Markets *may* be efficient, but they're not smart

A recent article in The American Prospect gets the problem with markets right:

Markets [...] are not "smart," and smart people do not claim otherwise. Rather, markets are supposed to be efficient. There is a difference. A man who finds himself in a convenience store without sufficient funds for his purchase might decide that the most efficient method of obtaining cash is to rob the register. He might well be right. But he would not be smart. He would not be wisely considering his own future or soberly evaluating the possible consequences of his actions. A Hostess cupcake is not worth jail time.

The market, however, carries no such capacity for self-reflection. It prizes efficiency above all -- at least, insofar as efficiency is reflected in profit. It will squeeze profit from coal and trap heat in the earth's atmosphere. It will sell cigarettes to children and mortgage-backed securities to adults. It will build bombs for dictators and deny medical care to the sick. It will support all manner of economic activity and do so rather mindlessly. And so we must intervene when efficiency contradicts wisdom, or profit does injury to ethics. We must impose intelligence on the market.

This may be the best description of how markets are a collection of self-interested firms, and so don't make socially optimal decisions. Most importantly though, they are run by self-interested managers. The principal-agent problem means we barely get real efficiency from a market, let alone societies best interests.

I have forgotten the source, but a few years ago I found a book by a mathematician that argues books on how to be a successful CEO may be a total scam, as there may be no real secret, but instead total randomness. Being a CEO of a top company doesn't mean you're any good.

Imagine 100 CEOs who make random business choices, 50% of which are right and make good profits that year. Next year, another 50% are right, etc. In 5 years, only 3 were right the whole time. Does that make them in the top 3% of CEOs, or just real lucky? This has always seemed like a reasonable model of business leadership to me.