When I first got interested in markets I thought technical analysis was a great tool. It was easy to learn, made logical sense and most importantly, appeared to work.

I was fooled by this cunning, seductive mistress. Technical analysis is the fools gold of finance. It is an illusion that is as real as Michael Jackson’s face. But to quote Levar Burton’s Reading Rainbow: “You don’t have to take my word for it”.

The most eye opening book I’ve ever read on finance is hands down Benoit Mandelbrot & Richard Hudson’s The (Mis)behavior of Markets. Who is Benoit Mandelbrot? He is the father of fractal geometry. A mathematician who looks at the markets as a form of organic randomness. But the markets can’t be random, can they? No. Markets are not random. However, understanding randomness is essential when judging technical analysis.

Technical analysis attempts to predict future price movements based on movements in the past. This is done through averages, equations and charts. This data is then used to determine when to buy/sell a stock.

There are a couple of problems with this philosophy.

Firstly, there are a large number of individuals who believe that the past does not reflect the future. Lets say you were thinking of buying 100 shares of Coca-Cola. What is important to you, what the stock did yesterday, last month, or last year? Or is it what the stock will do tomorrow? You probably care about the next earnings report, the next news article, the next executive decision. What happened in the past is done with and already built into the share price. The market is forward looking and all about future expectations.

The second problem with technical analysis has to do with human deficiencies in perception and randomness. It is human nature to look for patterns, even when patterns aren’t really there. In experiments conducted by Mr. Benoit, he created random charts using his fractal geometry and gave them to technical analysts with real charts using market data. The analysts were blind to the fact that some of the charts were computer generated, as they all had no identification, time, or price scales on the axis. The result? The analysts found trends in the random charts. How can randomness have trends?

They don’t. But we think they do.

Another example:

Our parent company runs a website called ConquerClub, an online game of world domination that involves “virtual dice”. When programming the dice they wanted to ensure complete randomness so the programmer turned to random.org as source for the data. Study after study has concluded that the dice rolls on ConquerClub are random. However, the largest complaint from users is that they are *positive* that the dice are in fact not random. They call them streaky.

Streaky dice & Stock trends.

It’s the same thing. An illution created by our brain in an attempt to understand randomness.

So now we know technical analysis is junk. Why is it still important?

It has become a self fulfilling prophecy. 

People are the ones making investment decisions. If many people believe that a stock will go up once it crosses a $35 dollar barrier, what happens? They buy the stock! When people buy the stock, it drives the price up and the system works! Another successful technical analysis prediction.

So go on looking for trends you crazy technical analysts, you control your own destiny.