and Mark Watson argue the reason oil supply shocks have historically been tied to subsequent weak growth is not because of the shocks themselves, but because of how monetary policy responded to those shocks. That is, central banks typically responded to the inflation created by the supply shocks in a destabilizing manner. With the advent of inflation targeting in the early 1990s, this wrinkle has become institutionalized across most central banks.
Now in theory modern inflation targeting should be able to handle these shocks. For the modern practice is to do 'flexible inflation targeting' which aims for price stability over the medium term and therefore allows some wiggle room in responding to supply shocks. The problem, as demonstrated above, is that in practice it rarely works. Responding to supply shocks in real time requires exceptional judgement and usually some luck. In fact, as I note in this policy paper, some scholars think that the successes of inflation targeting prior to the crisis were due largely to luck. There were simply fewer supply shocks during the early years of inflation targeting.
Ironically, The Economist article Ponnuru mentions above argues discretion is preferred because of the knowledge problem! It says that "until the day the economy is fully understood, human judgment has a crucial role to play." Really? It would seem the last eight years of ad-hoc, make-it-up-as-we-go-along policy by the Fed and ECB would give pause to this type of thinking.  Especially with the weak recoveries these responses have created.

As Ponnuru notes, one of the big appeals of a nominal spending target rule is that it gets around the knowledge problem. Under this rule the Fed would simply stabilizes the path of total dollar spending.  There would be no need to worry about changes in the real economy. If for example, there were a positive supply shock—say new technology or increased oil supply—that lowered prices the Fed would do nothing other than maintain stable money spending. The composition of the spending would change—more goods and services at lower prices—but the total amount of spending would not. Likewise, total dollar spending would not change if there were a negative supply shock—such as a natural disaster or an oil shortage—though its composition would alter too.  In other words, the Fed would let relative prices and markets sort out real shocks on their own while maintaining monetary stability. This means the knowledge problem would no longer be a constraint in implementing monetary policy. This is just another reason for the Fed to adopt a nominal GDP target rule.

Austrians vs Market Monetarists on Canadian Fiscal Austerity

One of the more popular tales of 'expansionary austerity' was Canada in the mid-to-late 1990s. During this time Canada slashed government spending and brought down its debt-to-GDP ratio. Despite this fiscal stringency, the economy grew steadily and the unemployment rate fell. How was this possible?

Ramesh Ponnuru and I have argued this outcome occurred because the Bank of Canada (BoC) provided a monetary policy offset to the fiscal policy tightening. The evidence we point to in support of this view was the BoC cutting its target interest rate more than 500 basis points between 1995 and 1997. Some folks like David Henderson and Bob Murphy did not find this evidence compelling. It did not help our case that we shared the same view as Paul Krugman. Apparently, we were taking the "Keynesian view". Guilt by intellectual association?

Nonetheless, I laid out further evidence for the monetary offset view in a later post.  Bob Murphy has now replied to me in a new post and concedes that at least one of my points, the permanent increase in the monetary base, does lend some support to our view. (However, he correctly points out there are timing issues with the increase in the monetary base.) So he does concede the story is more complicated than he originally envisioned. 

In my view, however, the strongest evidence for the monetary policy offset view is not to be found in the monetary base. It is found is the divergence between the BoC's and the Fed's target interest rates. I made this point before and reiterated it in the comment sections of his new post. Below is an edited and extended version of the comment I left there. 
Bob, the monetary base data does raise some questions for both sides as you note. But let me share what I think is the most compelling evidence for the monetary offset view by making a set of observations.

First, Canada is a small open economy. During the period in question, Canada’s GDP was about 7% of US GDP on average. Canada, in other words, is very susceptible to external economic shocks, particularly ones coming from the US.

Second, Canada’s financial markets are highly integrated with US markets. This is especially true with short-term money markets which means there should be over the long run very little arbitrage opportunities between the two countries interest rates once one controls for risk and other country specific-factors.

Third, given the points above it should be the case that when the Fed (large economy) sets its short-term interest rate target in the US it also influencing short-term interest rates in Canada (small economy). This understanding would imply that the BoC generally follows what is happening to US monetary policy. That seems to be the case as seen here:
 But now take a closer look at the period in question:
The figure shows that Canada’s central bank interest rate differed as much as 250 basis points for a sustained period from the federal funds rate during the period in question.
How can this be possible given the observations above? Are not Canadian financial conditions tied closely to US financial conditions? The answer is yes, but they can deviate if the BoC exogenously intervenes and tinkers with interest rates. Put differently, had the BoC not intervened Canadian short-term interest rates probably would have more closely tracked the Federal Reserve’s target rate. As shown above, they typically do. But this time the BoC defied external money market pressures coming from the United States and struck its own interest rate path.
As a robustness check on this understanding I estimated the relationship between the BoC's target interest rate against the federal funds rate as well against the core inflation rate and the output gap in Canada. I estimated the model from 1980:Q1 to 1994:Q4, the period right before the BoC eases. This way we can ask the following question: given how the BoC adjusted interest rates in the past, how would one have expected it to do so going forward into the 1995-2000 period as new realizations of the federal funds, core inflation, and the output gap occurred?
If there were a deviation between the actual path and predicted path of the BoC's interest rate target during the 1995-2000 period, then this would constitute an exogenous movement or 'shock' to monetary policy. The figure below shows the results of this exercise.
The actual easing, then, was not something one could have easily predicted based on past BoC behavior. What is nice about this is that it provides a kind of a natural experiment for the monetary offset view. It provides (1) an exogenous easing of monetary policy (2) in a period of fiscal tightening and (3) results in stable aggregate demand growth. In my view the evidence provided by this natural experiment is very clear.
Nick Rowe rightly notes in the comments, though, that one should also look at the exchange rate. It depreciated during this time indicating monetary easing was at work. So it is hard for me to understand how one could view this experience as anything but strong evidence for the monetary offset view.