A paper seems to offer a mechanical rule for beating the stock market by a substantial margin. Is there some trick?

Ask an Economist
Question: 

A recent Economist article www.economist.com/news/finance-and-economics/21639587-beware-heavily-tra... reports on an academic paper that seems to offer a mechanical rule for beating the stock market by a substantial margin. Is this a disproof of the efficient market hypothesis or is there some trick?

Answer: 

Assume that you decide to buy a new cell phone. Are you going to buy iPhone or Xiaomi, some unknown brand? For a typical US consumer, you have about 40% chance to buy iPhone, but almost 0% chance to buy Xiaomi. An iPhone costs you about $600, but a Xiaomi costs you about $200. Now, what if everyone wants to buy Xiaomi instead of iPhone? After a period of time, Xiaomi’s price will increase and iPhone’s decrease. The market works.

Now apply this logic to the stock market. For an unknown stock, it has small trade volumes and probably a cheap price. And for a fashion stock like Apple or Google, it has large trade volumes and a high price. But if everyone knows the trading strategy suggested by Ibbotson and Kim and applies this strategy to buy only unknown stocks. Then their prices and trade volumes will increase. If no one buys Apple or Google, they will have small trade volumes and cheap prices. So you see, what makes a fashion market? It is just you, the investor. And you are looking for some trading strategies to beat yourself.

I am not saying that the stock market is always efficient. A lot of evidence challenged the effect market hypothesis, such as the price index is too volatile, the trading volumes are too large, stock prices shows short-run momentum and long-run reversion, etc. Maybe fashion itself has some other value.