Archive for September 2010


There’s only one way to test the EMH

Before beginning, I like to point out that I have a new essay over at The Economist.

Commenters often present me with market anomalies, which supposedly “prove” the efficient markets hypothesis is wrong.    I always respond that they’re just engaging in data mining.  They retort that no theory that can’t be disproved is worth anything.  But the EMH can be disproved.  The tests have been done, and it passes.  Sort of.

Finance professors have done many different tests of the EMH, and I’d guess 90% of the published tests (but only 5% of the actual tests) show that the EMH is wrong.  (Yes, I’m pulling these numbers out of thin air, but you get the point.)  They’ve found January effects, small stock effects and value stock effects.  They’ve found the market does better when P/E ratios are low.  They found the market does worse on rainy days (a study published in the AER!).  Of course you’d expect to find 5 anomalies for every 100 tests you do, and for the most part you only get published if you find an anomaly, and finance professors have a lot of computer power, so . . .

Here’s my analogy.  Suppose Stephen Wynn was concerned that some mysterious gamblers were getting away with cheating at one of his casinos.  He’d heard rumors, but had no proof, or even suspects.  You are statistician brought in to investigate.  You study 600 slot machines, and find 30 of them produced three cherries more often than you’d expect from mere chance.  You suggest that the anomalous slot machines be destroyed.  How should the casino owner react to that “investigation?”  I’m guessing Wynn wouldn’t be impressed.

I think you see the problem.  The investigator assumed he could figure out how the cheating was occurring.  He assumed that certain individuals had spotted slot machines with predictable tendencies, and that he (the investigator) could also find those machines.  But it’s really hard to walk into a casino and walk out with lots of money.  So how likely is it that an academic statistician is going to know where to look for cheating?  How likely is it that they would know how it’s done?  Not very, and that’s why their statistical search for anomalous machines will almost certainly be fruitless.  Oh, the data mining will unearth a few suspicious machines, no doubt about that, but only the number predicted by chance.

So how would one determine whether people were cheating?  Start by being humble—admit that you probably don’t know how it’s done.  If someone did have a secret formula, they wouldn’t put it in a book called “How to cheat at casinos” for all the world to see.  Any system that works is almost certain to be kept secret.  It’s like alchemy.  If anyone did have a formula to turn lead into gold, they would keep it secret; as the formula’s value would plunge to zero the instant it was released.

To find out whether cheating is occurring you need to look at whether the winnings of gamblers are serially correlated.  Are those who win once, more likely to win next time.  That’s the proper test, indeed the only practical test, of whether people are cheating the casino.  And it’s also the only test of the EMH.  Don’t look for “the system,” the secret way to beat the stock market.  Look for whether other people have found it.  Look to see whether people who did better than average one year, tended to do better than average the next year.  Don’t look for market anomalies—look for evidence that other people have found market anomalies.

Of course the study has been done.  I recall that Fama and French found that mutual returns were approximately serially uncorrelated, but not exactly.  There appears to be a slight serial correlation among the very best funds (top 3%), but not enough to give the average investor any advantage.

And that’s what I would have expected.  The EMH is approximately true; indeed it’s almost impossible for me to imagine any other model of financial markets.  But it’s not precisely true, again, just as you’d expect.  After all, if the EMH were perfectly true then no one would have any incentive to estimate fundamental values.  We know people are imperfect and hence that any real world human institution, including markets, will be at least slightly imperfect.

A smart person like Eugene Fama should have been able to come up with both the EMH, and its limits, by just sitting in a room and thinking.  Much as David Hume got the QTM by imagining what would happen if everyone in England woke up one morning with twice as much gold in their purses.  Or Fisher’s theory of inflation and nominal interest rates.  Or Cassel’s purchasing power parity.  Or Friedman/Phelps’ natural rate hypothesis.  Or Muth and rational expectations.  Certain ideas are simply logical, and that’s why I have no doubt that despite all those economists on the left arguing the EMH has been discredited, it will still be taught in every top econ/finance grad program 100 years from now, whereas fiscal stimulus will be long gone from macro textbooks.

PS.  Why will fiscal stimulus be gone?  Because even Krugman admits it only makes sense at the zero bound.  And we are rushing headlong into a world of all electronic money–probably within 50 years.  There is no zero lower bound with electronic money, and hence the Taylor Rule is all you need.  Old Keynesian economics will vanish, leaving only new Keynesianism.

Lost in translation

I have a friend who teaches development.  He tells me that he would often run into the following problem.  He’d carefully work through the logic of some argument, explaining how A implies B, and B implies C. The student would nod his head in agreement.  At the end he’d say; “So here’s the policy implication, right?”  And the student would shake his head and say; “No, you don’t understand, China is different.”  (BTW, it could have been any country.)

Both Paul Krugman and I have felt this frustration when looking at the actions of the Fed, ECB, and Bank of Japan.  It goes sort of like this:

You say inflation is currently below target.  “Yes”

You say inflation is expected to remain below target for years to come.  “Yes.”

You say output is below the natural rate.  “Yes”

You say output is expected to remain below the natural rate for years to come.  “Yes.”

Ergo monetary policy should be made more expansionary.  “No.  You don’t understand, this cycle is different.”

I’ve racked my brains for an answer.  So has Krugman.  Once I even delved into pop psychology, arguing that monetary policymakers needed to have a relentlessly logical attitude in applying theory to policy.  No common sense judgments about whether interest rates “seemed to low.”  I argued this was similar to the autistic cognitive profile–blocking out any extraneous factors.

Stefan Elfwing just sent me a Google translation of an article on Lars Svensson, which led me to recall all this frustration:

Lars EO Svensson wants the federal funds rate is 1.75 percent in three years, not 3.8 percent as the Riksbank’s majority wants.

He was also against the latest interest rate rise from 0.50 to 0.75 percent.

His colleague on the Executive Board, Karolina Ekholm, voted for the increase but also want a lower interest rate over the next few years because of the doubts that exist about the global economy.

For Lars EO Svensson is the “obvious” to select a rate and interest rate forecasts that are considerably lower than the Executive Board a majority of four members chosen because it gives better effectiveness.

Our forecasts for CPIF inflation below the target most of the period and resource utilization is low.  If the interest rate path is lower we get a better effectiveness.

That sounds very logical.  So why don’t his colleagues agree?

Lars EO Svensson has repeatedly asked his colleagues on the Executive to explain why they disagree with, he has asked them to point out errors in their reasoning.  The answers are lacking.

Unfortunately I don’t speak Swedish, and I think something got lost in translation.  But I gather the other members of the board have not been able to point out errors in Svensson’s logic.

He points out that although the Finance Committee stated that it would be desirable that the Riksbank is very clear about why various policy decisions taken or not taken.

I am surprised by the disagreement about some things that should be obvious, for example, that one should not raise rates when the forecast for inflation and resource utilization is too low.

That’s why transparency is so important.  It isn’t because a central bank doesn’t know what it should be doing; it’s needed to make sure the central bank in fact does do what it knows it should be doing.

Lars EO Svensson has been accused of being an academic theorist who hide behind their models and research reports. Especially when he, in the midst of financial crisis, would have zero interest, and led a discussion about having a negative interest rate.

I see no contradiction between theory and practice.  Monetary policy should not be controlled by the brain without the ghosts of the facts and analysis.

I guess Svensson’s fellow board members did not benefit from being educated by John Stuart Mills’ father:

I recollect also his indignation at my using the common expression that something was true in theory but required correction in practice; and how, after making me vainly strive to define the word theory, he explained its meaning, and shewed the fallacy of the vulgar form of speech which I had used; leaving me fully persuaded that in being unable to give a correct definition of Theory, and in speaking of it as something which might be at variance with practice, I had shewn unparalleled ignorance. In this he seems, and perhaps was, very unreasonable; but I think, only in being angry in my failure.

Svensson continues:

The last interest rate path should not uncritically be taken as a starting point.  Any interest rate decision should stand on its own and not lean on the previous.

In a post entitled “Losing Face” I showed that central bankers don’t like to admit that their previous policy decision was wrong.

Svensson continues:

Although my colleagues do not seem to want to listen to his argument, he has not tired of sitting on the Executive Board.

Not a minute have I regretted that I took the job.  It was a challenge I could not say no to.  I got the chance to turn 15 years of research in practice.

All six members have an individual responsibility to make independent decisions and justify them.

My job is to do what I do and continue to do so.

Lars Svensson is one of the few policymakers who will come out of this debacle with his reputation intact.  I’m surprised Krugman hasn’t mentioned him in his blog posts.  (They are colleagues at Princeton.)

PS:  Two Swedish posts in one day!  What other American blog provides such complete coverage of Nordic events?  My half Swedish–half Norwegian grandmother would have been proud.

What do we mean by “economic reform?”

The term ‘economic reform’ implicitly meant more government between 1875 and 1975, and has implicitly meant less government since.  That’s how we can tell whether or not we are still in the neoliberal era.

When the recent financial crisis first hit I was worried that it would push us toward statism, just as in the 1930s.  And indeed that seems to have occurred in the US.  By the spring of 2009, however, I noticed that the rest of the world seemed to be moving in the opposite direction.  Parties of the left kept losing in places like India (the communists), Argentina, and Germany.  At the time many commenters said I had things wrong, that I was misinterpreting complex local conditions.

Now I think we now have enough information to know I was right.  The financial crisis has turned out to be a huge problem for parties on the left.  Here’s a recent headline on the election in Sweden:

(Reuters) – The crash of Sweden’s long-ruling Social Democrats to their worst defeat since 1914 highlights the decline of socialist parties in much of Europe, drained by social change, economic crisis and the rise of new issues.

Obviously Sweden will continue to be far more social democratic that the US.  But the effect of the crisis has been for countries like Sweden, Denmark and Holland to trim back government.

Look at how envious the British are of Sweden:

For decades, conservatives have played an important role in Swedish politics: they are there to be defeated. They advocate lower taxes, and are duly accused of planning savage cuts. So the voters traditionally stick with the Social Democrats who have held power for seven of the last eight decades. Every other decade Swedish conservatives come on for some light entertainment, before being booted out after a term. Never in modern Swedish history has a conservative prime minister been spared this fate. Until now.

This week Fredrik Reinfeldt, a bald and deeply dull 45-year-old who communicates with David Cameron by text message, is celebrating the first re-election in history of his party, the Moderaterna. He is also celebrating the success of an extraordinary experiment. His response to the recession was to cut taxes, a move his critics said the country could not afford. The European Commission warned him it would end in tears. But instead, the lower taxes were a spur to growth and Sweden now has the fastest-growing economy in the Western world.

When elected four years ago, leading a four-party coalition, Reinfeldt had a striking slogan. ‘We are the new workers’ party,’ he said, meaning he would cut taxes for those in employment, but not for those on benefits. When faced with protests about how the poorest would be paying a higher marginal tax rate, he appealed to voters’ innate sense of fairness – and resentment at the high level of welfare dependency. At every stage, his ministers would explain the basics of low-tax economics. Cut tax on wages, and you increase the incentive to work. ‘This will increase employment,’ Reinfeldt said. ‘Permanently.’

Not that he was believed – at first, anyway. The party fell 20 points behind in the polls, and braced itself for the ritualistic electoral ejection. It carried on regardless, with tax cuts for cleaners and baby-sitters (most home helpers were paid ‘black’, as the Swedes say, because the tax was so high). Tax on low-paid jobs fell sharpest. Nursing assistants, for example, saw their tax bill drop by a fifth. The aim was to make work compete more aggressively with Sweden’s famously generous welfare state.

Taxes for the rich also came down. Reinfeldt abolished the notorious wealth tax, which took 1.5 per cent a year from any Swede worth over about Skr1.5 million (£125,000). Anders Borg, the finance minister, faced predictable protests about a Bush-style tax cut for the rich. He replied: ‘The big winners are, in the long term, all Swedes, because we must create conditions for companies to match global competition.’ So while the Tories were endorsing Gordon Brown’s plan to increase the tax on the rich, the Swedes were cutting the tax rate – in order to collect more from the well-paid.

When the recession came, Sweden was badly hurt, as one would expect from an export-orientated economy trading with a stricken continent. But the damage was limited because Sweden had properly regulated financial institutions (having been stung by a serious financial crisis in the 1990s. Banks which had embarked on misadventures in the Baltic cleaned up their own mess. The government entered the recession with a surplus. A Gordon Brown figure would have been impossible in Sweden because its laws prohibit politicians running up the national debt in boom years.

Like most of Europe, Sweden launched a stimulus, but Reinfeldt set aside two thirds of his for a tax cut. Corporation tax fell from 28 per cent to 26.3 per cent [the US rate is over 40%], taxes on jobs were cut further still while income tax thresholds were raised. Determined not to let a crisis go to waste, he declared the tax cuts permanent. So while Brown was planning to increase National Insurance, the Swedes were doing the reverse and explaining why. ‘If you tax work higher, you will get fewer people in work,’ said the education minister, Jan Björklund. ‘But if you tax work less you will get more in work.’ This was a battle of ideas, and it was a battle that Reinfeldt and his coalition allies were winning.

In April last year, his party pulled ahead in the polls for the first time since his election. As election day approached, it became clear just how effective his tax cuts had been. Unemployment never hit the forecast 10 per cent – it was 8 per cent in July and 7.4 per cent in August. Two think tanks have confirmed Reinfeldt’s assessment that his tax cuts have created some 100,000 jobs. The deficit was tumbling as the economy recovered. Extraordinarily, Sweden has now overtaken the United States and reached second place in the World Economic Forum’s yearly competitiveness rankings.

Ask about the recession in Sweden now and you are met with a blank stare. Consumer confidence is at a ten-year high. Its recession was steep, but shorter than its downturn after 9/11. In fact, this had come to worry Reinfeldt. ‘Our medicine may have been just too effective,’ one of his state secretaries told me over the summer. ‘Voters don’t think about the economy now. The recession is becoming a distant memory.’

It helped that the deeply unpopular Mona Sahlin was leading the Social Democrats. It also helped that her party was intellectually exhausted. The left has had to embrace ‘free schools’ (the model adopted by Michael Gove) because they are now so popular among parents. The Social Democrats spent much of the campaign trying to persuade voters to forget they ever opposed them. ‘Pupils should choose schools,’ ran one of its posters, ‘schools should not choose pupils.’

The Social Democrats’ main economic argument – that tax cuts mean vicious spending cuts – was exposed as false by the Reinfeldt recovery. Ms Sahlin did not dare to propose reversing what had been the sharpest tax cuts in Swedish history.

And finally, a little comic relief from The Economist:

But Maria Wetterstrand, the Greens’ co-leader, demurred, saying her party could not support a government that “kicks sick people off health insurance, does not have a climate policy and wants to build ten new nuclear power plants.”

PS:  Why did we go the other way from the rest of the world?  I’ve always thought our liberals were motivated by unfinished business.  We never really completed the welfare state, leaving millions without health insurance.  Matt Yglesias says that now that Obama has finished the job, the real need is efficient neoliberal reforms, not even bigger government than currently forecast.

Six reasons to oppose inflation targeting

Tyler Cowen and James Surowiecki have recently discussed why it is difficult for the Fed to set a higher inflation target.  They’re right.  All you need to do is read conservative press accounts of the recent Fed statement, which figuratively roll their eyes at the suggestion that we might need a little more inflation.  “Imagine that!  The Fed thinks we need more inflation.  Don’t they know inflation is a problem?”

Of course this is frustrating to macroeconomists, because if we are targeting inflation at say 2%, then it stands to reason that approximately one half of the time inflation will be too low, and one half of the time it will be too high.

But the articles that mock the Fed are not written to be read by macroeconomists, they’re to be read by the general public.  And the public does believe that inflation is a problem, for several very good reasons.  First, for any given nominal income, higher inflation will tend to hurt an individual worker.  And second, most of the major increases in inflation have been associated with bad times.  Think about 1974 or 1979, or the first half of 2008.  Of course all those were associated with adverse supply shocks.  You have to go back to the second half of the 1960s to find a major increase in inflation that was associated with good times.

On the other hand, if AD rises briskly and brings inflation from below normal back up to normal, most people feel better off.  So when the Fed says it would like to see higher inflation, what it is really saying is that it would like to see higher AD, which as a side effect will raise inflation somewhat.  For any given level of AD, higher inflation would be a bad thing, as it would represent an adverse supply shock.  So why not just call for higher AD?  We know that an increase in AD raises both prices and output, so it seems like what they really want is more NGDP, not higher prices.  Higher NGDP is definitely needed in a demand-side recession, whereas higher inflation might or might not be a good thing.

Consider the following 7 questions:

1.  Which target best measures what the Fed is directly trying to influence, NGDP or inflation?

Answer:  NGDP targeting.  For a given level of NGDP, higher inflation would actually be harmful.

2.  Which policy goal sounds better to the public, higher inflation or higher nominal NGDP?

Answer:  Hmmm, would you rather tell the public you are trying to boost their incomes back to prosperity levels, or that you are trying to raise their cost of living?

3.  Which policy is more consistent with the Fed’s dual mandate?

Answer:  NGDP targeting

4.  Which policy doesn’t force government bureaucrats to make high subjective estimates of quality changes in products?

Answer:  NGDP targeting

5.  Which policy is most like to prevent bubbles; NGDP or inflation targeting?

Answer:  NGDP targeting, which calls for lower inflation during booms.

6.  Which target best avoids liquidity traps, NGDP or inflation targeting?

Answer:  NGDP targeting, as 1% deflation may or may not result in a liquidity trap, depending on the trend real GDP growth rate.

7.  Which target is preferred by the Fed and most macroeconomists?

Answer:  Inflation targeting.

The public instinctively feels the Fed should be targeting NGDP.  And if they were, inflation really would be bad.  Always.  The public’s not stupid; they’re one step ahead of macroeconomic elite!  🙂

Update 9/27/10:  David Beckworth pointed out several advantages of NGDP targeting that I forgot to mention.  It is easier to implement than a Taylor Rule, and it has been advocated by a number of respected macroeconomists, both center-left and center-right.

Back to aggregate demand, which has been the #1 problem all along

Long time commenter David Pearson recently made a couple of excellent points in the comment section of this post:

The epitaph for “Credit Easing”?

“Although financial markets are for the most part functioning normally now, a concerted policy effort has so far not produced an economic recovery of sufficient vigor to significantly reduce the high level of unemployment.” (from yesterday’s Bernanke Princeton speech)

Then he made this observation:

Also, Bernanke argues-correctly, IMO-that “Credit Easing” already succeeded in moving funds out of safe-haven assets. This was a boon to corporate debt issuance, but unfortunately, it did very little for corporate investment (as companies simply used the proceeds to refinance existing debt or build up cash levels):

“Mr. Bernanke noted that by withdrawing supply of mortgage bonds on the market, the Fed was able to get investors wanting to hold safe instruments to move into higher-quality corporate bonds, thus lowering yields there.”

This would seem to support the view that actors have no problem moving out the risk curve. There are plenty of opportunities for companies faced with attractive investment projects to access funds from creditors and invest them. For goodness sake, Petrobras just placed a $70b equity offering! Yet some continue to argue that the issue is “demand for safe haven assets”-some sort of extreme risk aversion.

“Credit Easing” did not try to directly stimulate AD; it tried instead to produce a working monetary transmission channel. Bernanke’s work on the GD supports the idea that he would be extraordinarily concerned with getting this channel to function. Now it seems that he admits that it is functioning, and therefore he turns to the problem of a shortfall in AD.

Which has been the #1 problem all along.  Of course there really was a severe banking crisis in late 2008, and I have an agnostic position on how that should have been addressed.  But it’s not an either or question, the Fed needed to address AD, regardless of what they did for banks.

Just to head off any “can’t walk and chew gum at the same time” arguments, recall that there is no such thing as “not using monetary policy.”  The Fed always can and does set a monetary policy stance.  When the FOMC met in September 2008, it would have been no more difficult to set the fed funds rate at 1%, as compared to leaving it unchanged at 2% (which is what they did.)  And in October it would have required even less effort to not pay interest on reserves, than to start up the IOR program.

I only met Mr. Bernanke once on my life, when he presented a seminar at Bentley discussing his theory that credit problems were a major factor depressing output in the early 1930s, even apart from the monetary shock.  I asked him just one question; how bad would the Great Depression have been if the Fed had targeted steady NGDP growth, but the bank failures had occurred anyway?  As I recall, he hemmed a hawed a bit in answering the question.  I still don’t think the Fed has a good answer to my question from way back in the early 1990s.

A few of us were vigorously arguing for much more monetary stimulus in October 2008.  The Fed wasn’t paying attention, nor were 99% of macroeconomists.  I still haven’t heard a good explanation for why that happened.  Why are we only now considering more stimulus?  Was there no AD problem in late 2008, when NGDP was falling fast?  And don’t say the Fed didn’t realize we had an AD problem—their own forecasts called for nominal growth rates far below any implicit Fed target.