Is Forex trading chaotic, deterministic or...manipulated?

Problems requiring analysis and resolution that can be assisted by powerful computers include those where the initial conditions and the rules for solution are well defined, even if the work required is onerous. An example of this type is the development of chess playing programs, which are said to fall into the category of being deterministic.  

At the other end of the spectrum are those problems where initial conditions have to be chosen by reference to the level of detail that is practicable, and where the circumstances that might be encountered during resolution are imprecise. Weather forecasting by computer falls under this heading. This type of problem has been described as “chaotic”.    

In all cases the way to proceed is to develop algorithms, or descriptions of the method for solution that proceed in steps that can be coded into a computer language. In both weather forecasting and the playing of chess, this process is now well advanced.  

The application of algorithms and computer power to electronic trading has been the focus of a lot of activity in recent times. It might be instructive to decide where algorithmic trading sits in the spectrum between the extremes of chaotic conditions on the one hand, and purely deterministic ones on the other. This is important. If algo trading tends towards determinism then the only barrier to outright success is the computer power that can be brought to bear on the question.  

However, there is another issue. The rules of chess are fixed - all efforts to build a computerised algorithm that can defeat the very best human grand masters proceed on the understanding that the number of pieces and the number and arrangement of the squares on the chess board are fixed, and the moves that are allowed for each piece are unchangeable. Long term suspected changes in climate due to global warming aside, the weather, also, cannot be made subject to human intervention.  But trading can be.

As only one example, the practice of stop-hunting is pervasive and well documented. Consider this recent message from the Dow Jones newswire service, on a day when a speech by Mr. Bernanke of the US Federal Reserve was anticipated after an FOMC meeting:

DJ MARKET TALK: Euro Drives Toward $1.30 Just Ahead of Bernanke; Algos?

14:21 EDT - Were high-speed traders trying to trigger stop losses, options before the Fed chairman starts answering questions? Whatever the motive, it was a similar pattern before the Fed statement at 12:30, when markets moved sharply just beforehand. This is what algo does to markets. They move before news, trying to catch slower players off guard when the news actually comes out. Healthy? You be the judge. Bernanke now speaking. (michael.j.casey@dowjones.com)

(END) Dow Jones Newswires

September 13, 2012 14:21 ET (18:21 GMT)