When discussing market analysis, we generally consider the two contending schools of thought to be Fundamental Analysis and Technical Analysis. However, in the early 1970’s, there emerged a third view known as the "Random Walk Theory", which was not so much an approach to market analysis as it was a critique of the other two methods.
The Random Walk Theory is the popular name for a market model known in academic circles as Efficient Market Theory. This model of the market contends that prices are "efficient" in the sense that all known information and market expectations are immediately factored into the market through the movement of prices. But these price movements are caused by so many different factors that they become random in nature, and the only thing that we can be sure of is that the present price is the correct one because it is based on all known information. We cannot, according to EMT, reliably predict the movement of prices which keeps the market at this point.
This model centers around the Efficient Market Hypothesis, of which there are three versions; the weak, the semi-strong, and the strong. In terms of using analysis techniques to outperform the market consistently, the weak version basically says that technical analysis cannot work, the semi-strong says that neither technical analysis nor fundamental analysis can work, and the strong version says that nothing, not even illegal inside information, will work! More recent academic developments such as Behavioral Finance have presented strong arguments against the EMH, and in my view the hypothesis is not valid even in its weak form.
Proponents of the Efficient Market Hypothesis or "Random Walk Theory" generally have two lines of defense when it comes to supporting the idea that technical analysis of the markets cannot improve returns:
The first is that they have tested many TA indicators and have not found any that are able to lift returns significantly enough to overcome transaction and slippage costs. The counter-argument to this is that since they cannot test every possible indicator, and since there are certainly market inefficiencies which have been documented such as the January Effect, there could be many untested indicators which can in fact lift returns significantly.
The random-walkers then move to their second line of defense which says that even if there existed some indicators which could predict the market and give higher returns, they wouldn’t work for long once they were discovered. This is because once everyone started using such an indicator, their collective behavior would cancel out the indicator’s value. This is commonly stated as, "If everyone knows that the market is going to go up tomorrow, then it will surely go up today instead." A rebuttal to this is that the discoverer of such an indicator may not make it public, so there could be many undisclosed indicators out there that work very well. The response to this is that if an indicator is good enough, it will eventually be discovered by more and more people, and this will lead to the same "self-defeating" phenomenon.
Let’s play a brief game of "What if."
What if we just decide to re-examine this whole paradigm of "indicators that work" and "indicators that don’t work" and "indicators that work until everybody knows about them and then they don’t work" and so on? What if at any given time there are actually thousands of indicator combinations that work really well and millions that don’t? What if the set of "good indicators" changes over short periods of time for any given market? And finally, what if we had technology for finding and testing the combinations of indicators that are working well currently?
This would make the whole argument moot. Consider this alternative vision:
Oh, you found a great indicator? But you’re worried that everyone else will start using it and it won’t be effective for long? Don’t be. It wasn’t going to be effective for long anyway. They never are. Besides, everyone else just found some great indicators too, and they’re all completely different from the one you found. So have a good time with your indicator. Make lots of money. Just remember that it will wear out in a week or two and you’ll have to find a new one. Just like the rest of us will…
This view is based on one of the preliminary models I have of the way the market might work. In a nutshell, I don’t think that the main reason indicators become less effective over time is because everyone uses them. I think the effectiveness of any given indicator rises and falls over time because the behaviour of individual markets changes over time. I think the behaviour of the market in a given instrument changes over time because the market participants change over time.
Market participants are the different traders, whether they’re individuals or institutions, who are actively engaged in moving into and out of positions in a given financial instrument. The only reason the price of the instrument moves around is because of these market participants. Note that my definition of "market participant" does not include people who are just passively holding a position in the financial instrument. This is because if they’re just holding a position then they’re not moving the price around, and therefore do not participate in the creation of "price behaviour."
Each market participant has his or her own particular quirks and unique approach to trading. Some are trend followers and others are contrarians. Some are quick on the trigger and others have nerves of steel. Some are scalping for small gains while others are swing trading. Some use tight stops, some use wide stops and some don’t even know what stops are.
At any given time, there is a specific set of these market participants involved in trading the instrument, and all their individual trading styles combine to make the price behave in certain ways. Different sets of participants will cause different behaviors. For example one FOREX currency pair may have a tendency to always turn on a dime, while another may often consolidate for long periods before changing direction. One may have a very orderly movement and form well defined chart patterns while another is very choppy and undefined, especially around major tops. And as the group of market participants changes over time as some drop out and others begin trading the pair, the price behavior will change with the gradual evolution of the group. This will cause some indicators to work well at some times and poorly in others.
So the way to get superior returns out of the market is not to just find a set of terrific indicators and keep them to yourself. It’s not to develop a single well thought out trading strategy. It’s to constantly scan the market’s price behavior, trying out many combinations of indicators and strategies to see which ones are currently working well. This kind of constant dynamic statistical analysis of historical price movements can only be done with the aid of computers.
While I agree that technical analysis in its classical form is not satisfactory, I believe that its practitioners should be supported and educated by the academic community…not ridiculed by them. It has become clear that the EMH is not widely applicable to markets in the real world. There are trading opportunities which arise in markets and persist for long enough periods of time for traders to profit from them, if they have the correct tools to find and validate these opportunities. The academic world and the trading world can and should work more closely together to explore these possibilities and develop these tools.
This is one of the ideas that I have managed to flesh out the most in my research into FOREX price behavior. I developed a tool called a Price Behavior Map which tracks how various TA indicators have been performing recently across a group of seven currency pairs. By using this tool, I hope to get a better idea of how price behavior evolves over time. And I may just make some nice FOREX trading profits along the way!