Efficient markets and their logic

By 1970 the question that had been hanging over this whole history since Cowles could finally be stated properly: can anyone beat the market? This chapter has three papers. The first says, in effect, no. The second explains why not in language a practitioner can use. The third shows that the answer has to be “almost, and only for a fee,” and that this is not a contradiction but the way markets must work. Between them they settle how I think about paying anyone to manage money.

Fama, 1970

Eugene Fama did not discover that prices are hard to predict. Bachelier had said it in 1900 and Cowles had measured it in 1933. What Fama did, in a long review article, was gather twenty years of evidence into one clear claim and give it a name: the efficient markets hypothesis. Prices, he said, reflect the information that is available. He split the claim into three levels. The weak form says you cannot profit from past prices, so charts do not work. The semi-strong form says you cannot profit from public news, because by the time you read it the price has moved. The strong form says you cannot profit even from private information, and that one Fama himself rejected, because insiders do make money.

It became the assumption every empirical study in finance had to argue against, and it gave index funds their intellectual license: if the price already knows what you know, hold everything and stop paying people to guess.

The important thing about Fama is what he did not claim, because people get it wrong constantly. He did not say prices are always right. He said they are very hard to beat, which is a different statement, and one that has survived every attack on it. Even the anomalies found later, and there are many, mostly turned out to be either compensation for a real risk or too small to profit from after costs. I use efficiency the way the curriculum says to: as a starting assumption, not a religion. The default for any pool of money is to hold the market. Anything that wants to deviate from that has to explain, in writing, which of Fama’s three levels it thinks it can beat, and why that edge will still be there in five years.

Ellis, 1975

Charles Ellis said the same thing to the people who actually manage money, and he said it in eight pages that have aged better than anything else I have studied. His observation was social rather than mathematical. When he began his career, professionals did ten percent of the trading on the New York Stock Exchange. By 1975 they did seventy. The person on the other side of every trade was no longer an amateur with a hunch. It was another professional with the same training, the same information, and the same computers.

That changes the game, and Ellis borrowed the perfect image to explain how. In tennis, professionals win points with shots their opponents cannot return. Amateurs lose points, with double faults and balls into the net, and the match goes to whoever makes fewer mistakes. Investing, Ellis argued, had quietly become the amateur kind of game, even for professionals, because the professionals were now playing each other. Nobody could win by brilliance. The winner would be whoever made the fewest errors and paid the least to play. And he did the arithmetic that Sharpe would later formalize: a manager has to beat the market by the full amount of fees and trading costs merely to tie.

The practical lesson is a question, and I ask it about every market before I put money in: who is on the other side of my trade? If the answer is professionals like me, then I am in a loser’s game, and the only sensible play is to minimize cost and error and hold. If the answer is someone else, someone forced to sell, someone uninformed, someone with a different clock, then there may be a game worth playing. Ellis is often read as counsel of despair. It is the opposite. He is telling you to choose your games. And for a family the edge is almost never picking securities. It is structure: taxes, a long horizon, the ability to hold illiquid things, and the absence of a boss who will fire you after a bad year.

Grossman and Stiglitz, 1980

There was a hole in the efficient markets argument, and Sanford Grossman and Joseph Stiglitz put a theorem through it. If prices really reflected all information, nobody would bother to gather information, because there would be no profit in it. But if nobody gathered information, prices could not reflect it. So perfect efficiency is impossible. Markets have to be inefficient by exactly enough to pay the people who do the research that makes them efficient. Not more, not less.

This sounds like a debating point and it is the most practical idea in the chapter. It tells you where active management can work: in markets where information is expensive to get and there are enough uninformed traders to profit from. And it tells you where it cannot: in large, liquid, well-covered markets where a thousand analysts already know everything and the profit from knowing it has been competed down to the cost of the analysts. It also tells you that the profit, where it exists, goes to the person doing the work, not to the client, unless the client is getting something the fee does not fully capture.

So the test I apply before paying anyone for active management, in one sentence: name the uninformed money you are going to take, and name the information cost you are recovering. If nobody can name either, the expected result after fees is negative, and no track record changes that.

What the three add up to

Passive is the default. Active management is a paid service whose price is set by how costly information is, and whose value to the buyer, after that price, is usually nothing. The two mistakes that cost the most money are the cartoon versions of each side: believing prices are always right, so valuation does not matter; or believing anomalies are everywhere, so trading a lot must be worth it. Both are expensive. The middle position, that markets are exactly as inefficient as they need to be to pay for their own efficiency, is where I live.

Gala: before you pay anyone to be clever with your money, ask them who they are going to be cleverer than. If they cannot name that person, the person is you.