Saturday, March 29, 2008

Charlie Munger, Ricardo and finance

Here is the text of an interesting talk given by Charlie Munger entitled Academic Economics: Strengths and Faults After Considering Interdisciplinary Needs. Among other things, he discusses physics envy, psychology, behavioral economics and efficient markets. Munger, a billionaire, was Warren Buffet's long time investing partner. He studied briefly at Caltech before attending Harvard Law School, and now lives in Pasadena.

“Charlie is truly the broadest thinker I have ever encountered. From business principles to economic principles to the design of student dormitories to the design of a catamaran he has no equal..." –William H. (Bill) Gates III Microsoft Corporation

...Another example of not thinking through the consequences of the consequences is the standard reaction in economics to Ricardo’s law of comparative advantage giving benefit on both sides of trade. Ricardo came up with a wonderful, non-obvious explanation that was so powerful that people were charmed with it, and they still are, because it’s a very useful idea. Everybody in economics understands that comparative advantage is a big deal, when one considers first order advantages in trade from the Ricardo effect. But suppose you’ve got a very talented ethnic group, like the Chinese, and they’re very poor and backward, and you’re an advanced nation, and you create free trade with China, and it goes on for a long time.

Now let’s follow and second and third order consequences: You are more prosperous than you would have been if you hadn’t traded with China in terms of average well-being in the United States, right? Ricardo proved it. But which nation is going to be growing faster in economic terms? It’s obviously China. They’re absorbing all the modern technology of the world through this great facilitator in free trade, and, like the Asian Tigers have proved, they will get ahead fast. Look at Hong Kong. Look at Taiwan. Look at early Japan. So, you start in a place where you’ve got a weak nation of backward peasants, a billion and a quarter of them, and in the end they’re going to be a much bigger, stronger nation than you are, maybe even having more and better atomic bombs. Well, Ricardo did not prove that that’s a wonderful outcome for the former leading nation. He didn’t try to determine second order and higher order effects.

If you try and talk like this to an economics professor, and I’ve done this three times, they shrink in horror and offense because they don’t like this kind of talk. It really gums up this nice discipline of theirs, which is so much simpler when you ignore second and third order consequences.

The best answer I ever got on that subject – in three tries – was from George Schultz. [Schultz was an MIT economics professor before becoming Secretary of the Treasury and Secretary of State.] He said, “Charlie, the way I figure it is if we stop trading with China, the other advanced nations will do it anyway, and we wouldn’t stop the ascent of China compared to us, and we’d lose the Ricardo- diagnosed advantages of trade.” Which is obviously correct. And I said, “Well George, you’ve just invented a new form of the tragedy of the commons. You’re locked in this system and you can’t fix it. You’re going to go to a tragic hell in a handbasket, if going to hell involves being once the great leader of the world and finally going to the shallows in terms of leadership.” And he said, “Charlie, I do not want to think about this.” I think he’s wise. He’s even older than I am, and maybe I should learn from him.

My favorite Munger quote:

I regard the amount of brainpower going into money management as a national scandal.

We have armies of people with advanced degrees in physics and math in various hedge funds and private-equity funds trying to outsmart the market. A lot of you older people in the room can remember when none of these people existed. There used to be very few people in the business, [and they were not] who were not very intelligent. This was a great help to me.

Now we have armies of very talented people working with great diligence to be the best they can be. I think this is good for the people in it because if you know enough about money management to be good at it, you will know a lot about life. That part is good.

But it's been carried to an extreme. I see prospectuses for businesses with 40-50 people with PhDs, and they have back tested systems and formulas and they want to raise $100 billion. [Reference to Jim Simons of Renaissance Technologies.] And they will take a very substantial override for providing this wonderful system. The guy who runs it has a wonderful investment record and his system is a lot of high mathematics and algorithms with data from the past." [...]

"At Samsung, their engineers meet at 11pm. Our meetings of engineers (meaning our smartest citizens) are also at 11pm, but they're working on pricing derivatives. I think it's crazy to have incentives that drive your most intelligent people into a very sophisticated gaming system."


Drawbacks said...

I remember having a similar thought when I read a while back that Bill Gates had expected to be the best mathematician in his classes at Harvard, but finally had to admit that he wasn't quite as good as his room-mate Andy Braiterman. Mr Braiterman is now a tax lawyer.

steve said...

Gates was overconfident and learned his lesson pretty fast. Does the math world really miss Braiterman? He was only magna, not even a summa ;-)

David Kane said...

Munger is a genius and I hesitate to disagree, but the big benefit of having so many smart people in finance is that it makes the allocation of capital much better than it otherwise would have been. Now, the magnitude of this effect is tough to estimate, but I bet that it is higher than Munger thinks.

Roger Bigod said...

It's one of the larger misallocations of capital in history. The social value added is in the investment decisions that involve taking a broad view, at which Buffett and Munger are clearly masters. A narrow focus on little greek letters in superscripts and subscripts may not be so productive.

tc said...

Did quants contribute to the crisis? Nicole El Karoui says blame the financiers. On the other hand, a commenter at Wilmott's points out that quants made securitization possible (which brought large benefits as well as drawbacks, of course).

Roger Bigod said...

It looks like the immediate cause of the meltdown is applying historical data based on prudent lending practices to mortages involving a lot of fraud. Quants weren't responsible for the misapplication of the models. And they weren't in a position to monitor the quality of the loans. Nor could they predict the enormous leverage that would be applied to create trillions in cross-loans on the derivatives.

OTOH, the intellectual prestige of theoretical physics probably contributed to the whole process. The justified confidence people have in successful theories led them to push too far.

Drawbacks said...

Braiterman maybe couldn't have done any really good pure math, but I'm guessing he could have applied his talents to some area of engineering or business that looks a lot less zero-sum than the world of tax law.
Of course I may be doing him a disservice, and also, there's plenty of people who'll tell you that Bill Gates' contribution to the software world has been a net negative.

STS said...


If the talk was in 1995, why is it dated 2003? And why does it reference a book published in 1997?

Dave Kane:

"...having so many smart people in finance ... makes the allocation of capital much better..."

I find that very hard to believe. Munger's type of intelligence (heuristic, syncretic, skeptical) is good for allocation. I suspect most of the high-brow academic approaches to trading contribute a lot more to liquidity. Valuable to be sure, but quite a different thing.

steve said...

STS: Thanks, I don't know why I thought it was given in 1995. I must have misread something somewhere -- I've fixed it in the post.

Roger Bigod said...

The idea of repackaging morgage securities into tranches was originated at Salomon and is covered in Liar's Poker, by Michael Lewis. Quants may have contributed to applying historical data, but relying on historical data was diastrous with so much of the later mortages being based on fraud.

The Black-Scholes formula created the options market. People had been pricing options with intuition and guesswork, and human intuition isn't good at this. It can be derived from a diffusion equation, but the developers' background was in economics.

Blog Archive