We often find that in the world of trading there is a constant battle between system developpers which argue for simple and complex system. It is quiet easy to think that the forex market -being extremely complex- would need a very complex system to achieve sustained profitability but such assumptions have proven to be wrong several times as systems which are exceedingly complex have proved to have their own pitfalls. On today's post I am going to talk about the level of complexity of automated trading systems and which approach I have found to be best. In particular, I am going to show you some of the pitfalls of added complexity and the advantages of simple trading strategies. In the end you will see that it is simply a matter of having an undestanding of how the market behaves.
So what do I mean by complexity ? The matter is pretty simple. What has more probabilities of being able to work in the market to achieve long term profits : a system which uses a simple indicator signal with adaptive money management or a system which has neural networks, takes fundamental data into account and analyzes the feed from 5 time frames and 12 indicators simultaneously ? The answer for many people seems to be "obviously the second one", however most of the time this is NOT the case. It is also true that simple systems can also fail so the answer is not a straightforward as many would like.
The fact is that the success or failure of a system depends merely on its ability to profit from a tradable market inefficiency. It has to have a positive mathematical expectancy. This can be achieved both with simple and complex systems, however - like in mechanical engineering - trading systems which are more complex are harder to make and more prone to failure (due to their higher complexity). Systems which do a higher number of computations and analysis do not necessarily reach better results and as a matter of fact, the failure of these complex systems is often very hard to predict and very hard to correct.
People often believe that adaptive behavior against market conditions is "very hard" to achieve, but when you understand the underlying properties of the market that change along different market conditions you can make a trading systems adaptive with less than 20 lines of code (often with even less). This shows that it is not a matter of making a computer do neural network analysis or other computationally intensive analysis, it is a matter of understading what you are doing.
Then the argument is reduced to robustness and simplicity. Why would you make an extremely complex system which is much more prone to failure when you can make a very simple system which is much more robust and bound to reach the same or even better results ? I have often found that the failure to reach simple trading systems is often a consequence of the lack of knowledge of the designer. The fact that few people do not understand how to exploit tradable market inefficiencies and how the market changes when market conditions change makes them resort to added complexity, often with disastrous results, such as excessive curve fitting.
However it is also true that people who aren't familiar with the world of automated trading and the algorithmic trading systems which have been developed over the past decades are often captivated by fancy words such as "genetic algorithm", "neural network", etc. These words do not mean anything, since what determines profitability is the ability to exploit a tradable market ineffiency and having added complexity simply does not guarantee this. Through all my experience in automated trading I have always decided to go with the simplest approach. I program the simplest trading systems I can which give me the results I desire, something which ends with the development of robust and adaptive, yet simple trading systems which I fully understand and trust. So as you see, it is not a matter of how complex the system is, it is a matter of how sound the logic behind the system is.
If you would like to learn more about my automated trading systems and the way in which they use simple logic to adapt to changes in market conditions please consider buying my ebook on automated trading or joining Asirikuy to receive all ebook purchase benefits, weekly updates, check the live accounts I am running with several expert advisors and get in the road towards long term success in the forex market using automated trading systems. I hope you enjoyed the article !
So what do I mean by complexity ? The matter is pretty simple. What has more probabilities of being able to work in the market to achieve long term profits : a system which uses a simple indicator signal with adaptive money management or a system which has neural networks, takes fundamental data into account and analyzes the feed from 5 time frames and 12 indicators simultaneously ? The answer for many people seems to be "obviously the second one", however most of the time this is NOT the case. It is also true that simple systems can also fail so the answer is not a straightforward as many would like.
The fact is that the success or failure of a system depends merely on its ability to profit from a tradable market inefficiency. It has to have a positive mathematical expectancy. This can be achieved both with simple and complex systems, however - like in mechanical engineering - trading systems which are more complex are harder to make and more prone to failure (due to their higher complexity). Systems which do a higher number of computations and analysis do not necessarily reach better results and as a matter of fact, the failure of these complex systems is often very hard to predict and very hard to correct.
People often believe that adaptive behavior against market conditions is "very hard" to achieve, but when you understand the underlying properties of the market that change along different market conditions you can make a trading systems adaptive with less than 20 lines of code (often with even less). This shows that it is not a matter of making a computer do neural network analysis or other computationally intensive analysis, it is a matter of understading what you are doing.
Then the argument is reduced to robustness and simplicity. Why would you make an extremely complex system which is much more prone to failure when you can make a very simple system which is much more robust and bound to reach the same or even better results ? I have often found that the failure to reach simple trading systems is often a consequence of the lack of knowledge of the designer. The fact that few people do not understand how to exploit tradable market inefficiencies and how the market changes when market conditions change makes them resort to added complexity, often with disastrous results, such as excessive curve fitting.
However it is also true that people who aren't familiar with the world of automated trading and the algorithmic trading systems which have been developed over the past decades are often captivated by fancy words such as "genetic algorithm", "neural network", etc. These words do not mean anything, since what determines profitability is the ability to exploit a tradable market ineffiency and having added complexity simply does not guarantee this. Through all my experience in automated trading I have always decided to go with the simplest approach. I program the simplest trading systems I can which give me the results I desire, something which ends with the development of robust and adaptive, yet simple trading systems which I fully understand and trust. So as you see, it is not a matter of how complex the system is, it is a matter of how sound the logic behind the system is.
If you would like to learn more about my automated trading systems and the way in which they use simple logic to adapt to changes in market conditions please consider buying my ebook on automated trading or joining Asirikuy to receive all ebook purchase benefits, weekly updates, check the live accounts I am running with several expert advisors and get in the road towards long term success in the forex market using automated trading systems. I hope you enjoyed the article !
Posts
No comments:
Post a Comment