I decided to share one of my trading secrets which I successfully use in finding or improving my trading systems and ideas. It was crucial discovery which helped me how to build my first profitable system. I know that this is not the only way leading to the  profitable trading, but this one works very well for me.

Why systems based on the technical analysis fail?

     I am short-term trader and technical analysis forms about 80% of my trading system. While  I was building my first system I was wondering why so many approaches fail after few months even in backtest. I bet that you know what I mean. You have some cool idea. You put it under some testing, it looks perfect after few months, but few months later you realize that this is only another dead end. I realized that there is a simple reason for it. It is a common truth that market conditions are changing all the time like a seabed. And we shouldn’t expect  that we can tie enviroment with so many elusive variables by some constant mathematic formula.

Example n .1:


There is equity of one of my trading ideas on the example screen where you can see how vulnerable this idea (which used about 4 different indicators) was to changing market conditions.  


What does it mean?

     It  means that using indicators with constant numbers, periods and so on, mostly can not work for a long time. I will give you an example:

Example n.2:We can say that trading system based only on crossing two moving averages with specified periods. For  example one with period 10 (MA10) and another with period 20 (MA20) is not working in long-term. You can easily prove it by putting it in historical tester in your platform (you can try all possible combinations of numbers).

     But it does not mean that systems based on crossing two moving averages in general do not work. Even crossing MA10 with MA20 is nicely profitable but in some periods of time, when market conditions change and there is time for different combination of moving average periods. And I believe that if you can adapt quickly enough and change periods of MA than you have good system… But as we know, it is far from being so easy.

Solution

    This problem covers all static aspects of technical analysis. Does it mean that technical analysis is wrong? No way, technical analysis is great thing. But you can not use it in naive form presented by thousands of tutorials, books and so on if you want to use it as a long-term profitable.

     Here is my solution.  I try to avoid all possible static variablilities in my trading since I have really realized this fact. It means that I do not use any kind of indicator unless I really know why it is on my screen and what it really shows to me. Here is an example:

Example n.3: I am using simple moving average with period 30 (SMA30). Why I put it on my screen? Because I am executing some of my trades on 2 minutes timeframe. And SMA30 shows me same information as SMA2 on 30 minutes period which shows smoothed close graph…

     I know exactly why it is there and what it tells me. I am looking for patterns which you can see even in little different timeframe. I usually execute on 2minutes timeframe so I want to see my pattern even on 1 minute or 3 minutes timeframes. I am not using rules like: “If it close one pip above…”. No, it is only another static variable in my system which means another source of problems and uncertain situations. The more adaptive system you have the more durable to changing market conditions it becomes.

Conclusion

     My advice is to try to use as little as possible static variables but if you want them than be sure what they are doing on your screen exactly. This approach leads to patterns and systems which is more robust and can work under more types of market conditions.

Example n.4:


There is equity of the same basic idea as you can see in Example n.1. But I removed lot of static variables from this strategy. The result is much smoother equity, on the other hand it is little harder to trade it due to more discretion approach.

    

But usually it is not possible to convert system like this into automatic strategy, because it is more discretionary. It is not perfect approach and you can get lots of dead ends, but it is a step forward, that helped me a lot and I believe that it can help somebody else too. I am sure that this style is not for everybody and I hope you share your own experience too.

I wish you all good luck and many pips.

Lukas