Log in

General Brain trend system discussion

  • Public
Started by JohnLast 2163 days ago Replies (4)

Here I add this topic for the general  discussions of the Brain Trend system. This is the general discussion topic for this system. Refer to the pages and the blogs in this group for the main issues. The download files are within the files in this group. 

 

    JohnLast 2163 days ago
    image
    От 21 октомври 2011

    OK the beginning of this year we see a shift in the market state for the EURUSD. We are not talking about trend and direction and bullish or bearish bias. We are talking about specific market state that can be appreciated as a visual pattern and quantitatively evaluated.

    The market state was within the parameters of the Brain Trend system, you can see that it was profitable under all possible parameters. That is why I show that on the picture. This is a robust system working on low fractal dimension time price series.

    What is fractal dimension, yes I see a lot of guest here. You can read more on in the the most simple manner in the group about market state analysis.

    The basic rules are not so difficult, used in a discretioanry manner folloe the brain trend signal (or any other good directionnal system you prefer) when in low fractal dimension. In that way you would add a statistical element in your discretionary trading.

    Look when the iVAR is in the blue area, that is related with the range characteristics in the market. What for a technical break - out that is linked with a shift in the fractal dimesion characteristics (going below 0.5).

    However I have tested this pattern in the current market conditions this pattern gave less profitability than just following the market action. I would say that the market was relatively simple and overall we had quite a few fake directionnal signals. The Brain Trend filtered out quite well those signals, in fact we did not had the killer of the brain Trend system, antipersistence combined with high volatility, the volatility was directionnal and with impulse character. 

    I know that it is quiete strange and those reasoning is not typical nor standard for a technical analyst. 

     

    JohnLast 2161 days ago
    image
    От 21 октомври 2011

    Here I would add the continuation of the  same system. You would see that the performace decreased sharply when we are in a range bound market.

    The quuestion is if we can know beforehand that the system is not likely to perform. My hypothesis is that it is possible. The fractal dimension measures give us  that possibility. Here on the example whe we have iVAR > 0.5 the probability for a directionnal signal is weaker.

    Howeve it is not as easy to use this in practice, because it is hard for me to automate the decision making. However using a discretionary approach together with the human pattern recognitions it is quite usefull.

    The fractal dimension estimators would help the dicretionary break - out trader to avoid some traps.

    So basically there are two main instruments: the directionnal indicator, the fractal dimesnion estimation and the human pattern recognition.

    I use the brain trend expert in order to know what are really the best settings for the current moment.

     

    JohnLast 2156 days ago
    image
    От 21 октомври 2011

    Today I add another screen shot of combination of market state analysis and brain trend. 

    As you know Brain Trend is not a grail, however at particular market states it performs pretty well. 

    From the beginning of 2012 we have very nice predictability windows with low fractal dimension iVAR < 0.5 or FGDI < 1.5. 

    The theory says that when we have this we can take into account the directionnal systems. Why? 

    The reason is because we have a greater mathematical probability that the next movement will be in the same direction as the previous.

    In practice that means that if you have a signal from your directionnal signal that means you have identified a signal into a particular direction. If the fractal dimension is low you have by independant way a confirmation of this.

    As you can see that explains why the Brain Trend expert was profitable with all the settings of the digital filtering (regulating the smootheness of the filtering). That was because the market state was there.

    On the other hand if we look at the market state we can see that according to the basic classification we have a Volatile Trend. The direction is DOWN. So basically you could use this knowledge in order to optmize the risk.

    - A Conservative approach: Trade only in the direction of the main trend

    - A Risk taking approach: Take all the signals from your system. This may sound unwise but I do not think so. The idea here is that those market conditions of high level of predictability are pretty rare and when they happen they have to be exploited as much as you can.

    However everyone has its own risk tolerance and appetite.

    Another approach is when you see a predictable market structure is to try to time the enty at lower market frame where you have also a predictability window.

    By the way since the last 3 months as it was stated there was a high degree of predictability on the 1 h time frame that was exploited with neural nets.

     

    JohnLast 2152 days ago
    image
    От 21 октомври 2011

    After working unsuccesfully with some piece of code and unfortunate mathematical formulas I decided to get back to the basic brain trend and to see what he gave following the last shot. 

    Take a look it is most interesting. In fact even if the fractal dimension of the iVAR is close to the 0.5 point the Brain Trend calculations were not confused by it. 

    Why then?

    The fact is that Brain Trend is based on market reality calculated by a break - out over calculated treshhold. This is a kind of variation of the Wilder's break - out strategy. So even if the market action is not predictable the Brain trend still may give you the right direction provided that the market action does not have a big volatility. The bigger volatility will jam the break - out treshold mechanism. 

    So the market action was trending and quiet despite the fact that the fractal dimension was close to the unpredictability. 

    You can have another inerpretation too.

    The question is if this climb is due to the closing of orders because of the end of the week or it is a genuine break - out.

    I would speculate that it is not a genuine break - out because if the fractal structure, it is close to 0.5. I would like to see an upward impulse, breaking technical levels but with low fractal dimension.