The Efficient Market Hypothesis Almost Always Works

Date: September 10, 2022

There’s a classic joke where two guys are walking down the street when one of them sees some money. “Look, there’s $20 on the ground!”, he says.

His friend, an economist, doesn’t even glance down before replying, “Can’t be. If there were really a $20 bill, someone would have picked it up already.”

This principle is known as the Efficient Market Hypothesis, and it states that there are no easy ways to make a lot of money, since otherwise someone would’ve exploited them by now. It will tell you that a “get rich from home” ad is probably a scam, but it can also smother innovation.

Imagine you’re talking with your friend when she pitches you a startup idea. Without sparing a thought, you reply “Nah. If that were really a good idea, someone would’ve thought of it by now.” And in this instance you’re probably right. But you’ve just given a fully general counterargument against ever starting a company![1]

Scott Alexander writes about Heuristics That Almost Always Work. For some classification problems, you can get to 99.9% accuracy with a heuristic that takes basically no effort. The problem is all the value of the classification lies in the 0.1%.

Of course in some cases the EMH is more applicable than others. If you see a dime on the sidewalk, you probably wouldn’t doubt its legitimacy, but if it’s a $100 bill in Grand Central Station, you can be pretty confident it’s a trick to get your attention. Consider factors like how many other people would be in a position to claim the prize, and how big the opportunity is.

So go ahead and enter that competition; start that startup; write that blog post. Someone has to be the one picking up the free money and it might as well be you.

But maybe you shouldn’t listen to my advice: after all, if this post were any good, someone would’ve written it already.


[1] This is why startup founders tend to be overconfident. Any reasonable person would’ve done the sensible thing and worked for Metabet.