Given how much energy I’ve put into this blog, I’m hoping my readers can help me out a bit.  I have been getting suggestions that I really need to finish revising my manuscript.  So I decided that to better focus on it, I’d devote my blog to the Great Depression for the next two months.  Hopefully this will allow for synergy, as I will work on the manuscript and also post bits of it on the blog to get feedback.  I will also set a deadline of April 15 to finish revisions, in the hope that this will overcome the commitment problem (since I have little self-control.)  The fear of a public embarrassment for not finishing by this arbitrary deadline should force me to get it done. 

Of course you readers will have to sacrifice a bit, as the Great Depression may be much less interesting than our Great Recession.  But after mid-April I plan to go back to current events.  I will also have to sacrifice a bit, because I much prefer blogging, and my inflated ego may not be able to go long without excessively flattering commentaries like this and this.

Today I will sketch out a very brief summary of my explanation of the Great Depression, and the models I use to evaluate it.  This will summarize the first three chapters of the book.  I’d like to put the entire manuscript on my blog, but I don’t think the publisher would allow that.  But here is what I think I can do, I can probably post the three chapters that cover the early 1930s, as they are merely slightly modified versions of papers already in the public realm, published in the early 1990s.  So I will do that over the next few weeks.

So the plan is that you can give me feedback on my basic approach, and then later when I post chapters 4, 5, and 6 of the manuscript, a few comments on those as well.  You will notice that the writing style of those chapters (much of which were written in the early 1990s), is far inferior to this blog.  I still think the content is pretty good, but unfortunately my later chapters are more convincing in terms of economic analysis.  But what can I do, we are stuck with the fact that the earliest parts of the Great Depression are the most difficult to explain.  And I can’t very well do my narrative in reverse order, like a Hollywood movie with flashbacks.  (Or can I . . . )

Part 1.  What caused the Great Depression?

I believe the Great Depression had two primary causes.  One cause was deflationary monetary policies during 1929-33, and 1937-38.  The other cause was five nominal wage shocks during 1933, 1934, 1936-37, 1938 and 1939.  These 5 shocks were caused by New Deal programs, and slowed the recovery. This is why the Great Depression lasted 12 years, ending about the time of Pearl Harbor.  I don’t know what ended the Great Depression, and don’t really discuss it in my manuscript, but I suspect it had something to do with the German invasion of France, which led to a military mobilization in the US and elsewhere.  This may have directly raised AD, and indirectly increased expected NGDP growth by raising the expected inflation rate.  Wars are usually inflationary.  

I also think there were many secondary factors that depressed output.  During 1929-33 there were at least three important adverse supply shocks; Hoover’s high wage policy, Smoot-Hawley, and the big jump in MTRs during 1932.  But without the deflationary monetary policy, the tax hike wouldn’t have happened, the wage policy would have had little effect, and even Smoot-Hawley would have had only a modest effect.  And of course after 1933 there were many other New Deal polices that inhibited recovery, but I focus on the 5 wage shocks, as I am convinced that they are far more important than the other policies.

You may have noticed that the two variables I focus on are prices and nominal wages.  If you look at a real wage series (manufacturing wages over the WPI) it captures both of these effects.  I’d rather use NGDP as an indicator of deflationary policies, but lack monthly NGDP data.  But I think the WPI does almost as well, at least during periods where monthly industrial production is moving in the same direction.  Thus during 1929-33, and 1937-38, we observe both monthly wholesale prices and monthly industrial production falling sharply.  It is a pretty good bet that if we had monthly NGDP, it would also have been falling almost continuously during those periods. 

If you invert monthly real wages, and graph it against monthly de-trended industrial production, the correlation is amazing.  Indeed I regard it as the single most astounding graph I have ever seen in business cycle theory.  Especially since modern theory predicts no correlation at all.  But this does not mean that changes in real wages drove changes in IP, nor do I make that claim.  Rather I claim that the two components of real wages, deflationary monetary policies and autonomous nominal wage shocks caused by New Deal policies, caused the changes in industrial production.  My model is completely ad hoc, and I am proud of that fact.  I have a model that fits the Great Depression.  I don’t believe there are “timeless Platonic models” that fit all business cycles.

Understanding the GD is like peeling an onion, after you remove one layer of causality there is always another inside.  So now we know the proximate cause of the Great Depression; deflation, and wage shocks.  But what are the deeper causes of each shock?  The wage shocks were simply policy initiatives by FDR, all the complexity is in the deflationary shocks. 

Recall that the Great Depression was a global phenomenon.  Prices fell almost everywhere.  The theory of PPP provides one explanation of the linkage.  Thus prices fell even in countries like Canada, which did not see a single bank fail.  And of course exchange rates were fixed by the international gold standard. 

Part 2.  How do we evaluate monetary policy under an international gold standard?

Except in 1933, gold was the medium of account in the US during the interwar years.  Thus changes in the price level were identical to the inverse of changes in the value of gold.  This isn’t a theory, it is an identity.  If the price level fell 20%, then ipso facto the purchasing power of gold rose 25%.  Thus to explain why prices fell sharply in the US after 1929, we need to explain why the value of gold rose sharply after 1929.  That can’t be so hard!  And surprisingly, is really isn’t all that hard.  The value of any commodity is determined by supply and demand.  The fact that gold equaled money doesn’t change that at all.

The supply of newly mined gold is pretty stable, roughly 2% of the stock of existing gold.  So to explain any big change in the price level, we will need to focus on shocks to gold demand.  Unfortunately, I didn’t have good data on total gold stocks.  Instead, I relied on monetary gold stocks.  Actually I could have constructed reasonably accurate estimates for total gold stocks, but I found monetary gold stocks served my purposes just as well.  All I had to do was consider changes in the non-monetary demand for gold (hoarding or jewelry) to be factors that affected the supply of monetary gold. 

So there are two things that could have caused the great deflationary shock of 1929-33; more private demand for gold, and more central bank demand for gold.  Both happened, although central banks seemed by far the biggest culprits.  Central banks held gold reserves to back their liabilities, basically the monetary base.  I had to take a few short cuts for foreign countries, where I used currency as a proxy for the base.  This ratio is called the “gold reserve ratio” and is the ratio of monetary gold stocks to currency or the monetary base.    The model starts with this identity:

Price level = (monetary gold supply)/(real monetary gold demand)

P = (monetary gold supply)/[gold reserve ratio*real currency demand]

All three variables on the right side played a role in the Great Deflation.  But the supply of monetary gold actually increased over those four years, so in net terms it pushed prices higher.  Nevertheless, private gold hoarding reduced world monetary gold stocks at certain key moments in 1931-33, and this played a role in the intensification of the Great Depression.  During a depression gold supplies will normally increase even faster than usual, as gold is brought to central banks in response to the increase in the real value of gold.  Private gold hoarding in 1931-33 (and 1937-38) due to devaluation fears slowed this process, and worsened the Depression.

Even so, the big problem was central bank gold demand, which soared for two reasons.  First, people hoarded lots of cash, especially in the US and France.  This was due to low interest rates and fears of bank failures.  More gold was required to back up this increase in real cash demand.  Even worse, central banks dramatically increased their gold reserve ratios.  They essentially hoarded gold.  Under the “rules of the game” this wasn’t supposed to occur.  As gold flowed into your country you were supposed to increase the currency stock in proportion to the rise in gold stocks.  Had this occurred, had central banks played by the rules of the game, it is very likely (though not certain) that the Great Depression would not have occurred.  There might have been some deflation in the 1930s, but much milder than what actually occurred. 

There was almost no currency hoarding during the first 15 months of the GD.  Instead, the initial sharp contraction was triggered by a huge increase in the world gold reserve ratio, which drove prices and output much lower in most countries.  The only major exception was France.  It had recently devalued its currency and prices had not yet fully risen to reflect PPP.  As a result, the onset of the Depression occurred about a year later in France.

Part 3.  Beyond ideology

I’m not interested in producing a narrative that fits anyone’s ideological bias.  Right-wingers may not like the first half of the book.  As I indicated, I think Hoover’s supply-side screw-ups were secondary factors.  If NGDP falls in half, you’ve got a major depression on your hands, regardless of trade, tax and wage policies.  Look how much trouble we’re having adapting to a 3% fall in NGDP in the 12 months after July 2008. 

Left-wingers won’t like the last half of the book, as I attribute the slow recovery to FDR’s high wage policies.  But although I have a right wing reputation, FDR comes off far better in the narrative than Hoover.  Much of it is devoted to FDR’s dollar devaluation policy, which Keynes correctly called “magnificently right.”  In contrast, Hoover did absolutely nothing right.  He didn’t intervene where he should have (monetary policy) and the areas he did intervene just made things worse.  No single ideology was capable of creating a disaster this big (Just as WWII in Europe was jointly produced by Hitler and Stalin.)  The first half of the Depression was a failure of right wing economics, as it was conceived at the time.  The second half was a failure of left wing economics, circa 1935.  If a single screw-up was capable of creating a Great Depression, we would have had many of them.  Instead, this 12 year Depression was three times as long as any other Depression in American history. 

The current recession (which was also caused by a failure of both right and left wing ideas) may end up being the second longest, at least if we measure it from the onset to the point where unemployment falls below some reasonable figure like 8%.  But at least right-wingers no longer believe in the gold standard, and left wingers no longer support the NIRA.  So we won’t have another Great Depression.  We are making all the same mistakes, but in much milder forms.

Part 4.  Proposed Table on Contents  

THE MIDAS CURSE: GOLD, WAGES, AND THE GREAT DEPRESSION

Table of Contents

Preface

Part 1:  Gold, Wages, and the Great Depression

Chapter 1:  Introduction

Chapter 2:  A Model of the Great Depression

Chapter 3:  Monetary Policy under a Gold Standard

Part 2:  The Great Contraction

Chapter 4:  From the Wall Street Crash to the First Banking Panic

Chapter 5:  The German Crisis of 1931

Chapter 6:  The “Liquidity Trap” of 1932

Part 3:  Bold and Persistent Experimentation: Macroeconomic Policy During 1933

Chapter 7:  A Foolproof Plan for Reflation

Chapter 8:  The NIRA and the Hidden Depression

Chapter 9:  The Rubber Dollar

Part 4:  Back on the Gold Standard

Chapter 10:  The Demise of the Gold Bloc

Chapter 11:  The Gold Panic

Chapter 12:  The Midas Curse and the Roosevelt Depression

Part 5:  Conclusion

Chapter 13:  The Influence of the Depression on Macroeconomic Thought

Chapter 14:  Concluding Remarks

In my comment section, I keep getting Keynesian commenters claiming that fiscal stimulus is the only way to boost AD in a recession.  Here is a similar example from the New York Times:

WHAT CAN BE DONE NOW? Here is an unpopular but undeniable fact of life: When private sector demand is weak, the federal government must serve as the spender of last resort. Otherwise, collapsing demand sets in motion a negative, self-reinforcing spiral in which lack of demand — for goods, services and new employees — leads to ever deepening economic weakness.

That is why when the banks and the economy began to crumble in 2008, President Bush responded with a $700 billion bank bailout and a $168 billion stimulus package.

Now I suppose you could argue that they meant it is necessary when interest rates are at zero.  But of course rates were at 2% when Bush implemented this stimulus.  So that argument won’t work.  As we will see, it doesn’t even work when interest rates are at zero.  But first I want to quote from Paul Krugman, as my Keynesian readers won’t trust anything I have to say:

I notice that commenters keep citing this paper by Alesina and Ardagna as if it were a definitive rejection of Keynesian economics. So I guess I should explain why I’m not convinced.

First, the whole stimulus debate is supposed to be about what happens when interest rates are up against the zero bound. Everything is different if the central bank is busy adjusting rates in response to conditions, and may well raise rates to offset the effects of any fiscal expansion. Yet the Alesina-Ardagna analysis doesn’t make that distinction; Japan in the 90s, which was up against the zero bound, is treated the same as a batch of countries in the 70s and 80s, when interest rates were quite high.

Krugman and I believe that if the central bank is targeting some sort of nominal aggregate like the price level or NGDP, level targeting, it is pretty hard for fiscal stimulus to have much effect.  Any anticipated effect would be mostly offset by monetary policy, indeed completely offset if monetary policy is efficient.  Where we diverge is what happens when rates hit zero.  I assume that at zero rates the central bank can still engage in inflation targeting.  Krugman assumes that . . . well it isn’t always easy to figure out what Krugman assumnes.  Here are some options:

1.  Monetary stimulus is ineffective at the zero bound

2.  Monetary policy is potentially effective if central banks are credible, but they often are not.

3.  Monetary policy is effective but central banks simply refuse to do what is necessary, such as unconventional QE and/or inflation targeting.

Lately he has emphasized the 3rd option, which is the most defensible.  No one can seriously claim that the BOJ has been powerless in recent years.  They could always depreciate the yen.  A few decades ago people used to argue that yen depreciation wasn’t an option because the US would object.  But that view is totally indefensible today, for three reasons.

a.  The focus is mostly on China, not Japan.

b.  The yen recently rose from 120 to the dollar to 90.  So even holding it steady would have been far more expansionary, and obviously there weren’t many complaints from the US when the yen was at 120.  Even if you don’t believe that, surely they could have limited the appreciation to a level closer to 110 than 90.

c. Most importantly, the Japanese government is actually pushing for more expansion, and the BOJ stubbornly resists.  Recall that any US pressure is directed at the Japanese government, not the BOJ directly.

There can no longer be any doubt that the BOJ prefers stable prices or mild deflation over the alternative of mild inflation.  This explains why the Japanese fiscal stimulus failed.  Even if, as Krugman implies, the 1995 stimulus worked, all it did was postpone the problem for a year.  As soon as the stimulus stopped the economy slipped back.  You can’t run big deficits forever, so if stimulus is to work the central bank must be accommodative.  The BOJ was not accommodative, hence the stimulus didn’t work.  If the BOJ had been accommodative I suppose you could claim it would have worked.  But of course if the BOJ had been accommodative then fiscal stimulus would not have been needed in the first place.  It is a fifth wheel.

Part 2.  People also got moody when they saw the Titanic was sinking

I recently argued:

I’m not convinced mood swings are as obvious as they might seem.  I’ve argued that the stock market crash of 1929 was a rational response to the sudden awareness that we were rushing headlong into Depression.  I wonder if that stock market crash was one of those examples where Bryan thinks it’s “obvious” there was a mood swing.  Even if Bryan doesn’t believe that, I’d estimate about 99.9% of historians do look at the crash that way.

Bryan Caplan responded:

I’m not going to argue the Depression with an expert like Scott.  But I saw the 2008 crash and subsequent downturn with my own eyes, and I’m convinced that mood played a key role.  The world freaked out, big time.  It was the economic analog of a riot.

But hasn’t Sumner shown that the fundamental problem was falling nominal GDP?   I’m sympathetic, but he never really explains why money velocity suddenly plunged.  (Yes, the Fed started paying interest on reserves, but that’s far from the whole story).  After the 2008 crash, people clearly became much more reluctant to spend, holding their income constant. 

I argued that people were responding to expectations of a fall in NGDP.  The following year NGDP did fall at the fastest rate since 1938.  Bryan is arguing that the public panicked and saw a ghost.  Or more specifically that got very worried about the economy.  Then the economy got much worse precisely because people had become worried.  Self-fulfilling expectations.  BTW, it is rational to be reluctant to spend if your current income is stable, but you expect it to fall very soon.

There are two things wrong with the argument.  First it relies too much on common sense.  When people respond correctly to fundamentals that are hard to discern, it can look like panic.  And the fundamentals are often hard to discern, because if they were easy to spot then people would have seen the problems coming even earlier.  Imagine the point where the smartest person in the world sees a problem.  The market will see it even earlier.  Stocks crashed in early October, not September 2008.  So even in September it clearly was not yet obvious that the economy was about to fall through the floor.  (BTW, don’t say person X saw it before the market.  That’s cheating.  You have to pick the smartest person before the event, not after.  Buffett lost billions in the crash.)

The more significant problem is that Bryan unintentionally ignores the fact that fundamentals drive NGDP in the long run.  And the important fundamental in this case is monetary policy.  The great flaw in Keynes’ General Theory is that it takes the current level of NGDP as a historical datum.  It has no explanation for why current US NGDP is $14 trillion rather than $14 billion or $14 quadrillion.  Instead, the model merely tries to explain changes, starting from a level that is assumed given.  But of course the factor that explains why NGDP is currently $14 billion also will determine what NGDP is three years from now.  And only monetary policy can do that.

So any panic about future levels of NGDP going several years out cannot be a purely self-fulfilling prophecy.   Even if monetary policy operates with a lag (and I think lags are exaggerated) the Fed can always try to steer NGDP back on course at a later date.  But if people expect them to do that, then current NGDP would change much less.  It was pretty obvious that if we had had a NGDP futures market in late 2008, you would have seen NGDP contracts for 1, 2, 3, 4, and 5 years out plunging sharply.  The public wasn’t just losing confidence in current monetary policy, they were also saying, “we don’t expect the Fed to correct its mistake.”  And they would have been completely right.

Bryan also makes an error when he focuses on velocity.  The Fed doesn’t target the base, nor should it.  Under interest rate targeting the base is endogenous, and under the NGDP futures targeting that I favor the base is also endogenous.  The question is not what caused velocity to fall, we know that was an endogenous response to two factors; interest on reserves and expectations for falling NGDP (I won’t argue the relative importance here.)  The question is what caused NGDP expectations to fall.  And that was a failure of monetary policy, as it is their job to keep NGDP expectations on course.  Falling velocity, falling commodity prices, falling stocks, a falling dollar price of euros, falling prices of houses in non-subprime areas, falling commercial real estate prices, were just some of the many symptoms of sharply falling 2010 NGDP expectations in late 2008.

How did the market read the Fed’s intentions back in October 2008?  Why ask me?  The market figured this out way before I did, because they are much smarter than me.  Don’t confuse me with my many commenters who are smarter than the market.  You should ask them how the market correctly saw the Fed would let NGDP stay depressed, and wouldn’t try to bring it back up with the sort of rapid NGDP growth we had in the 1983-84 recovery.  I stand in awe of the market’s wisdom.  If I was that smart (here comes the cliche) . . . I’d be rich.

PS.  I am going to try to stay away from current events for a while, unless I need to respond directly to someone commenting on my posts.

This WaPo story is slightly worrisome:

The Federal Reserve would consider reopening its program to support the mortgage market if interest rates spiked or the economy showed new weakness, Federal Reserve Bank of New York President William C. Dudley said in two new interviews.

Low long term rates are usually the sign of a weak economy, and rates normally rise as the economy recovers.  We saw this during 2009, when rates moved up somewhat after the economy seemed to pick up a bit in the second half.  I would be happier if the Fed was saying that they stood ready to respond with more stimulus if long term rates declined.  

Why do I say that I am only slightly worried?  Because Dudley also mentions that signs of further economic weakness would trigger additional stimulus.  Of course that begs the question:  How weak does the economy have to get before the Fed decides the US would be better off if aggregate demand were a bit higher?

Part 2.  I found the WaPo excerpt in an Arnold Kling post.  Kling makes the following observation:

Rates on 30-year, fixed-rate mortgages are 5 percent. The market wants those rates to be higher. Down the road, the market probably will want those rates to be much, much higher. If so, the ultimate lenders are going to take huge losses, as the interest rates they pay to keep these mortgages in portfolio will exceed 5 percent.

Who will bear these losses? As taxpayers, we will.

I’m no expert on the default risk on these securities, but I’d like to make one observation about interest rate risk, which is the subject of Kling’s remarks.  He is of course correct in noting that if rates rise sharply, the value of the Fed’s MBS portfolio will drop sharply, and ultimately the taxpayers will bear the cost.  I would add, however, that by far the strongest determinant of long term interest rates is the level and growth rate of NGDP.  Rates tend to be low when NGDP is falling, or is rising from a very low level (like right now.)  In my view the sort of macroeconomic environment that would produce much higher interest rates would also produce a robust recovery in the economy.  If we abstract from default risk, and consider Treasury bonds for instance, the worst thing that could happen would be if the Fed made a large profit on the T-bonds it accumulated in the so-called QE program.  That would indicate that long term rates had fallen to Japanese levels, and that for every dollar the Fed gained in higher asset prices, the Treasury was losing $10 in lower tax revenue and higher unemployment and welfare payments.

This is not to excuse the Fed for its purchases of MBSs.  I’d rather they would buy enough ordinary Treasuries to boost NGDP very sharply.  That’s the best way to help housing.  Trying to micro-manage specific sectors almost never works.

Part 3.  Tim Duy; devil’s advocate

Tim Duy has also been a critic of the Fed’s passivity in the face of a severe recession.  In this post he tried to play the devil’s advocate and look at things from the Fed’s perspective.  I think he makes a lot of excellent points, but I want to comment on this assertion:

We need to see sustained growth at more than twice that rate to bring unemployment quickly down to acceptable levels.  But there is simply no faith that such a feat can be achieved.  Most doubt there is sufficient pent up demand in the consumer to do the job, while Calculated Risk has repeatedly stated the case against a quick rebound in housing.  With fiscal stimulus set to slow, that leaves investment and the external sector to big up the slack – neither of which packs the weight of the consumer.  Consequently, the Fed can hold policy steady on the back of the current forecast, which lacks the post-1982 surge.

This is a fairly standard Keynesian approach to evaluating the prospects for a rise in AD.  I read similar things every recession.  But do you recall the phrase “it’s darkest just before the dawn?”  In almost every single recession things look bleak right before a rapid recovery occurs.  If you are deep in recession, and consumers are suffering massive job losses, you’d naturally expect sluggish consumer spending going forward.  The same would be true of business investment, as factories have excess capacity.  I grant you that housing is one area where we are worse off than usual.  In past recessions low rates have sometimes led to an early recovery in that sector.  But I still maintain that this approach to AD is wrong.  It seems to look at AD as a real variable, a collection of sectoral demands that must be added together.

The fastest AD growth in US history probably occurred between March and July 1933, when by all accounts no sector of the economy should have been doing well. We had 25% unemployment.  Where was all that AD going to come from?  Now instead of thinking of AD as a real variable, think of it as a nominal variable; NGDP.  And think of monetary policy (broadly defined as MV, not just M) as the driving force behind AD.  Then the only question is whether or not when NGDP grows rapidly (as it did after March 1933), the growth is prices or output.  It was mostly output in the spring of 1933, and I expect it would be today as well.

In the end I agree with Tim Duy’s conclusion that the recovery will muddle along at a slow rate.  After today’s drop in unemployment I am a bit more optimistic than last night.  But I think the real AD approach (C+I+G+NX) obscures the transmission mechanism of monetary stimulus in a deep recession, leading the casual observer to think more G is the only answer.  If people did draw that conclusion, it would be unfortunate.  At the same time I do understand the appeal of this approach.  Remember I am also a teacher.  I know how easy it is to explain ideas like the expenditure multiplier to students, and how hard it is to get them to grasp the essence of monetary economics–that people’s attempts to get rid of excess cash balances drives AD higher, even if you cannot see the effect in your own behavior.

Part 4.  Brad DeLong solves the age-old problem of estimating crowding out.

Many researchers have tried to estimate the extent to which government expenditures crowd out private expenditures.  We basically know that the crowding out is roughly one for one if at full employment, and also if the central bank has some sort of nominal target, such as inflation.  In other cases it is hard to tell.  As you know I am skeptical of the estimates for all sorts of reasons.  But as David Henderson points out, DeLong basically ignores the ceteris paribus problem in his criticism of a recent post by Steven Horwitz.  Here is Henderson making the sort of comment I had planned to make:

If wages are not falling, then that well could be due to extension of unemployment benefits and some of the additional spending in the stimulus package. DeLong has arbitrarily chosen zero real-wage increase as his baseline. But in a readjustment, what Arnold Kling calls a recalculation, there’s a case to be made for some real wages to fall. At those lower real wages, some of the currently unemployed would be employed. Those jobs that aren’t created, therefore, are a cost of the stimulus package. No one, including Steve Horwitz, claimed that there was a one for one. So the jobs not created by the private sector are indeed a cost of the stimulus package.

I was going to make a similar point about both wages and interest rates.  If DeLong’s evidence was really as definitive as he claims, then the whole crowding out debate should have been resolved long ago.  Just look at wages and interest rates!  They tell us everything we need to know.  No need for messy multiple regressions. 

I’ve never met Mr. Horwitz.  But when I saw him insulted in DeLong’s headline, I knew he must be a fine economist.  One of my proudest days was when DeLong treated me in the same way he treats distinguished economists like Fama and Cochrane.  I still proudly display the post on my office door.  “Scott Sumner simply loses his mind.”

PS.  If you ever forget DeLong’s blog address, just Google ‘Scott Sumner loses his mind.’  It will take you right there.

Bryan Caplan asks the following question:

My main complaint: Scott’s still not buying my simple modification that makes the EMH far more plausible.  It’s pretty obvious that investors’ mood swings matter a lot.  Why not just say that their mood swings are yet another important hard-to-predict variable that the best minds in the world struggle to forecast?

I’m not convinced mood swings are as obvious as they might seem.  I’ve argued that the stock market crash of 1929 was a rational response to the sudden awareness that we were rushing headlong into Depression.  I wonder if that stock market crash was one of those examples where Bryan thinks it’s “obvious” there was a mood swing.  Even if Bryan doesn’t believe that, I’d estimate about 99.9% of historians do look at the crash that way.  But they are wrong, as they don’t understand the underlying fundamentals driving AD.

Just to show I’m not being dogmatic, I admit that the 1987 crash is much harder to explain, so perhaps a mood swing did occur.  Does this violate the EMH?  Not as long as the mood swings are unpredictable.  But it is hard to claim that mood is unpredictable.  That would mean that it followed a random walk.  But variables that follow a random walk tend to eventually drift off toward infinity or negative infinity.

I am pretty sure that most people who believe in mood swings (including Bryan) have in mind some sort of mean reversion.  Think of 100 as a normal mood, and let higher numbers represent optimism and/or less risk aversion, while lower numbers represent pessimism and/or high risk aversion. 

Now suppose the mood index hit 120 right before the 1987 stock crash, and bottomed out at 85 after the crash.  If people could observe current mood levels, the fact that there was some mean reversion would make future movements somewhat predictable.  And that would violate the EMH.  You might ask “Then why do I believe in the EMH?”  The answer might surprise you.  If events like the 1987 crash happened frequently, I would not believe in the EMH.  I think 1987 is a real problem for the EMH.  Even the 2000 bubble had at least the excuse of the massive uncertainty about the future prospects of the internet companies as a plausible explanation for some of the excesses.

Update:  Oops.  The first commenter below (OneEyedMan) pointed out that this is not correct.  It would not violate the EMH, as the changing expected return in 1987 (as stocks fell) would properly reflect the changing level of risk aversion.  So it looks like Bryan was more correct than I thought.  I think at some point we may enter deep metaphysical discussions of what the term ‘efficient’ means, in a world of investors as moody as 14 year old girls.  But since I’ve already made one silly error, I won’t hazard a guess right now.  End of update.

My other problem with Bryan’s argument is that if mood were truly unpredictable, then the EMH would not have to be modified.  Rorty said “that which has no practical implications, has no philosophical implications.”  If mood was a random walk, it would serve precisely the same role as randomly time-varying risk aversion.   It would be observationally equivalent.  But the EMH already allows for time-varying risk aversion.  (Actually I’ve never taken a finance course, so I don’t know that.  Let me rephrase that; the version of the EMH that I believe in allows for random time-varying risk aversion.)

Part 2.  What about those who did see the crisis coming?

Good for them!  That’s been my answer in several comment sections.  This relates to a post by Adam Ozimek: 

I think this is part of the reason Sumner is frustrated by the anti-EMH crowd:  he isn’t distinguishing between those that identify profitability as sufficient proof from those who don’t, so he is faced with some people arguing that “people reliably profit off of bubbles, therefore EMH is false!” and others arguing that “market irrationality prevents reliably profiting off of bubbles, and EMH is still false”. There may be people who hold both of these contradictory points at the same time, and they are wrong. But the existence of two separate arguments against EMH that happen to contradict each other is not evidence against either theory or for EMH.

That’s a good point, and my post was meant sort of half jokingly.  I understand that the two complaints are logically distinct, and somewhat in conflict.  My point was more to show the difficulty with debating the anti-EMH crowd.  No matter what happens in the natural experiment performed by LTCM, it will be viewed as a defeat for the EMH by a sizable contingent of the anti-EMH crowd. 

But he also raised another issue that is more serious:

Take Calculated Risk, for instance. He clearly and specifically identified the housing bubble using his extensive knowledge of the entire housing industry- from realtors to securitizers. Are we really going to demand that he identify the next bubble in gold, oil, Beanie Babies, or some other market he knows nothing about until we accept that he identified the housing bubble? I see no reason to expect this to be true. Nor do I see any reason to expect that all bubbles be equally identifiable. Using Sumner’s criteria, I’m not sure how you tell the difference between a world in which 1/5 bubbles are identifiable and a world in which 0/5 bubbles are identifiable.

I have actually made two separate arguments regarding the people who have become famous forecasters (Roubini, Calculated Risk) or famous investors (Soros, Buffett.)  I believe that group includes people who are both lucky and also people who actually did in some sense beat the market, or forecast something that the market as a whole missed.  But this is important, it is almost impossible to distinguish between the lucky and the smart.  As a result, their forecasts don’t have practical implications for ordinary investors, for public policymakers, or for social scientists like me.  They are flukes.  If they did it reliably we could simply observe where they invested, and mimic their stock investments.  Soros has even written accounts of his strategy.  Does anyone think I would do as well as he did if I read his explanation?

I hope I don’t botch this, as I am relying on memory.  But a commenter named 123 recently sent me a recent article by French and Fama that I think perfectly encapsulates what I am trying say.  Years earlier, some studies showed that the excess returns of mutual funds are not serially correlated.   These studies seem to me to be almost the only reliable test of the EMH that I have seen.  Other studies done by people like Shiller just strike me as data mining.  In any case, if markets could be beaten by smart investors, then mutual funds run by above average managers should do better on average.  Not every year, but on average.  This means that excess returns should be serially correlated.  But they aren’t.  The successful investors during one year tend to just do average the next, suggesting it was all luck. 

Recently Fama and French took a closer look at the data, and found evidence that if you just looked at the top few percentiles, above 97%, there did seem to be a bit of serial correlation.  But even in that group there are also some funds that were lucky.  Interestingly, I am pretty sure that they found that if you could identify which of the top 3% were smart, and not just lucky, you could slightly beat the market.  But if you invest in all funds in the top 3%, you will get a mixture of skilled managers and lucky managers, and you won’t be able to beat an index fund (which has lower expense ratios.)  Odds are I’ve at least slightly misinterpreted their findings.  If so, someone let me know and I will add an update as soon as possible.    

So for me the bottom line is that the EMH is not literally true, indeed no model in the social sciences is precisely true.  There are probably occasional mood swings like the 1987 crash, which set up at least slightly forecastable mean reversion.  And there are probably a few people who do on occasion see things the market misses, but they are hard to identify.  In the end, the EMH is still useful for ordinary investors, regulators and social scientists.  But I hope pundits like Calculated Risk and Nouriel Roubini will keep looking for bubbles on the assumption the EMH is not true, and I hope investors like Buffett and Soros will continue assuming the EMH is not true, so that they can help make it more true (or more useful, if like me you deny objective reality.)

I don’t have the copy at home, but I just finished reading Tyler Cowen’s book on globalization and culture.  I seem to recall that he ended the book arguing that us cosmopolitans can only enjoy sampling from a rich and diverse global environment if many segments of the world’s population refrain from cosmopolitanism, but instead impose constraints on people that generate distinct local cultures, and their associated artistic achievements.  The EMH is like that, perhaps Cowen made the connection himself.

Update 2/7/10:  Commenter 123 just sent me the following:

Fama&French study Scott mentioned is really great, you can see it here:
http://www.dimensional.com/famafrench/2009/11/luck-versus-skill-in-mutual-fund-performance-1.html
But they did not study autocorrelation of excess returns there. Their approach was to have a Monte Carlo simulation of distribution of lifetime fund returns where no funds have alpha, and to compare that simulation with a distribution of actual lifetime mutual fund returns.

People might also want to check out his blog, as he knows far more about the market efficiency literature than I do.  And he often disagrees with me.  So you will get a different perspective.

 

We are now more than two years into the Great Recession, which began in December 2007.  In the Great Depression, this was the point where the Fed decided to raise interest rates to keep the dollar from depreciating (after Britain left the gold standard.) 

Mr.  Bullard of the St. Louis Fed wants to see “at least one month of positive jobs growth” before raising rates.  Maybe it will come out tomorrow, but stocks fell sharply today on worries that the recovery is sputtering out:

NEW YORK (AP) — Stocks buckled Thursday under the growing belief that the global economy is weaker than many investors expected and is likely to stop the U.S. labor market from rebounding in the coming months.

.   .   .

The drop was similar to stumbles the market began having in mid-January. Stocks fell then in response to China’s attempts to curb its overheated growth. Those moves raised fears that the other world economies could suffer as a result. The pullback in stocks worsened as leaders in Washington said they would impose tighter regulations on U.S. banks.

So stocks crashed because investors are worried about Chinese moves to curb demand, and tighter regulation of the banking system.  Interestingly, Krugman has recently been calling for a higher value of the yuan (which is a deflationary policy for China) as well as tighter regulation of the banking system.  He also favors tariffs on big carbon emitters (i.e. China.)

I still think we will avoid an outright double dip (or perhaps I should say that I infer that the markets think this will be avoided, as I don’t do witchcraft.)  But I am increasingly worried about the Japanese scenario (which to his credit, Krugman has repeatedly warned about.)  The CBO just put out an estimate of 2.8% NGDP growth in 2011.  There is no way to overstate how depressing that number is.  During the recovery from the 1981-82 recession we had roughly 10% nominal growth for much of 1983 and 1984.  The normal NGDP growth rate has been just over 5% in recent decades.  In this recession we fell about 8% below that trend, and could now use some rapid catch-up.  Instead, we may fall even farther behind trend. 

Why does any of this matter?  Consider the following from a study of the Great Depression:

We find that, once we have controlled for lagged output and banking panics, the effects on output of shocks to nominal wages and shocks to prices are roughly equal and opposite.  If price effects operating through nonwage channels were important, we would expect to find the effect on output of a change in prices (given wages) to be greater than the effect of a change in nominal wages (given prices).  As we find roughly equal effects, our evidence favors the view that sticky wages were the dominant source of nonneutrality.

Let me explain what Steve Silver and I did when we studied the Great Depression.  We regressed (differences of logs of) monthly industrial production on monthly wholesale prices and nominal hourly wages.  We found higher prices had a positive effective, and higher nominal wages had a negative effect of a similar magnitude.  So it looks like real wages were the problem.  You can get higher real wages through deflation (for a given nominal wage) or through higher nominal wages (for a given price level.)  Either way you get high unemployment.

Actually I tricked you.  The quotation above does not come from the time series paper with Steve Silver that I just described, but rather from a different paper that looked at cross-sectional data for many different countries during the Depression.  Isn’t it interesting that they got almost the same findings, using a completely different technique.  BTW, the authors of the paper I quoted were Kevin Carey, and . . . some fellow named Ben Bernanke.

Late last year I expressed a lot of concern about how difficult it was going to be to get all workers to accept pay cuts that put them 8% below their previous trend line, assuming that NGDP doesn’t fully recover but rather starts a new 5% growth rate from this point forward.  I pointed out that NYC teachers, for instance, were about to get a 4% raise negotiated way back in 2005.  Little did I imagine that even 8% adjustments wouldn’t be enough.  Faster than 5% NGDP growth would allow us to get back to a reasonable unemployment level without painful and prolonged wage cuts.  Instead they’re estimating only 2.8% NGDP growth in 2011.  So after a painful round of wage cuts workers can look forward to  . . .  even more painful wage cuts.

I’m not a leftist, but if I was a conspiratorial leftist I would attribute today’s stock crash to investors waking up to the fact that the pain was spreading beyond the labor markets.  Here’s what a left-wing Sumner would assume was going through the mind of investors :

“For a while we had the economy in a sweet spot.  It was depressed enough to keep labor and other inputs really cheap, but at least we could look forward to steady 5% NGDP growth going forward.  Yeah, workers were suffering, but corporate profits were doing alright.  But now the hawks at the Fed and ECB and BOJ have gone too far.  If NGDP growth slows further, it won’t just be workers suffering; there won’t even be enough revenue to underpin higher corporate profits.  And now the Chinese are even tightening.  China; the one bright spot in the world economy during 2009!  And I just read that the ‘Davos men’ are pressuring China to adopt the Hoover/Mellon strong currency policies.  If the China recovery sputters out, who’s going to be the engine of the world economy?  I’m all for conservative policies, but not so conservative we end up in a depression.”

I’m not saying that’s what happened on Wall Street recently; just saying a left-wing version of me would entertain that hypothesis.

Part 2. Krugman wants a stronger yuan

Paul Krugman is again attacking the weak yuan policy, using a zero sum exchange rate model that takes no account of how the policy impacts world aggregate demand.  But Krugman is not a vulgar protectionist.  In the past he carefully noted that the weak yuan would not hurt the US if interest rates weren’t stuck at zero.  Normally, the Fed could ease policy to offset any negative effects on AD.  Instead, he only seems like a vulgar populist because he has a special “depression economics” model that allows for all sorts of unsavory policy advice once rates hit the zero bound.  Unfortunately, he is still wrong for three different reasons. 

1.  Start with his assumption that currency adjustments are a zero-sum game.  If that were true, and it’s not, then the weak yuan should help China and hurt the US.  Yes, some other poor countries might also be hurt, but they could offset the effect via devaluation.  In contrast, the US and Europe cannot easily devalue without triggering all sorts of international tensions.  But even if this view is right, his advice to the Chinese is wrong.  China has 1.35 billion mostly poor people.  The US has 310 million mostly affluent people.  A good liberal like Krugman should argue for a weak yuan, as the Chinese need the money much more than we do.

2.  The second thing wrong with Krugman’s argument is that (unless he has become a vulgar protectionist in the last few weeks) he himself notes that it only applies to a situation where the Fed is powerless to boost AD.  But even Krugman admits that they aren’t powerless right now.  They could set a higher inflation target.  They simply don’t want to.  So now he wants to impose hardship on millions of poor Chinese workers in export industries because our own Fed (and ECB) are too lazy to lift a finger to boost NGDP growth in the US.  Indeed they are just itching to tighten policy further.  So even if Krugman isn’t a utilitarian liberal like me, even if (unlike me) he doesn’t care more about the poor in China than the middle class in America, it’s still a bad argument, as the solution isn’t a higher yuan, it’s unconventional monetary stimulus in the US.

3.  But it’s even worse than that.  Because even if Krugman doesn’t care about the Chinese, and even if you assume the Fed is powerless, a stronger yuan is still a bad policy.  Why?  Because money isn’t a zero-sum game.  Currency policies are also monetary policies.  A stronger yuan is a deflationary monetary policy.  A weaker yuan is expansionary.  Krugman has spoken highly of FDR’s bold expansionary policies undertaken when he took office, even citing a 2008 AER paper by Eggertsson.  But the FDR policy that was far and away the most influential at boosting inflation expectations was dollar devaluation.  That’s right; he praises his hero FDR for pursuing exactly the sort of “beggar thy neighbor” policy that he wants the Chinese to stop doing.  Just to be clear, I understand that China has begun growing briskly again and that at some point they will need to start appreciating the yuan again to prevent overheating.  But Krugman was certainly wrong about the yuan in 2009, when China experienced deflation.  And the recent stock plunges all over the world, which seemed to follow the Chinese decision to tighten policy in January, should be a shot across the bow of those who want to tighten prematurely.  Krugman quite rightly points to the errors of 1937, when the US tightened policy during a period of rapid growth, but also a period when neither the US nor world economy was yet out of depression.  Let’s not make the same mistake again.  There’s plenty of time to tighten policy when things get a bit better.  If the Chinese recovery is real then by late 2010 a higher yuan will probably be appropriate.  But it is too soon to tighten now.

If China slows, our recovery will also be threatened, regardless of the fact that US exports to China are relatively modest.  China affects the entire world economy.  It has a huge effect on East Asia, on Australia, even on capital goods producers in Germany.  If the entire world economy slows, it will affect the US. 

And the effects are not just from exports, that’s more zero-sum thinking.  It will lower the world Wicksellian real interest rate, and that will drag central banks deeper into a Japanese-style liquidity morass.

I hope I am just overreacting to the stock market tonight (it fell 3.1%), and that we get a “strong” jobs number tomorrow.  (These days ’strong’ would mean 10% unemployment, and a few hundred jobs gained after 8 million lost.)  But please, let’s err on the side of recovery.  I haven’t even talked about all the other problems with slow NGDP growth—more banking problems, debt crises in Southern Europe, and lots more.  Those who have studied the Great Depression know about all the ugly side effects.  And how about this Gideon Rachman’s description of the “best and the brightest” who represent us at Davos:

With the Americans and the Europeans experiencing a crisis of confidence, Davos man was keen to learn from China this year. American businessmen could be heard ruefully contrasting their own “dysfunctional” political system and flaky politicians with China’s decisive and meritocratic leadership.

Oh yes, China has strong leaders who can make the high speed trains get built on time.  Wish we had a leader like that here.  Our businessmen made many similar comments during the 1930s, but not about China.