Home > Europe, Federal Reserve, Politics, Theory > Copernican Economics

Copernican Economics


I start the second quarter on the road in beautiful Cincinnati, but I have some time to pen a few thoughts tonight (aside: will our grandchildren even know what it means to “pen” a few thoughts, or will they “pad” them?)

My first thought is that for the second month in a row, commodities ended weakly and then surged on the first day of the new month. Today, despite weak global data – although ISM was slightly stronger-than-expected – industrial metals and energy led the commodity complex higher. This is of course what we expect as the price level rises: microeconomic concerns of supply and demand in individual markets operate on the real clearing price, not the nominal clearing price; this latter should rise with the transactional money supply (though defining that is sometimes problematic) times money velocity. It continues to amaze me that commodities are as weak as they are, but today the DJ-UBS index rose 1.3% and national gasoline prices reached $3.92/gallon (and are still rising).

My second, unrelated, thought concerns the impact that weakening European growth (Fiat sales were -36%!) will have on the global economic crisis, which is assuredly not over and will not be over until the Euro changes membership or disintegrates completely – and, more importantly, on the participation/support of the United States for European sovereigns and institutions.

Until now, the U.S. has wisely remained above the fray (and the fraying); although the Federal Reserve has aided the ECB, the U.S. has pointedly refused to add its heft to the IMF and demurred against providing other aid. I say that this is wise because the European experiment should succeed or fail on its own merits. It does no good for the U.S. to keep alive a failed institution, whether that institution is industrial (as with GM or Chrysler, one of which ought to have failed) or quasi-sovereign (FNMA or FHLMC, at least one of which ought to have failed) or sovereign. On this final point, at least, the Administration has concluded that the costs of involvement far, far outweigh the potential benefits of involvement.

But if the European crisis starts its next unraveling in the next few months, the U.S. will enter the arena with, I think, a large contribution of support. What is a couple hundred billion when an election is at stake? The current “firewall” has barely enough height, once already-committed funds are deducted, to contain a modest campfire. When Portugal and possibly Spain step up to get aid, it will be a damned-if-you-do, damned-if-you-don’t moment for the Administration because both paths are painful. Americans clearly do not want to subsidize European institutions, and so (in contrast to the usual tin ear this government has displayed with respect to the People’s wishes) it has been relatively easy to deny the IMF the help it desires. But if there is an election in the offing, and a crisis threatens to burst into bloom in August or September? I believe they will quickly choose to take action they can characterize as “decisive action” to avert the crisis and portray themselves as stewards of global economic health. It isn’t clear to me that it will play to the positive for the President’s party, but I think given the Hobson’s choice they will choose to be involved.

And that might spell the end to the dollar’s strength, which is more due to every other developed currency’s weakness than to the strength of our own system.

The third, also unrelated, thought concerns St. Louis Fed President Bullard’s presentation in China last week, which a friend thoughtfully sent to me.  The presentation concerns the question of whether the proper metric for monetary policymakers is not the domestic output gap but the global output gap. See, the problem is that the large domestic output gap is not consistent, as I’ve pointed out ad nauseum, with the fact that core inflation with or without housing has been accelerating for well over a year in the U.S., Europe, UK, and Japan. Thoughtful policymakers by now should have dispensed with the many-times-discredited notion that output gaps matter to inflation. I said in 2008 that the crisis would be an outstanding test of the two main policy theories: one, that money causes inflation; two, that growth causes inflation. For the first time in many years, money growth and economic growth were moving sharply in opposite directions by meaningful amounts. In the event, it has been a slam-dunk win for the monetarist crowd. More money has meant prices rose, even with a huge output gap. And even with declining money velocity. And even with 40% of the consumption basket, Housing, collapsing from a bubble.

While addressing the question if it’s the global output gap, rather than the domestic output gap, that matters (a reasonable question, since we know that something like 65% of domestic inflation comes from common global sources), Bullard’s Powerpoint presentation is initially encouraging. When he cites evidence, he includes the clear observation that “One study for Europe found that the global output gap did not appreciably impact Euro-area inflation from 1979-2003,” and there are several others. Unfortunately, he then argues that since the global economy is now much more fully integrated, maybe it will have an effect in the future. Indeed, he says “the global output gap idea may be the ‘wave of the future’ rather than an explanation for past economic outcomes.”

It is incredible to me that economists can’t bring themselves to simply abandon a theory that has not worked in practice. Not in the 1970s, not in the 2000s, not in Zimbabwe, and so on. They work hard to posit tweaks to the model that can explain the most-recent aberration.

When I was first entering the business, I worked for a company that did technical analysis, and I helped design quant models. One of the greatest sins was to design a model and, the moment it didn’t work in real time, to make a new rule to carve out the recent underperformance. If the out-of-sample test doesn’t work, you need to question the whole theory, from first principles.

But I’ll go further with my analogy. Prior to the development of the heliocentric model of the universe, courtesy of Copernicus, the previous-best theory was that the heavens revolved around the Earth. The problem was that certain observations did not comport well with theory. In particular, astronomers noticed that some stars – dubbed “wanderering stars,” the Greek word for which became “planet” – occasionally seemed to reverse course and head in the opposite direction they had previously been observed to travel. This “retrograde motion” clearly did not agree with a model in which all of the stars were fixed on a spherical firmament that rotated around the Earth. So, astronomers did the only thing they could do.

They fudged it.

Astronomers invented the concept of “epicycles.” These ‘planets,’ it seemed, existed on other spheres that rotated in the opposite direction but which were attached to the grand sphere. Epicycles made the theory fit the observations. Of course, it was wholly wrong, and Copernicus in time developed a model that was entirely consistent with observation without needing “epicycles” – simply by noting that if the Sun, rather than the Earth, was the center of the universe then we could observe such phenomena.

President Bullard, abandon this theory. You were so close to saying it! Just say that the theory that output gaps affect nominal prices (as opposed to real prices and exchange rates of the factors of production, which they can reasonably affect) doesn’t fit the observations, and be the Copernicus of economists.

None of this has anything to do with market action, but then market action these days has very little to do with anything in the news. Equities will eventually falter, and then the comeuppance could be severe; bonds rallied today but the future is grim aside from occasional flight-to-quality or Fed-frontrunning, and the risks of downside relative to potential upside gains make long duration positions foolhardy. And everyone loses to inflation, which continues to accelerate. These are the trends, although other people would read the trends as being “bullish stocks, bullish bonds, and inflation to fall.” I question the prevailing wisdom.

Advertisements
  1. bixbubba
    April 2, 2012 at 10:42 pm

    Hi Michael,

    Love your comments, as always. Just one little little comment, from a history of science nerd to an econ nerd: Copernicus didn’t do away with epicycles. Ptolemy invented epicycles to account for the retrograde motion of the planets (the time of year when then shift directions and go backwards in the sky). When you move the sun to the center of the system, you can explain the _idea_ of retrograde motion motion without epicycles (and that was Copernicus’ genius) but you cant get the details just right without just as many epicycles as Ptolemy used. That’s because Copernicus still used circular motion. You cant get rid of the epicycles until you follow Kepler and move to elipses. There’s a great quote of Kepler’s about circular motion being a wonton wench that beguiled all the astromers for centuries, but I digress…

    • April 3, 2012 at 2:33 am

      Thanks Bubba! As usual, I can’t let the facts get in the way of a good story! I had the whole story right, although I usually mention Ptolemy when I use this example…except that I didn’t realize Copernicus still needed epicycles! Of course, you’re right about the need to do so w/o the ellipses. Although I wonder, given how close the motions of the planets are to circles, if computationally they would have been able to discern that little difference? In other words, did Copernicus KNOW that the model didn’t QUITE work?

      • bixbubba
        April 3, 2012 at 9:39 am

        Yes. Copernicus knew. Astronomy was incredibly sophisticated in those days. And they had very detailed observation sets going back hundreds of years. ( A lot more data than economists have!) Most of Copernicus’ book was devoted to working out the exact structure of epicycles needed. In fact, saying it didn’t QUITE work is a vast understatment. Copernicus needed just as much epicycle structure as Ptolemy did. His only advantage is that IN PRINCIPLE, he could explain retrograde motion without an epicycle. But he had a few disadvantages too! Why dont we feel the earth moving?? This is way before Galileo, so there is no physics to explain how this could be. And also, why don’t we see stellar parallax? The true answer is that the stars are thousands of trillions of miles away!! But such an answer would have been laughed at at the time.

        It takes about 100 years for Copernicus model to get accepted, and contrary to popular belief, this was not primary because of stubbornness or religious dogma (though there was plenty of both). Its because the best evidence wasn’t really on his side until at least 1610, and arguably not fully until 1687.

  2. April 3, 2012 at 4:47 am

    Michael,

    What about the claim that no inflation is possible until wages start growing?

    • April 3, 2012 at 12:30 pm

      Wages follow, not lead, inflation. You can probably search on my blog for wages and inflation and find some of the things I’ve written about that hypothesis.

  3. Jim H.
    April 3, 2012 at 8:17 am

    ‘One of the greatest sins was to design a model and, the moment it didn’t work in real time, to make a new rule to carve out the recent underperformance. If the out-of-sample test doesn’t work, you need to question the whole theory, from first principles.’

    Probably every economist gives lip service to the principle that curve-fitting is bad. But it doesn’t stop them endlessly doing it. Why is this?

    Most econ papers, with their elaborate statistical panels, are of little interest to me because THEY DON’T MAKE ME ANY MONEY.

    Apparently, only real-world experience building profit-seeking quant models can instil the visceral instinct that self-deception just doesn’t pay. Constantly having to move the goal posts, and add ‘that don’t count’ exceptions, means the model had weak explanatory power to begin with.

  1. December 27, 2013 at 12:00 pm

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: