Bond Vigilantes and the Risk Premium

Some folks seem to be having a hard time with my previous post, bond vigilantes to the rescue.  They assume that there could be an actual default by the U.S. Treasury Department that would be reflected in a rising risk premium. While this is certainly possible, I find it highly unlikely since the U.S. government could always  print dollars to buy up its debt.  It could gradually "monetize the debt" and allow slightly higher inflation to slowly erode the burden of the national debt as it did after World War II.  This is, in my view, the most likely worst-case scenario, not an outright default.  The real risk for treasury holders then is a higher inflation risk premium, not a higher risk premium. 

But even this outcome seems unlikely in the near term.  The most likely development treasury holders face over the next year or so is a temporary bout of higher-than-expected inflation associated with Fed easing or more rapid economic growth.  This was the premise of my post, not an outright default.  Given my view that a robust recovery has not taken hold because the demand for safe assets remains elevated (i.e. portfolios remain overly weighted to low yielding, liquid assets), a temporary rise in inflation would cause the much needed treasury sell off that would start a recovery.  That is, treasury holders would sell their treasuries and other safe assets and move into riskier, higher yielding assets. We already see this in the relationship between expected inflation (using 10-year treasury breakeven) and stock prices.  The treasury sell off, therefore, would catalyze the rebalancing needed for a strong recovery. 

Note in this story, the risk premium would actually fall with the recovery.  Currently, it is too high as indicated in this post and as suggested by the figure below from Ed Bradford.  The figure shows the S&P500 earnings yield less the 20-year treasury yield.  This equity risk premium spread has been hovering around 5% over the past two years which seems unreasonably high. 

These indicators of inordinately high risk premiums correspond to the ongoing high demand for safe assets. Once the demand for safe assets is normalized--via the portfolio rebalancing--these risk premiums should decline too.

Bond Vigilantes To the Rescue

Paul Krugman claims that should the much-dreaded bond vigilantes show up, they actually would be good for the economy.  He notes that unlike Greece, the United States has its debt denominated in its own floating currency.  Consequently, the appearance of bond vigilantes would lead to an expansionary decline in the value of the dollar, not a contractionary rise in interest rates.  Tyler Cowen is not buying this story, but Nick Rowe sees some merit in it.  I do too, but from a slightly different perspective.  

Currently, investors around the world have their portfolios inordinately weighted toward safe, liquid assets.  This is because of the ongoing economic uncertainty caused by the Eurozone crisis, fiscal cliff, China slowdown, etc.  They also have a seemingly insatiable demand for these safe assets as evidenced by the ongoing decline in their yields across the globe (see below).  These developments, however, mean that investors are avoiding higher yielding, riskier assets more so than normal. Consequently, these unbalanced portfolios are suppressing asset prices, keeping household balance sheets weak, and ultimately are holding back robust aggregate nominal spending.  Another way of saying this is that risk premiums are currently too high relative to fundamentals.  

 Source

The appearance of bond vigilantes would indicate their economic outlook has changed and are in the process are rebalancing their portfolios.  This rebalancing, whether it was driven by higher expected inflation or higher expected growth, would catalyze more aggregate nominal expenditures and given the significant economic slack, more real economic growth.  The problem, as noted by Nick Rowe, is that we want some portfolio rebalancing, but not too much  That is why an nominal GDP level target is important.  It would clearly set expectations on how much nominal income growth and, by implication, how much portfolio rebalancing would be allowed.  In other words, a nominal GDP target would guarantee we get the just the right dose of bond vigilantism needed to shore up the recovery.  And note that the recovery in nominal GDP would push up interest rates too. Using Paul Krugman's terms, this would be an expansionary rise in interest rates. So let's not fear bond vigilantes, but learn to manage their expectations in a way that will spark a real economic recovery.

Update: Just to be clear, the expansionary rise in interest rates does not mean the Fed would raise rates before the recovery.  Rather, recovery would naturally cause yields to rise (i.e. demand for credit increases, desired savings falls) and the Fed would respond by raising its target federal funds rate.  For more on this point see here.

Update II: Further thoughts on bond vigilantes here.

John Taylor Tips His Hat To NGDP Targeting

John Taylor gave a recent talk at the Milton Friedman Centennial Celebration where he made some interesting points.  He makes the case that that monetary policy has been unpredictable and ad-hoc over the past few years, including some periods where monetary policy was too tight.  As a result, he believes a more systematic, rule-based approach to monetary policy would be beneficial to the recovery.  To this end, he tips his hat toward nominal GDP targeting in his discussion of what he thinks Milton Friedman would endorse:

Another question raised during the discussion at the centennial conference is “what about nominal GDP targeting?” In my view, Milton Friedman would have been positive about a proposal to keep nominal GDP growth stable, but would have wanted also to have a specific rule for the instruments of policy to achieve that target.

I agree. Friedman in this 2003 WSJ article indicates he might have liked a nominal GDP level target. As Michael Woodford notes, nominal GDP targeting can be thought as the heir to the Friedman monetary target rule once one acknowledges that velocity is not stable. The predictability, transparency, and certainty this rule would create would also make it appealing to Milton Friedman.  Here is how I explained this point before:

So yes, Milton Friedman did call for buying longer-term securities until a robust recovery takes hold... I suspect, however, that Friedman would have preferred that such a monetary stimulus program be done in a more systematic manner than that of announcing successive, politically costly rounds of QE.  Imagine how much easier all of this would have been had the Fed announced a level target from the start and said asset purchases will continue until the level target was hit.  There would have been no need to announce the large dollar size of the asset purchases up front that attracts so much criticism.  There would also have been no need to announce successive rounds of QE that make it appear the previous rounds did not work.  More importantly, it would have more firmly shaped nominal expectations in a manner conducive to economic recovery.  The question is what type of level target would Friedman have supported? 

I think the answer is clear.

Casey Mulligan Nails It

At least on this part:

To put it another way: for every worker that construction lost between 2007 and 2010, the rest of the economy lost at least another five workers, rather than gaining workers. I agree with Professor Krugman and other opponents of the “sectoral shifts theory” that something must have happened — in less than a year or two — that profoundly affected practically all industries and practically every region.

Mulligan is right that something suddenly happened that affected economic activity in every region and every industry.  I date this development to about mid-2008 as can be seen in the following figures.  The first one shows that despite the start of the housing recession in April, 2006 employment in the rest of the economy continued to grow through early 2008:  

Similarly, despite the fall in dollar incomes tied to housing, other dollar incomes continued to grow until about mid-2008:

So, as Mulligan notes, something turned a two-year sectoral recession into an economy-wide one.  Many folks attribute it to the worsening of the financial crisis.  I think a better story is that the Fed passively tightened monetary policy around mid-2008.  A passive tightening occurs whenever the Fed allows total current dollar spending to fall, either through a endogenous fall in the money supply or through an unchecked decrease in velocity.  Given a proper measure of the money stock--one that includes both retail and institutional money assets--this can be shown to be the case for the U.S. economy during this time.  This reduction in broad money assets and the drop in velocity amounts to an excess money demand (i.e. safe asset shortage) problem. 

The Fed's failure to stabilize total dollar spending had implications for household balance sheets.  Households had come to expect about 5% annual nominal income growth over the past 25 years or so.  These expectations were assumed by household and firms when they signed long-term nominal fixed debt contracts.  A debt crisis was therefore inevitable when these long-term nominal income forecasts were not realized.

Now Casey Mulligan thinks it was a change labor market incentives brought on by government policies that caused the economy-wide collapse.  I agree with Mulligan's premise that these policies do change incentives, but am not convinced that magnitudes are large enough to explain the severity of the past four years.  A far easier story to tell is there has been an excess demand for money assets (i.e. a shortage of safe assets).  For households this is a particularly compelling story:

Here's hoping that Mulligan, who has done extensive work on money demand, can weave this story into his narrative of the crisis.

What the Obama Win Means for NGDP Targeting

Okay, so Greg Mankiw will not be Fed chairman after all and usher in a golden era of nominal GDP targeting.  All hope is not lost, though, on the nominal GDP targeting front.  Here is the Cynthia Lin of the WSJ:

President Barack Obama‘s re-election has made markets more confident that the Federal Reserve will continue on its current path.

With Obama winning a second term, the odds are higher that any leadership change at the central bank will follow in Chairman Ben Bernanke‘s footsteps. The path of Fed policy now seems clearer to the market, as evidenced in the fed-funds futures market.

November 2014 fed-funds futures now price in no chance of an interest-rate increase by then, compared with a 24% chance priced in at Tuesday settlement. The central bank has said it plans to keep rates near zero at least until mid-2015. But Bernanke’s current term ends in 2014, and some had speculated that a Romney administration would have appointed a Fed chairman who would push for higher rates sooner

So the Fed has less to worry about it as it executes QE3.  To the extent QE3 is a step in the direction of the Fed adopting an explicit nominal GDP level target, this would be a positive development. President Obama also has an opening to fill at the Board of Governors and he could reinforce the Fed's move toward a nominal GDP target by appointing someone who endorses it.  How about his best friend at Goldman Sachs, Chief Economist Jan Hatzius? He would be a nice complement to the Baord of Governors.

If Mitt Romney Becomes the Next President...

I hope he appoints Greg Mankiw as the next Fed chairman.  Here is a post from October, 2011 that explains why Mankiw would be a great choice:

Back in May, 2011 I wrote the following on my blog:

Greg Mankiw recently referred to a paper where he assess which inflation rate should be targeted by the central bank.  Here is his conclusion:

[A]central bank that wants to achieve maximum stability of economic activity should use a price index that gives substantial weight to the level of nominal wages. 

There are several good reasons laid out in the paper for targeting nominal wages.  Here I like to point out that stabilizing nominal wages is similar to stabilizing nominal income per capita.  It is not too much of a stretch to go from this to a nominal income or nominal GDP target.  In fact, Greg Mankiw and Robert Hall have a 1994 paper that sings the praises of a nominal GDP target, especially one that that targets the consensus forecast of the nominal GDP level. 

So where does Greg Mankiw stand today on nominal GDP level targeting?  If he still supports it, does he see the need to return nominal GDP back to its pre-crisis trend or at least higher than its current level?

Though I have never got a direct answer from Greg Mankiw, there is now enough circumstantial evidence to know his answers to my questions. First, he and coauthor Matthew Weinzierl have a recent Brookings Paper on the optimal stabilization policy.  They go through a menu of policy options, but reach this conclusion if monetary policy is not constrained:

The second level of the hierarchy applies when the short-term interest rate hits against the zero lower bound. In this case, unconventional monetary policy becomes the next policy instrument to be used to restore full employment. A reduction in long-term interest rates may be sufficient when a cut in the short-term interest rate is not. And an increase in the long-term nominal anchor is, in this model, always sufficient to put the economy back on track. This policy might be interpreted, for example, as the central bank targeting a higher level of nominal GDP growth.

In other words, monetary policy targeting a nominal GDP level is sufficient to bring the economy back to full employment.  That sounds like a rather favorable view of nominal GDP level targeting to me.   If that were not enough, Greg Mankiw today implicitly endorses nominal GDP level targeting by linking on his blog to the Goldman Sachs paper on nominal GDP level targeting.  I'd say Mankiw has answered my questions clearly.

Now Mankiw is not just a big-time academic economist at Harvard.  He is also an economic adviser to  Mitt Romney, the likely GOP candidate for president.  That means NGDP level targeting might eventually find its way into the White House.  There are good reasons for Republicans to endorse such an approach to monetary policy.  I hope Mitt Romney is hearing them.

We will have to wait another day to see if Greg Mankiw will get a chance to implement nominal GDP level targeting.  

Update: Here is Joe Weisenthal earlier this year talking about Greg Mankiw as Fed chair.

Steve Hanke: a Market Monetarist?

This past weekend while traveling, I was able to listen to Steve Hanke on EconTalk with Russ Roberts.  Hanke has been out in front of the hyperinflation story in Iran and has been doing some really interesting work on it.  While this is an important discussion, another topic that was brought up in the interview was Hanke's views on the stance of U.S. monetary policy.  Hanke argued that monetary policy in the United States has been effectively tight over the past few years.  This is evident, he says, by looking to broad monetary aggregates like the M4 divisia, a view that I share.  Elsewhere, he has argued the Fed should target final nominal sales.  With these views, Hanke could almost be confused for being a Market Monetarist.  He even calls himself a monetarist in the interview.  

There are a few areas, though, in the interview where he appears to differ from Market Monetarists:

(1) He believes that the Fed has been extremely loose with its large expansion of the monetary base that began in late 2008.  Market Monetarist disagree and say that the large run-up in the supply of Fed liabilities has been more than offset by a large run-up in the demand for them.  Monetary policy is not loose if the Fed fails to check an excess demand for the monetary base.  Moreover, monetary policy is not loose if the run-up in the monetary base is not expected to be permanent. If, on the other hand, some part of the increase were expected to be permanent then nominal spending and nominal income should go up today in anticipation of this development. The fact that this has not happened indicates Fed policy has been too tight.

(2) He thinks the Fed has bought up an inordinate share of U.S. treasuries and has therefore kept interest rates too low. Market Monetarist note that the Fed actually holds only about 15% of total marketable treasury securities and therefore is not the main reason for the low treasury yields.  In other words, despite the large run up in U.S. public debt, households, their financial intermediaries, and foreigners are more than willing to hold treasuries.  Blame them and the weak economy causing them to buy more treasuries for the low interest rates, not the Fed (at least not directly).

(3) He thinks the weak recovery in M4 is the result of onerous bank capital requirements that prevents financial firms from producing many safe assets.  There may be some truth here, but this view overlooks the ability of the Fed to catalyze the private creation of more safe assets.  Here is how I explained this process before:

[T]he Fed and ECB should create an environment conducive to monetary asset creation that would support the return of robust aggregate nominal spending.  Since most of the money assets are created by the credit, maturity, and liquidity transformation services of financial firms, policymakers should aim to create an environment conducive to increased financial intermediation.  The easiest way for monetary policy  to do this is to raise the expected growth path of aggregate nominal expenditures. This would raise expected nominal income growth and the demand for money assets.  This, in turn, would catalyze financial intermediation and  lead to the creation of more money assets.  And of course, the way to raise the expected growth path of aggregate nominal expenditures is to adopt a nominal GDP level target.  It is time for monetary regime change!

I suspect the absent of this action by the  Fed is a much bigger factor behind the weak M4 growth than the regulatory burden.  

These differences, though, probably overstate the gap between Steve Hanke and Market Monetarists.  And maybe his views on these issues are not as different as they seem. In any event, it is good to know that someone with his views is still being heard at the CATO institute.