October 08, 2003

Nobel Prize Rumor du Jour

The rumor is that it is going to be Jagdish Bhagwati and Paul Krugman for work on international trade.


Nope. Clive Granger and Robert Engle. Reuters reports: "U.S. economist Robert Engle and Britain's Clive Granger have won the 2003 Nobel economics prize for inventing models used to evaluate investment risk and study the relations between simultaneous economic phenomena. 'This year's Laureates devised new statistical methods for dealing with two key properties of many economic time series: time-varying volatility and non-stationarity,' the Royal Swedish Academy of Sciences said on Wednesday. The two men share the 10 million Swedish crown ($1.3 million) prize, which has been awarded since 1969.

"Engle, born in 1942 in Syracuse in the state of New York, teaches at New York University and got the prize for improving methods of evaluating portfolio risk and asset pricing. 'Investors and financial institutions need forward-looking evaluation -- forecasts -- of volatility during the next day, week or year. Robert Engle formulated a model which allows such evaluations,' the Academy said. Granger, born in 1934 in Wales, is an emeritus professor at the University of California [at San Diego]. The academy said his work was used in studying links between wealth and consumption, exchange rates and price levels, and short and long-term interest rates."

Posted by DeLong at October 8, 2003 10:40 AM | TrackBack

Comments

That would definately bring great gnashing of the teeth to the right....schradenfreude is fun.

Posted by: fester on October 7, 2003 03:50 PM

When would this be announced? I've got to buy a copy of the WSJ the next day, just to watch them whine.

Posted by: Barry on October 7, 2003 03:54 PM

Bhagwati and Krugman, but not Dixit? Odd... but I'd still be overjoyed.

Posted by: Kimon on October 7, 2003 03:56 PM

Where's that futures market in Nobel Prize Winners?

Posted by: jerry on October 7, 2003 03:59 PM

Bhagwati and Krugman, but not Dixit? Odd... but I'd still be overjoyed.

Posted by: Kimon on October 7, 2003 04:01 PM

Where's that futures market in Nobel Prize Winners?

Posted by: jerry on October 7, 2003 04:04 PM

Further proof that the Swedes are decadent and depraved, perhaps even in bed with the Communists. It would be a clear slap in the face to the Great Leader. (I very much enjoyed the wailing and gnashing of teeth over at NRO when Carter won.)

(Heh. I very much doubt that his NYT column is among the reasons to consider him, but shallow minds conflate things so very easily.)

Posted by: Saam Barrager on October 7, 2003 04:14 PM

Agreed. The omission of Dixit would be odd--especially since the Swedes seem to like to do things in threes...

Brad DeLong

Posted by: Brad DeLong on October 7, 2003 04:23 PM

Luskin will call for the immediate invasion of Sweden.

Posted by: Jon h on October 7, 2003 04:29 PM

Barry - the announcement is "Wednesday, October 8, 4:00 p.m. at the earliest" (Swedish time), a few hours after the far more interesting Chemistry announcement. See http://www.nobel.se for details.

Posted by: Tom Slee on October 7, 2003 04:45 PM

"especially since the Swedes seem to like to do things in threes..."

It's an upper limit not a specific preference - they're only allowed to give a Nobel prize to at most three people.

Posted by: Andrew Kanaber on October 7, 2003 05:09 PM

How accurate has the rumor mill been in the past?

Posted by: richard on October 7, 2003 05:50 PM

Makes sense. Both Krugman and Bhagwati made serious contributions in trade theory. I think Krugman's full of it on a lot of his political columns, but his contributions to trade theory are clear. If you think of a Krugman win as a rebuke to Bush, then you just don't know much about the recent history of economic thought. I think it's funny that Krugman's biggest fans here on the comment thread tend to be so ignorant of his actual quality work, and worship his silly little columns. It's as if Emeril spends hours cooking up a dish and you never bother to eat it, but then bends over and craps on a plate and you gobble it up.

Posted by: Keith on October 7, 2003 06:03 PM

If this comes to pass (and I dearly hope that it does), I too am sure that the conservatives with flail about in anger for some time about this. Imagine, one of the very demons of their pantheon wins the nobel prize. As one of the conservatarians might call it, 'triumph of the shrill'.

Posted by: non economist on October 7, 2003 06:23 PM

yah, they'd give it to dixit, too, then, i would assume.

Posted by: Atrios on October 7, 2003 07:03 PM

For those of us that are clueless, could someone post a link as to the contributions of the winners? Including Dixit. I'm willing to do some reading to mitigate my ignorance.

As for the WSJ, they will probably just say something like "Krugman was OK when he stuck to economics."

Posted by: Alan on October 7, 2003 07:12 PM

Oh man. I would love for Krugman to get it. That would be sweet.

Posted by: Cindy on October 7, 2003 07:17 PM

Who leaks Nobel picks?

Posted by: EssJay on October 7, 2003 07:29 PM

I don't care if Krugman is the worst economist ever. If he gets it, it will help wash away the bad taste of Governor Arnold.

On the other hand, how many divisions does Krugman have?

Posted by: mike on October 7, 2003 07:30 PM

What can I say?

You used Emeril as your example of world-class chef.


And let's not forget that the head that created all of that quality work sees nasty, nasty things going on in Bushland.

Posted by: mario on October 7, 2003 07:32 PM

What can I say?

You used Emeril as your example of world-class chef.


And let's not forget that the head that created all of that quality work sees nasty, nasty things going on in Bushland.

Posted by: mario on October 7, 2003 07:38 PM

So, his serious economics work is worthy of a Nobel prize yet his observations about the Bush administration are "silly" Oh boy!

Posted by: SW on October 7, 2003 07:41 PM

What's funny is that if you read his old columns and writings and his new columns and writings, Keith, he hasn't changed what he's been saying noticably since 1994.

About the only new thing is some skepticism about the efficacy of free movement of capital.

Unless you believe that his popular and technical work are somehow the work of different people, methinks there's some problem in thine reasoning.

Posted by: Demosthenes on October 7, 2003 07:48 PM

A bad economist getting the Nobel Prize would wash away the bad taste of a bad politician becoming governor of California? I suppose it makes sense in a way, sorta like how if you bang your elbow against something hard, then rub it, it makes the pain hurt less because your nervous system is overwhelmed by all the signals.

Not that I actually *do* think Bhagwati or Krugman are bad economists, mind you. (based on mindless parroting of authority figures, I think they're good economists, though I've never read their work, besides Krugman's pop-economics books, like The Accidental Theorist, Peddling Prosperity, and the Return of Depression Economics.)

Posted by: Julian Elson on October 7, 2003 08:39 PM

Funny how when I predicted this, a few years back - circa 2 BCR (before the Conservative revolution), a few of my peers looked at me like I was the dummest kid in town... Similar experience with predicting George Akerlof. Some people must have a problem with great minds...

BTW, are we going to read soon that from now on the White House will reserve the right to preemptively & unilaterally announce NB's in Economics? :-)

Posted by: Jean-Philippe Stijns on October 7, 2003 08:57 PM

I agree, Demosthenes. When perusing some of PK's earlier "popular" works, it seems there is continuity of approach, even with the NYT Op Ed columns.

Much of the non-scholarly work seems to centre on his usual "debunking" of myth (whether it be an assault on supply-side or strategic trade policy entrepreneurs). For the most part, most of his works have relied on publicly available data or well know and generally accepted economic principals. Not much in the way of voodoo.

One of the things I most notice when I read his detractors' rebuttals is that they rarely, if ever, address the core evidence presented to undermine their arguments. The responses usually comprise of more myths, personal attacks or simply sidestep the issue altogether.

Posted by: Stephane on October 7, 2003 09:05 PM

well said Mario. People in New Orleans hate Emeril's cooking.

Posted by: KevinNYC on October 7, 2003 09:51 PM

"If you think of a Krugman win as a rebuke to Bush, then you just don't know much about the recent history of economic thought. I think it's funny that Krugman's biggest fans here on the comment thread tend to be so ignorant of his actual quality work, and worship his silly little columns."

Keith, please justify your inane, pointless, scatological comment. One of the things about the internet is that any idiot can stand up ans say any dumb thing and get away with it. And that's why you're here. And you do it with such pride, like a little kid who conquers the world with his toy soldiers.

Posted by: Zizka on October 7, 2003 10:26 PM

Who leaks Nobel picks? Posted by: EssJay on October 7, 2003 07:29 PM

Wrong question!

The right question is John Ashcroft going to investigate?

Posted by: snore on October 7, 2003 10:35 PM

I will e-mail Luskin and let him know the good (rumored) news!

Posted by: Frederick on October 7, 2003 11:01 PM

Was the public choice theorist, James Buchanan, a good pick in the mid 80's? This whole school of thought was demolished in the 60's. This prize is meaningless.

Posted by: skillet on October 7, 2003 11:28 PM

Public choice wasn't demolished in the 1960s; I happen to agree that it's a load of crap, but it's probably worth a Nobel Prize.

I would imagine that the paper "Directly Unproductive Prophet Seeking Behaviour", AER 1984, co-authored with Gene Grossman (look it up if you don't believe me) might count against Dixit, as it is in my professional opinion the most worthless academic paper ever published on any subject. Yes, including cultural studies.

Posted by: dsquared on October 8, 2003 12:07 AM

The NYT says it will be Phelps and Prescott

In the annual betting pool, with each entry costing $1, professors and graduate students in the nation's top economics programs have made Edmund S. Phelps of Columbia and Edward C. Prescott of the University of Minnesota the two favorites for the prize.

Posted by: MAtthew on October 8, 2003 12:26 AM

I would say public choice reemerged in the whole incentive compatibility stuff.

I also think Robert Aumann deserves a nobel. Considering his age better sooner than later.

Posted by: Michael Greinecker on October 8, 2003 02:38 AM

actuley the nobel price in economi was not made be alfred nobel bot by the swedhis bank...

Posted by: hans on October 8, 2003 03:43 AM

actuley the nobel price in economi was not made be alfred nobel bot by the swedhis bank...

Posted by: hans on October 8, 2003 03:44 AM

O frabjuous day, calhoo callhay
He chortled in his joy....

(Not sure about all the spellings.)

Posted by: richard on October 8, 2003 03:57 AM

O frabjuous day, calhoo callhay
He chortled in his joy....

(Not sure about all the spellings.)

Posted by: richard on October 8, 2003 03:57 AM

"Barry - the announcement is "Wednesday, October 8, 4:00 p.m. at the earliest" (Swedish time), a few hours after the far more interesting Chemistry announcement. See http://www.nobel.se for details."

Posted by: Tom Slee on October 7, 2003 04:45 PM


Thanks, Tom.

Posted by: Barry on October 8, 2003 04:00 AM

O frabjuous day, calloo, callay!
He chortled in his joy.

Posted by: richard on October 8, 2003 04:00 AM

O frabjuous day, calloo, callay!
He chortled in his joy.

Posted by: richard on October 8, 2003 04:01 AM

Yes, Auman would make an interesting choice as
would Thomas Schelling -- my favorite for the last
few years. I don't suspect either will win the
prize since the Swedes have already given a game
theory award. By the way, if the trade award is
given as speculated, would this make Krugman the
youngest Nobel (in Economics and in general)?

Posted by: malcolm on October 8, 2003 05:46 AM

Dean/Krugman 2004!

Not only would be it be very emotionally satisfying, but it would bring some other benefits:

1. The addition of Krugman could give Dean an excuse to abandon his anti-trade policies.

2. It would give the Democrats more credibility on the economy than anything else they might do could.

Krugman for Veep! You heard it here first.

Posted by: Mitch on October 8, 2003 06:06 AM

To put Mr. K's merits to one side for the moment, let's not forget (see comments by SW, Demosthenes) that the Nobel is given not for the overall quality of work in a career (or lifetime), but for a specific discovery. Hey, Linus Pauling got two of them, and he was worth them despite his oddness over Vitamin C. Also, let's not forget that when Einstein was given his for interpreting the photoelectric effect, someone (one of the judges?) said something like "we'll forgive him his relativity stuff".

Posted by: Tom Slee on October 8, 2003 06:29 AM

I think it's very possible for a very smart economist to be a terrible and biased analyst of politics. The Paul Krugman who sits down and precisely models a problem and develops new insights is a genius. The Paul Krugman who writes a thrice-weekly column for the New York Times ain't. Without the constraints placed upon them by modeling, a lot of economists lose their precision when they have to analyze things outside a modeling context, especially when their emotions overtake them.

Krugman is an excellent economist and WAS an excellent politcal-economic analyst. I liked Krugman back in the 90s, when he was hammering both parties for various stupidities. The 1998 piece that DeLong put up as Krugman's best (where Krugman nails Armey on some vital points) was quite good. Krugman was a clever commentator who brought good writing and solid analysis to bear on politics.

That all changed in 2000. Krugman lost it. I'll even discuss the column that tore it for me. First, Krugman criticzed Bush's tax cut. Fine and dandy. Good for Krugman. Then, Krugman praised Al Gore's Medicare prescription drug plan! What??!! Expanding an already massive entitlement when we face the retirement of the Boomers??!! And Krugman surely knows that actual costs of entitlements have always been WAY WAY WAY above projected costs (the 1965 projection of Medicare's cost in 1990, adjusting for inflation, was $12 billion. The actual cost of medicare in 1990 was $108 billion. Even if the 1965 inflation adjustment was too low, it's safe to say that nobody knew in 1965 that Medicare would be so hugely and terribly costly), so that Al Gore's "modest proposal" would almost surely turn into a money-sucking monster, dwarfing the fiscal stupidity of tax cuts (which are easier to repeal than entitlement expansion). At that moment, I knew that Krugman had sold out to the left. The integrity that had led him to be so honest with both right and left had been overtaken by a blinding hatred of Bush and desperation for acceptance in the Democratic Party apparatus. And Krugman's columns since have tended to confirm that sad hypothesis. Sigh.

Posted by: Keith on October 8, 2003 06:56 AM

Pauling did receive two Nobels, but they weren't in the same category. They were Chemistry (1954) and Peace (1963). Now maybe the science-related Nobels honor a specific accomplishment rather than a career, but the same is not true of the Literature Nobel.

Posted by: JB on October 8, 2003 06:58 AM

Pauling did receive two Nobels, but they weren't in the same category. They were Chemistry (1954) and Peace (1963). Now maybe the science-related Nobels honor a specific accomplishment rather than a career, but the same is not true of the Literature Nobel.

Posted by: JB on October 8, 2003 06:59 AM

Robert F. Engle Clive W. J. Granger

So much for the rumor . . .

Posted by: Bobby on October 8, 2003 07:22 AM

Rumor was wrong - winners are:

"for methods of analyzing economic time series with time-varying volatility (ARCH)"
Robert F. Engle
USA

"for methods of analyzing economic time series with common trends (cointegration)"
Clive W. J. Granger
United Kingdom


Posted by: Cecilia on October 8, 2003 07:23 AM

He'll get it but not this year it seems.

Posted by: GT on October 8, 2003 07:26 AM

It looks like Krugman would have to wait till next year:

http://www.nobel.se/economics/laureates/2003/press.html

Posted by: Tommaso Sciortino on October 8, 2003 07:35 AM

Alan,

I can't help you on links to their work on short notice (and nobody else has tried), but rest assured, there will be plenty of links available soon to the work of whoever is named. Now, a really high minded project for those of you with your own favorites for the prize is to cough up links that justify you nomination. That would create a source of ready education on some of the most notable recent thinking in economics...

...unless economics, like other intellectual pursuits, is an insiders game, among the rules to which is keeping outsiders satisfyingly ignorant.

Posted by: K Harris on October 8, 2003 07:35 AM

ARCH - that's so academic. Today you can estimate time varying volatility direct from high frequency data, or from hi-lo ranges, or you can read volatility expectations from options markets. IMO this is one example of where university conserves not only the good old ideas, but also some soon-to-be forgotten ones.

Posted by: Mats on October 8, 2003 08:05 AM

"The Paul Krugman who sits down and precisely models a problem and develops new insights is a genius. The Paul Krugman who writes a thrice-weekly column for the New York Times ain't."

Calvin called: he wants his duplicator back.

(He needs to turn it on its side and transmogrify Susie Derkins into a weasel.)

Posted by: Demosthenes on October 8, 2003 08:10 AM

Clive was just a matter of time. It's nice to see econometrics pulling down some more Nobels. Heckman won a few years back for his limited dependent variables-type work, so it makes sense to give out some time-series awards.

Krugman will get it, and he deserves it. So will Kevin Murphy. So will Andrei Schleifer. Basically, pencil in anybody who has won the JB Clark award and just wait.

Posted by: Keith on October 8, 2003 08:10 AM

Cointegration and the concept of Granger-causality are probably fine though - didn't say anything about those.

Posted by: Mats on October 8, 2003 08:15 AM

Demos,

Michael Jordan was brilliant on the basketball court but was a mediocre baseball player. This reasoning is hard for you to grasp?

Posted by: Keith on October 8, 2003 08:23 AM

Demos,

And Brad DeLong is a brilliant economist but may not be a very good poet. Whoa, it's blowin' my mind!

Posted by: Keith on October 8, 2003 08:26 AM

>[Keith] That all changed in 2000. Krugman lost it. I'll even discuss the column that tore it for me. First, Krugman criticzed Bush's tax cut. Fine and dandy. Good for Krugman. Then, Krugman praised Al Gore's Medicare prescription drug plan!

This the article that offended you so? With the following conclusion?

"So I'm not happy about Mr. Gore's tax proposals: instead of clearly distancing himself from Mr. Bush's truly bad idea, Mr. Gore has tried to meet it partway.

"Still, Mr. Bush's tax plan is grossly irresponsible, while Mr. Gore's is merely uninspiring. And this is a case where less than half a loaf is better than one. "

Is that your idea of "praise"?

Posted by: TK on October 8, 2003 08:39 AM

Those of us who know Paul Krugman primarily (or in my case entirely) as a political columnist, think he's a great columnist. He could be a thoroughly mediocre, garden-variety hack economist and I'd feel exactly the same about his newspaper work. There's no irony there; I've seen some of the stuff he does (but not the economics part) and I like what I've seen. To my mind a Nobel would have been a good thing to my mind simply because it would have given him added prestige, but no big deal.

Whenever Krugman's name is ever mentioned, a flood of kneejerks show up here to denounce his political writing. It has become impossible to believe that the reasons for this are non-psychiatric -- Keith here today is not unique, but in fact quite typical in his evident belief that Krugman's worthlessness is obvious to everyone, and that no reasons or arguments are required.

But yes, Keith, in answer to your specific question, it IS indeed quite possible for someone to be good at one job, and at the same time not good at some other job.

I wasn't a Krugman-worshipper until I encountered some of his enemies here. Come to think of it, I was lukewarm about Clinton before the phony impeachment hysteria. Some people you support because of their enemies.

Posted by: Zizka on October 8, 2003 08:53 AM

Sorry, the link didn't take for some reason. The Krugman article is here: http://www.pkarchive.org/column/91000.html

Posted by: TK on October 8, 2003 08:56 AM

"Keith here today is not unique, but in fact quite typical in his evident belief that Krugman's worthlessness is obvious to everyone, and that no reasons or arguments are required."

Ziska, there's a big new fad called reading. From an earlier post:

"Then, Krugman praised Al Gore's Medicare prescription drug plan! What??!! Expanding an already massive entitlement when we face the retirement of the Boomers??!! And Krugman surely knows that actual costs of entitlements have always been WAY WAY WAY above projected costs (the 1965 projection of Medicare's cost in 1990, adjusting for inflation, was $12 billion. The actual cost of medicare in 1990 was $108 billion. Even if the 1965 inflation adjustment was too low, it's safe to say that nobody knew in 1965 that Medicare would be so hugely and terribly costly), so that Al Gore's "modest proposal" would almost surely turn into a money-sucking monster, dwarfing the fiscal stupidity of tax cuts (which are easier to repeal than entitlement expansion)."

There's a clear reason and a clear argument. QED.

But I do think you have a point that my scatalogical comment wasn't necessary. I shouldn't have said that. Now I wonder if you're enough of an adult to explicitly realize that your last comment was inaccurate and wrong, and to clearly say so here in this forum.

Posted by: keith on October 8, 2003 08:59 AM

Well, so much for *that* rumor. :-)

> STOCKHOLM (Reuters) - U.S. economist Robert Engle
> and Britain's Clive Granger have won the 2003 Nobel
> economics prize for inventing models used to evaluate
> investment risk and study the relations between
> simultaneous economic phenomena.

So now we can continue to the ongoing Krugman bash/love fest, already in progress.

Posted by: Jonathan King on October 8, 2003 09:09 AM

We have a radical right Administration and Congressional leadership intent on vastly changing American social, economic, and foreign policies. Paul Krugman and Al Franken have shown that the intents and policies of our leaders can be written about no matter the attempts at intimidation. In a time of Faux News, Paul Krugman and Al Franken and cohorts have done us an enormous service.

Remember the WMDs, remember Valerie Plame, remember the surplus, remember environmental protections, remember our jobs...?

Posted by: lise on October 8, 2003 09:12 AM

Don't know. I don't really think ARCH is so important.

Posted by: Andrew Boucher on October 8, 2003 09:14 AM

The act of mentioning Al Franken and Paul Krugman in the same breath is ridiculous. Paul Krugman, whatever you think of his economic or political ideas, is a trained social scientist, whereas Al Franken is an entertainer.

The litany posted by lise of the allegaed sins of the current administration is similar to what an Al Franken or Michael Moore would write, which is entirely different from the academic writings of a competent economist. Don't be too sure that Krugman would want to be placed in the same company.

Posted by: EcoDude on October 8, 2003 09:27 AM

Where's the outrage? It's a radical-right conspiracy to transfer wealth from Princeton to LaJolla! It's a revolution! It's the next Argentina!

(wiping froth away from mouth)

Posted by: R-three on October 8, 2003 09:35 AM

Well, Keith, how much of Brad DeLong's poetry have *you* read?

I suspect little more than I've read of Krugman, Bhagwati, Granger, or Engle's economics work. How the heck can you know how good a poet he is?

Typical slime and defend, slime and defend...

Posted by: Julian Elson on October 8, 2003 10:15 AM

>> And Krugman surely knows that actual costs of entitlements have always been WAY WAY WAY above projected costs .... it's safe to say that nobody knew in 1965 that Medicare would be so hugely and terribly costly ... << Keith

Not really. I have no idea how much Krugman knows about health care markets, but it was widely known in 1965 that an inflationary bias was built into the structure of the Medicare program. Congress recognized this in 1966 when it passed comprehensive health planning legislation. Rather than a "liberal" giveaway, the structure that led to the inflationary bias reflected efforts to accommodate the program to the structure of the private health care market. So, in this sense, cost inflation could be attributed to an effort to accommodate “conservative” principles in the program’s design.

Posted by: RM on October 8, 2003 10:20 AM

Exactly, EcoDude.

If one looks at the fans of Krugman's political columns, they breathlessly mention his "genius" in the same breath as luminaries like Al freakin' Franken. Christ. It's like hearing "Milton Friedman" and "Anne Coulter" in the same sentence. Not easy to bear on a full stomach. It's like loving Mother Teresa because she got mad once and kicked a kitten.

Posted by: Keith on October 8, 2003 10:23 AM

>>ARCH - that's so academic. Today you can estimate time varying volatility direct from high frequency data, or from hi-lo ranges, or you can read volatility expectations from options markets. IMO this is one example of where university conserves not only the good old ideas, but also some soon-to-be forgotten ones.

But what happens if you're interested in things other than securities prices, Oh Omniscient One? Some of us are, you know. Fool.

Posted by: dsquared on October 8, 2003 10:26 AM

The significance of Cointegration (from Economy.com):

"If two series are nonstationary (statistically a random walk), then a regression involving the two yields nonsense parameters with biased test statistics. Intuitively, if GDP is a random walk, then it makes no sense to attempt to explain it in a regression involving 'fundamental drivers' such as interest rates or consumption."

"The issue quickly faded from the literature, however, and was then revived by Charles Nelson and Charles Plosser in a 1982 Journal of Monetary Economics article entitled 'Trends and Random Walks in Macroeconomic Time Series: Some Evidence and Implications.' The Nelson and Plosser article, which was the most-cited academic paper of the 1980s, argued that most macroeconomic time series were indeed random walks, and that the vast majority of econometric research was, therefore, trash. New methods were needed to relate the statistical behavior of one time series to another, especially if researchers and economic forecasters were to have any methodological basis for their profession."

"Granger and Engle then introduced the concept of Cointegration, which is a statistical attribute of two or more data series. Even if two time series, such as GDP growth and interest rates, are statistical random walks, they may share a common trend (they may be cointegrated). The presence of cointegration, or a common trend, can then be exploited to yield economic insights and for the purposes of analysis and forecasting. The work of Engle and Granger was then extended by Soren Johansen to include the presence of a common trend among many variables."

http://www.economy.com/dismal/pro/blog_main.asp#50AD34A0-A820-4B40-85A8-39A65AAC6D60

Posted by: Kosh on October 8, 2003 10:46 AM

I'm under the impression that in the real world, there are a non-trivial number of people out there using GARCH to account for stochastic volatility since the standard Black-Scholes assumes constant vol.

Posted by: ETC on October 8, 2003 10:50 AM

"A bad economist getting the Nobel Prize would wash away the bad taste of a bad politician becoming governor of California? I suppose it makes sense in a way, sorta like how if you bang your elbow against something hard, then rub it, it makes the pain hurt less because your nervous system is overwhelmed by all the signals."

Since when is Krugman a bad economist?

Posted by: Brian on October 8, 2003 11:06 AM

>>But what happens if you're interested in things other than securities prices>> - well, set up a betting exchange or futures market so it becomes a security price ;)

No, my impression is that the original success of ARCH-type specs. was built in modelling stock index return series. This was further underlined by the feedback from (conditional) volatility and expected return that was estimated to have the right sign (positive risk premia) in "ARCH-in-mean" specs. I'm sure this was good work and a long awaited step forward, but with more high frequency and detailed price records, and more liquid options market, I do think ARCH is becoming obsolete (in this respect) and is hampering more forward looking efforts.

However, ARCH might still be an important time series specification in other settings. Sunshine.

Posted by: Mats on October 8, 2003 11:16 AM

Engle's work is also fundamental to a lot of other areas apart from "Generalized auto-regressive conditional heteroskadestic" processes {quite a mouthful isn't it? }
His work in intraday duration based models has been used in a lot of market microstructure related work. Same things can be said about Granger Causality.

Its interesting to see that Engle's GARCH gets a lot of Publicity. Simply because it seems to help come up with the most important fudge number in a black scholes world.
GARCH Volatility as someone said "is the wrong volatility we use in the wrong model in order to get the right price"

Posted by: Kris on October 8, 2003 11:31 AM

Engle's work is also fundamental to a lot of other areas apart from "Generalized auto-regressive conditional heteroskadestic" processes {quite a mouthful isn't it? }
His work in intraday duration based models has been used in a lot of market microstructure related work. Same things can be said about Granger Causality.

Its interesting to see that Engle's GARCH gets a lot of Publicity. Simply because it seems to help come up with the most important fudge number in a black scholes world.
GARCH Volatility as someone said "is the wrong volatility we use in the wrong model in order to get the right price"

Posted by: Kris on October 8, 2003 11:32 AM

Engle's work is also fundamental to a lot of other areas apart from "Generalized auto-regressive conditional heteroskadestic" processes {quite a mouthful isn't it? }
His work in intraday duration based models has been used in a lot of market microstructure related work. Same things can be said about Granger Causality.

Its interesting to see that Engle's GARCH gets a lot of Publicity. Simply because it seems to help come up with the most important fudge number in a black scholes world.
GARCH Volatility as someone said "is the wrong volatility we use in the wrong model in order to get the right price"

Posted by: Kris on October 8, 2003 11:32 AM

Engle's work is also fundamental to a lot of other areas apart from "Generalized auto-regressive conditional heteroskadestic" processes {quite a mouthful isn't it? }
His work in intraday duration based models has been used in a lot of market microstructure related work. Same things can be said about Granger Causality.

Its interesting to see that Engle's GARCH gets a lot of Publicity. Simply because it seems to help come up with the most important fudge number in a black scholes world.
GARCH Volatility as someone said "is the wrong volatility we use in the wrong model in order to get the right price"

Posted by: Kris on October 8, 2003 11:36 AM

I think I'm both disapointed and relieved that Krugman didn't get the Nobel this year. I'm not fluent enough with his academic work to have an competent opinion myself, but it's my understanding that he is, at least, not a shoe-in for a Nobel. So I wondered if awarding him the Nobel here and now might not have been (had it happened) a teensy-bit politically motivated. That would have been fun, but it's probably best if all that doesn't come into it.

Posted by: Keith M Ellis on October 8, 2003 11:39 AM


Does anyone feel that David Hendry should be up there with Engle and Granger. He's done as much as those two on studying long-run data. If I was him I might be a little pissed off today.

Posted by: Mike on October 8, 2003 11:42 AM

>>the standard Black-Scholes assumes constant vol.
Yes it does, but still works fine for time-varying vola. if you just know its average over the relevant period. If it's uncorrelated with the return that is, which it isn't of course. However, "ARCH-in-mean" specs. looks into this correlation, but does not do a good enough job in modelling vola-return dependencies. Furthermore it is complicated to make an analytical ARCH/GARCH-based options formulae anyway.

I would plainly suggest that if you use ARCH/GARCH to price illiquid options in finance industry, you don't get much bang for the buck and are probably victim of this ARCH-hype.

Even worse: ARCH is very difficult to generalize from its original 1-d formulation into multi-variat settings. Estimation of ARCH parameters is an ill-conditioned numerical problem which requires special optimizers (like BHHH). However, these complication makes the model look advanced, hyping it even further. Huh.

Posted by: Mats on October 8, 2003 11:44 AM

"Furthermore it is complicated to make an analytical ARCH/GARCH-based options formulae anyway."

What do you mean? You can use the BS formula and use a garch type volatility. Or even use the heston type closed form formula both of which can be implemented by a junior high student.


"Even worse: ARCH is very difficult to generalize from its original 1-d formulation into multi-variat settings. Estimation of ARCH parameters is an ill-conditioned numerical problem which requires special optimizers (like BHHH). However, these complication makes the model look advanced, hyping it even further. "

So what are you suggesting we go back and use slide rules?

Posted by: matmat on October 8, 2003 12:15 PM

>>So what are you suggesting we go back and use slide rules?>>

I am suggesting that you should look at the problem of modelling time varying volatility in the way you think would yield the most precise and parsimonius results. I am further suggesting that you should stop overemphasizing and tweaking an old model just because its inventors and proponents has hyped up its status inside the academic world.

Posted by: Mats on October 8, 2003 12:31 PM

I agree with Mats. ARCH strikes me as a lot of hype. Maybe there are hedgefunds out there making lots of money using ARCH-models and high-frequency data, but then maybe there aren't. I work in one of the biggest Equity Derivative departments in the world, and while we use models where volatility is not constant, ARCH isn't one of them.

Posted by: abo on October 8, 2003 12:37 PM

Agreed with Mats. ARCH does not strike me as all that meaningful.

Posted by: anne on October 8, 2003 12:49 PM

Keith not-Ellis:

You proved in your second post that you disagreed with the political Krugman, and did give reasons why. You did not in any way prove that your dismissive comments in the two posts were justified, however. But perhaps my second slam was more applicable to your first post, which had no content whatsoever.

Several here seem to have the idea that Science (economics)is pure and wonderful, but that Politics is base and unworthy. Not really an appropriate point of view for citizens of a democracy, though it was widespread in Germany and Austria-Hungary before the First World War.

I think that, in stepping into a political role, Krugman has performed an important and desperately-needed service.

To me the kneejerk slamming of Krugman based on the anti-democratic Pure Science argument is about as bad as the knee-jerk right-wing BS. Especially because economics has always been politically implicated, and many other economists have got their hands dirty in the real world in less admirable ways.

Even when Krugman does win his Nobel in Economics, he will just become one more name on a considerable list (a list which, incidentally, now has two felons on it, one of them being Robert Merton).

Posted by: Zizka on October 8, 2003 12:53 PM

Keith not-Ellis:

You proved in your second post that you disagreed with the political Krugman, and did give reasons why. You did not in any way prove that your dismissive comments in the two posts were justified, however. But perhaps my second slam was more applicable to your first post, which had no content whatsoever.

Several here seem to have the idea that Science (economics)is pure and wonderful, but that Politics is base and unworthy. Not really an appropriate point of view for citizens of a democracy, though it was widespread in Germany and Austria-Hungary before the First World War.

I think that, in stepping into a political role, Krugman has performed an important and desperately-needed service.

To me the kneejerk slamming of Krugman based on the anti-democratic Pure Science argument is about as bad as the knee-jerk right-wing BS. Especially because economics has always been politically implicated, and many other economists have got their hands dirty in the real world in less admirable ways.

Even when Krugman does win his Nobel in Economics, he will just become one more name on a considerable list (a list which, incidentally, now has two felons on it, one of them being Robert Merton).

Posted by: Zizka on October 8, 2003 12:58 PM

Disregard my earlier statement.

Posted by: Brian on October 8, 2003 01:10 PM

Ziska:

"Several here seem to have the idea that Science (economics)is pure and wonderful, but that Politics is base and unworthy. Not really an appropriate point of view for citizens of a democracy, though it was widespread in Germany and Austria-Hungary before the First World War."

Science is about an honest quest for knowledge, politics is about rhetoric and manipulating the ill-informed by the use of propaganda. It is not well-informed citizens that determine the outcome of an election, it is the masses who get their information from television, newspapers and talk radio. The right wing understands this all too well. Control the mass media and you control the political process. Sad, but true. Repeat over and over again "Liberals are un-American traitors and are guilty of treason" "all Liberals want to do is tax and spend".

Hitler would be proud of these folks.

Posted by: Kosh on October 8, 2003 01:13 PM

Keith: I saw Paul Krugman speak last evening at the Los Angeles County Library. He is first and foremost an academic. You can tell that the more he speaks in public the better he becomes with delivering his scripted comments. But he is no Clinton, who could prattle on for an hour without looking at his notes. He is not freightening. He's a little guy. He is speaking his mind. But, in fairness to him, he is an academic at heart and is willing to listen to, and consider, evidence you present contrary to his position. Further, like many professional people who are smart, and know they are smart, he has an ego, so do not challenge his without some facts and coherent analysis -- not ideological ravings. Which reminds me of a sports anecdote that I thinks applies here. Kareen Abdul-Jabbar in his autobiography had a passage about comraderie on court amongst competing players. He tells about how Calvin Murphy (who was 5'7" and the shortest player in the NBA) would come down the court and complain to Kareen about how his team's management was always trying to get rid of him because he was so short and therefore considered a detriment on defense. So he would plead with Kareem to let him shoot jump shots from various distances without much resistence from Kareem. Each time down the court getting closer and closer to the basket when he would take a shot. Then, one time, Murphy came in for a lay up and Kareem blocked the shot into the stands. "Kareem!" Murphy screamed, "what are you doing to me? They're trying to take my job! Take food off my table!" To which Kareem responded, "Calvin, you can have the jumper from 25", you can have the jumper from 15" but don't ever bring the weak stuff inside." That is what I saw of Krugman last night. Polite. But don't bring the weak stuff. He sees what he considers to be weak economic policy being sold to an unsuspecting public by people who, in most instances -- and especially for the real economists in the Bush Administration -- know better but who have an ideological agenda to put in play -- without regard to sound public policy. Further, they dress it up as sound public policy but, from Krugman's vantage point, it is not. So, the political commentary goes hand in hand with the economic commentary. He only has 700 words a day to make a point, so you have to read these things in their totality, which is what he is trying to do in his book. My bet is that over time he becomes a better political commentator as he does it more and more. Finally, my prediction, Krugman should get ready for a back lash from the left (which plays into the hands of his detractors in the Bush Administration). He will not back down from his support of free trade. They will tar and feather him, but he won't budge. However, it is very hard to explain complex economic theory in 700 works twice a week. He will lose some luster with people on the left of this because he won't pander to them on trade issues as protectionism and China-India bashing becomes a greater public issue if the recovery does not produce jobs.

Posted by: Cal on October 8, 2003 01:14 PM

EcoDude sez:

"The act of mentioning Al Franken and Paul Krugman in the same breath is ridiculous. Paul Krugman, whatever you think of his economic or political ideas, is a trained social scientist, whereas Al Franken is an entertainer."

Maybe so, but an entertainer like Al Franken is far more likely to have an impact on the political process than Mr. Krugman.

Posted by: Kosh on October 8, 2003 01:44 PM

"Agreed with Mats. ARCH does not strike me as all that meaningful."

I don't see the rational in making blanket statements such as the above markets are not moving on your voting yes or no for a model. Believe it or not BlackScholes world has come to rule has for some time to come there is nothing in the horizon that is going to displace this formula when it comes to pricing contingent claims.

Regarding the other gentleman who talked about worlds biggest equity department just because you didn't see that model doesn't mean its of no use.

If any of you could let us know what is really so bad about GARCH/ARCH then I promise to buy you lunch.

Posted by: matmats on October 8, 2003 01:51 PM

>>just because ... [the other gentleman] [on the] worlds biggest equity department ...didn't see that model doesn't mean its of no use.>>

No, of course, but that's exactly my point. Because Engle has hyped a half-good model for over 20 years, and made a lot of other people follow him, doesn't mean its good enough for a Nobel Price. It is probably just obsolete, and to me it seems that the ARCH-hype has crowded out resources from more fruitful approaches. I tried to compile some links to research papers on this at:
http://blogofpandora.blogspot.com/2003_10_01_blogofpandora_archive.html#106564514994325784

Posted by: Mats on October 8, 2003 02:33 PM

What a bunch of stuffy nerds. Socialist nerds to boot!

Any economist who actually believes in a socialistic economy should be targeted as a quack.

Socialism does not create wealth, it consumes it.

Freerepublic.com

Posted by: myself6 on October 8, 2003 03:02 PM

Myself sez:

"Any economist who actually believes in a socialistic economy should be targeted as a quack."

So who have you noticed advocating a socialistic economy? Or is there no economy in the world today that you don't consider "socialistic"? Just asking for a little definition of terms, but don't strain yourself.

Posted by: Kosh on October 8, 2003 03:30 PM

Too young, too respectful, too righteous, too hot, too good!

Posted by: Gabriele on October 8, 2003 03:31 PM

To myself6:

Even having X*10EX*10 on my current account, I would rather prefer to spend the "money" in necessary things like: For some children in Afrika (U know, now very bad there => Somalia, Ethiopia, World very big, not only USA = eradicating terrorism) , because I know that I am only a "spirit in a material world"

Posted by: Gabriele on October 8, 2003 04:00 PM

Zizka, "To me the kneejerk slamming of Krugman based on the anti-democratic Pure Science argument is about as bad as the knee-jerk right-wing BS. Especially because economics has always been politically implicated, and many other economists have got their hands dirty in the real world in less admirable ways."

True. But you (and others) have seemed to inevitably assume that a) the "pure science" argument is necessarily "knee-jerk" and that, b) it's "undemocratic".

When you wrote in this comment thread earlier that you came to know Krugman only from his political writings, that was revealing to me.

I came to Krugman years ago through his popular economics writings, as I've mentioned, I have all but his most recent book. And I've only sporadically read his NYT column. I'm a social liberal who is pro-globalization and who despises the Bush administration. If there's a disgreement between Krugman and myself on any substantive matter, I don't know what it is.

Yet I have here and elsewhere criticized Krugman's NYT focus and style. Well, in your worldview, such a criticism is prima facie "knee-jerk" and "anti-democratic". That doesn't gibe with my self-professed politics and preferences, and so the end result is you write that "you long ago stopped trying to make sense" of me.

Why? I mean, it's *particularly* interesting that you assume that anyone that is critical of Krugman in this way is necessarily a "knee-jerk" reactionary. That's interesting because, in economics in general and in Krugman's case in particular, given the politicization of trade and globalization, the same sort of faulty reasoning has been repeatedly applied to him. I find it interesting that he's the toast of the left today when, five years ago, I was watching them excorciate them for his criticism of the antiglobalization movement. I mean, clearly an apologist for free trade and globalization is a lacky for the corporate right...right?

My criticism of Krugman's style and focus of his NYT column comes down to two things. The first is strategic: I think he's less of an asset to the left in this capacity than he would be as a less apparently partisan "Professor Krugman". You and I disagree on this, of course. But consider just for a moment that I am making that argument in complete good faith. The second reason is more ambiguous. It has to do with my association with the scientific culture and hearing my astronomer friends ridicule Carl Sagan. It also has to do with my related deep distaste for ideology and partisan politics. It has to do with the fact that every single damn day of this abhorrent Presidential administration, as my hatred of them inexorably grows, I fight against an internalized anger that becomes an irrational bias. I watch every thoughtful person I know on the left struggle with this, and it makes me sad to see any of them losing ground in it. People like you cheer that on, because you're so damn sure that our side is right and you *need* an angry partisan here and there to yell into the echo chamber. But I don't need the vox populi as reassurance that I'm right; and I think it's sad to see Krugman reduced to being a pundit. He has better things to do. You and yours believe that he's making a real difference as a columnist. That's a load of bull. Nothing would be different in the current political landscape were Krugman to have never written a NYT column. But things would be different in the economic landscape--both academic and practical--had he not done his professional work.

Posted by: Keith M Ellis on October 8, 2003 04:09 PM

Anyone know how many Nobel winners in economics have been under age 60 at the time they won?

Posted by: Patrick R. Sullivan on October 8, 2003 04:20 PM

Great point, Patrick. If a layman ever wants to be reasonably close to knowing what is "cutting edge" in economics, they should look to the JB Clark awards more than the Nobel Prize.

Unlike the "hard" sciences, the Nobel Prize in Economics is a little closer to a lifetime achievement award, although specific contributions do get cited. Usually, the work that wins somebody the Nobel prize is a bit dated. The JB Clark award, given to the best economist under 40, usually goes the economists who, at the very least, are not too far beyond their most productive research years.

To Zizka: I have no problem with great economists getting into the political fray, but I have a problem when great economists devalue their legacy by doing a poor job of things in the fray. Milton Friedman and Paul Samuelson represent two men who were great economists and entered the political fray, raising the level of discourse of the body politic. Krugman has tarnished their fine legacy as of late, losing his precision and sense of fairness to his emotional and not entirely rational distaste for the current President (not that all of Krugman's objections are wrong).

Cal: Maybe you're right, and Krugman will display the past courage that he has shown in also standing up to lefty shibboleths. Let me give credit where such credit is due to Brad DeLong, who took a very real and personal risk by being willing to criticize both Castro (not too risky) and Hilary Clinton (hugely risky).

But Cal, I've seen no sign of Krugman criticizing Dean's idiotic statements on trade. Maybe this is pure ignorance on my part, and you'll show me otherwise. But I believe (and I can be proven wrong) that Krugman has personally crossed some awful Rubicon, where he has lost his capacity to criticize Democrats because he has reached some personal conclusion that those evil Republicans must be stopped at any and all costs. If Krugman shows me up on this belief, then I would be overjoyed.

Posted by: Keith on October 8, 2003 05:45 PM

Keith Ellis, insofar as I understand what you're saying you're exactly the kind of antidemocratic purist I'm talking about.

"It also has to do with my related deep distaste for ideology and partisan politics." Politics is how democracies make their decisions. The problem with US politics is too few Krugmans, not too many. The only nations without some sort of partisan politics are one-party states and places like Saudi Arabia.

"Nothing would be different in the current political landscape were Krugman to have never written a NYT column". That's just not true, and the defeatism and fatalism is appalling. I personally am not optimistic, but I feel that everyone has a duty to try.

Posted by: Zizka on October 8, 2003 05:50 PM

I didn't say I had a problem with politics. I said I had a problem with *partisan* politics. And my point about Krugman's influence on the political landscape is not "defeatist" or "fatalistic". It's "realistic". If you think that any single pundit has a significant effect on American politics you're delusional. A real journalist can have an effect by creating a story where none existed before. A politician has a practical effect and a few of them have the bully pulpit. Even a few novelists or screewriters/directors can have an effect by presenting a strong point of view in an engaging way to large segment of the population. But even the pundits with the largest audiences, either television or radio, are speaking only to a small minority of the population, and then usually they are only telling those folks what they already believe. If you look at the actual numbers of people that read "Treason" or "Liars..." or the like, you'll find that huge political bestsellers such as these still only reach a tiny portion of the population. And all of these have bigger audiences than does Paul Krugman, Robert Novak, Michael Kinsley, Maureen Dowd, Buckley, or any of the others. Being a member of the political commentariat is, at best, like being a writer for the beltway version of "People Magazine". You have no influence with the people who make policy, almost none of the voters who matter read newspapers (much less political columnists), and for the most part you're stuck with the talking points determined by the political strategists who really shape the national discourse. It's about as important to the game as being an NFL cheeleader, without a similar presumption of chastity.

There are numerous ways to make a substantial difference in American politics. Being a pundit isn't one of them. Being a pundit is being a part of the fucking show they are putting on to entertain us to distraction. Punditry is not part of the solution, it's part of the problem.

Posted by: Keith M Ellis on October 8, 2003 06:25 PM

matmats: Like I said, maybe it is of use to a hedgefund somewhere trading on high-frequency data. But our general experience is that the more complicated you make the model, the more difficult it is for the trader to follow his position. For market making and flow business, simplicity counts. For volatility arbitrage the model counts more of course, and maybe others have made ARCH work as a predictor which can make money. On the other hand, if they have, they aren't likely to speak too loudly about it, so you could well be right.

Posted by: abo on October 9, 2003 12:15 AM

Keith, since about 1800 American politics has been partisan politics. George Washington didn't like it either, but he ended up being a founder of the Federalist Party.

Your assertion that no pundit has any influence needs support. I don't know where to begin in criticizing it. Without Krugman, the average, well-informed American would think that Bush's economic plan is a good one. As it is, many Americans do anyway, but that's because the proeponderance of pundits are letting Bush get away with it. If there's a pundit problem, having one less good pundit would make the situation worse rather than better. Obviously. (It's my assumption that for a host of reasons, pundit Krugman is better than almost all the others).

Posted by: Zizka on October 9, 2003 03:45 AM

Can't let this one pass without noting that, of course, to a grammarian with a grounding in probability theory, the phrase "time-varying volatility" is a repugnance.

Posted by: dsquared on October 9, 2003 04:25 AM

"ow, a really high minded project for those of you with your own favorites for the prize is to cough up links that justify you nomination. That would create a source of ready education on some of the most notable recent thinking in economics..."

Here´s a essay by Sergio Hart about the work of Robert Aumann:

http://www.ma.huji.ac.il/~hart/get/introduction.html

Posted by: Michael Greinecker on October 9, 2003 11:51 AM
Post a comment