Replies

I'm reading a book that seems related: The Wisdom of Crowds, by James Surowiecki

The premise is that a group will always predict outcomes better than any single expert. He says stock analysts do not perform much better than random choices, that groups of people from diverse backgrounds are superior to any single expert in making corporate decisions, and that beginners' luck has a significant role in problem solving. (I haven't finished the book so there's probably more.)

He takes a stab at why "experts" often make bad decisions; it's because they tend to follow accepted practices within their profession. A conventional failure, he states, is accepted more easily than an unconventional success. He also addresses Trends and fads like the information boom and bust.

I started reading this book half expecting that it would be a socialist screed playing to leftist populist conceits. Surprisingly, so far its anything but that; it affirms democratic institutions as more efficient and accurate, given a minimum level of fluency. The last point is important. Taken together this affirms conservative principle.

He says stock analysts do not perform much better than random choices, that groups of people from diverse backgrounds are superior to any single expert in making corporate decisions, and that beginners' luck has a significant role in problem solving.

You've got to be careful about that, too: how does one select among crowds to see which are successful, and which are not? If you pick only the "successful crowds," then the comparison is a priori biased, and likely invalid.

Friedrich Hayek predicted the fall of Communism based on the premise that a few people couldn't even access, much less properly process, the amount of information inherent in an economy.

His big insight was that the free market works better, because it allows huge numbers of people to make small decisions on things they know about, and they don't have to worry about everything else. The net effect over millions of such people and decisions is that the total information in the economy is processed more optimally. That doesn't mean that lots of people don't make bad decisions (they do). It's just that more people are able to hold enough information in their heads to make small decisions, than they are for large decisions.

Smarter people, though, can make larger decisions more accurately, which is probably why there's a such a strong correlation between IQ and income.

However, individuals are less likely to make good decisions over a broad range of fields, because they don't have enough information to do so. That's where the run-of-the-mill stock analyst resides -- he doesn't really know about the business behind a stock, he just knows the transaction side of the business, and perhaps some of the short-term business aspects, but as a generalist across many, many businesses he doesn't have much more in the way real insights on the performance of individual stocks than your average person.

Moreover, there is a certain randomness to the stock business, because there is a strong element of randomness in other peoples' investment decisions. For some reason the idea of market fluctuations reminds me of how passengers' heads all move the same way when an airliner hits turbulence. Nobody's immune to the short-term random fluctuations -- it's more a question of what happens over the course of time.

The problem with the premise of this article is that it assumes that there is no way to gain that expertise. However, there are investment companies that historically do very well -- precisely because they can afford to hire people to understand small sectors of the investment world. Those companies are just as subject to random market fluctuations as anybody else, but over the long term, they've shown the ability to pick investments that can ride through the fluctuations.

It seems to me that there's little more to this article than the seed of a sound business strategy (which has already been discovered many times).

The premise is that a group will always predict outcomes better than any single expert.

This is generally false unless the group has a clue what's going on. A group of non-medical people will do poorly in diagnosing a disease, for example. Likewise when about 50% of the people in the US do not know that the Earth goes around the Sun in once per year, such a crowd is poor at doing most scientific stuff.

In market analysis, the crown is probably fairly knowledgeable relative to the expert anyway. This is one reason for peer-review and for having committees. It's true that a single expert can miss things that a group (by virture of its varying experience) will detect.