Adaptive Indicators 4 - Adaptive MACD

It is weekend and time for my experiments…on Technical Analysis. I am still stuck on the topic of Adaptive indicator. Of course I have been doing a little bit of reading on my favourite subject Volume. The little time I get during the week days have been spent on the book “Investing with Volume Analysis” by Buff Dormeier.  Now I am working on how the VPCI can help in VSA studies.

More on VPCI later. Right now let me continue with the current topic Adaptive Indicator. I have been experimenting with making the simple moving average adaptive to the dominant cycle. This adaptive moving average was used for smoothing in the ADX calculation to make it adaptive which was explained in the last post.  The next obvious move was to create a MACD of the adaptive moving average. The MACD is basically the difference between two moving averages one short and another long. However in case of the moving average adapted to the dominant cycle the period itself is not fixed and is varying. So I calculated the adaptive MACD with two adaptive moving averages, one adaptive to the dominant cycle and the other adaptive to twice the dominant cycle.   As the basic behind the MACD is the difference of two moving averages we cannot find much difference between the conventional MACD (12, 26) and the adaptive MACD. However the adaptive MACD is less prone for less whipsaws and the catch the trends very well at the same time the catches the turning points in time. The Adaptive MACD is definite one notch better than the conventional MACD.

The Afl -  Adaptive MACD V.1.0

Adaptive Indicators 3 - ADX Adaptive

I am continuing with my experiments with adaptive Indicators. The possibilities are extensive and exciting. The next indicator I was looking was the ADX. It is one of my favourite indicators and I had done some work on it already like the Gaussian smoothed  ADX and a Volume biased ADX.

If we look at the the calculation of the ADX, there are two places where we can make it adaptive. First in the calculation of MDI and PDI we have the smoothing of the UP move and down moves. Then we have the average true range. SO we can make the Smoothing period adaptive tot he Dominant cycle period. Also the averaging of the True Range would be based on the Dominant cycle period. Finally the smoothing in the calculation of the ADX itself will be made adaptive to the Dominant cycle. The images are provide below. We can notice that though the Adaptive ADX does not drastically change the ADX it  is more pronounced and the moves are swifter and faster. The AFL is provided for those who are keen to experiment with this Indicator.

The AFL - KADX - Adaptive

Adaptive Indicators - 2 . Adaptive Stochastics

Last week we looked at an Adaptive CCI. Now let us look at extending our experiments in creating a adaptive stochastic. As you are aware the stochastic indicator calculates the difference between the current close and the lowest value in a range as a ratio of  the entire range. This ratio is converted to a percentage and then a small smoothing is applied.

%K = (Current Close - Lowest Low)/(Highest High - Lowest Low) * 100

The Lowest low and Highest High is calculated over a specified look back period. Here we will calculate these values on the Dominant cycle (DC) period. The  DC is again calculated using the Hilbert transform of John Ehler.  The DC period is a varying parameter depending on the market condition and the Stochastic calculation is automatically adjusted as per the varying DC period making it adaptive to the Market condition.

You can see in the chart below that the adaptive Stochastic is giving less whipsaws and is smoother. The afl is also provided for you to experiment. Do share your finding and thoughts.

The KStochastic afl

ADAPTIVE INDICATORS - 1 - Adaptive CCI

All of us who use technical analysis are used to Indicators with fixed period of calculation. Also certain periods like 9 and 14 are taken for granted. These indicators do not taken into account the market conditions. Sometimes the market is volatile and not at other times. The velocity and acceleration also differs at different times. The indicators are oblivious to all these are mainly static. For example for CCI most people use 14 as the period whereas the inventor of this indicator Donald Lambert has suggest one third of cycle period.

I always have been a fan of John Ehlers. . Ehlers suggest that making the indicators adaptive to the cycle period of the concerned market would be a better solution. However he suggested using fraction of the cycle part in adaptive formulas. One of my first trials on this concept was an adaptive CCI which I called as KCCI. Of course many of you who had visited my web site (Now almost defunct) would be aware of this. I am enclosing a image of the adaptive CCI(KCCI) with the conventional CCI. You can notice that the whipsaws are lesser in the KCCI compared to the conventional one. You can experiment with it by optimizing the cycle part.

I will discuss more about other adaptive indicators in the coming posts. The afl is shared here

http://www.4shared.com/file/nH9Ss4-p/KCCI_-_Adaptive_CCI_-_V20_.html

SUPER ADX !!

Finally we have the so called Super ADX, I mean a clone. I could get the time to finalize it after a long time. It is based on the MT4 super ADX indicator. First of all I wish to clarify that it is not the Holy Grail.  Of course we should understand that if one had the Holy Grail Indicator we would never share it any body at whatever price. Since he is not able to make enough money with that system from the market he is trying to make money by selling it. The newbie trader should first understand that there are always many people with their eye on his purse. Everyone wants a few bucks from his purse. Some take it legally by selling something and some steal it.

Coming back to the Indicator, Let us see what this does.  The indicator first calculates the difference between the moving average of the average price and the moving average of the close.  This is taken as a measure of the strength of the trend. The change of directions of this strength measure is a reversal point. However the validity of the reversal point is made when the ADX (Another measure of trend strength) is either above a value of 35 (normally end of a trend) and below a value of 10 (for starting of a Trend). It also looks at reversal of the MACD. Finally   it also looks at multiple reversals within the last 15 bars. Different signals are generated for different type of reversal signals.  The different signals are as follows.

1.       A red bar would indicate that a big move is imminent.

2.       A white bar would indicate that it is the end of the big move and hence exit.

3.       A yellow bar or a violet bar would indicate a possibility of a reversal of the current trend.

4.       A magenta bar would indicate that multiple signals have occurred and possibility of a reversal is very strong.

Also it should be noted that the exit signals is not generated in all cases. Similarly the Big Move indication is also not generated every time. But whenever it happens it will be a good opportunity. Also reversal signals are not generated every time reversal is imminent.

The super ADX also uses a indicator which they call Money line. This line is used as the exit point though I did not find it correct. Normally a Linear weighted Moving average of period 40 should be good enough as a money line. 

So try out the indicator and provide your feedback. I only hope that unscrupulous elements don’t take this and commercialize this as they have down with other my other indicators.

The Super ADX Indicator

The Money Line

The ADX Afls...

I have been experimenting a little on the volume biased ADX which I had propose in my last post. I found that volume spikes could drastically upset the the indicators. SO i had to use a little smoothing to overcome this. That is the reason for not releasing the afl. Now  I am ready to release it for those who would like to experiment. I am also posting my earlier work, a Gaussian smoothed ADX called KADX which is more responsive ans smoother than the classical ADX. Please share your opinion about these indicators.

1. Volume Biased ADX

2. Gaussian Smoothed ADX - KADX

A Volume Biased ADX

I am returning to blogging after a long hiatus.  It is usual in life that we face intervals of disturbing events and confused priorities.   Now that the dust is settling down I am slowly picking up the threads and I am coming back to my passion, Technical Analysis.

I am  sharing some thoughts on the ADX indicator. ADX is one of the very popular indicators and widely used by technical analyst. Dr. Charles Schaap’s  book “ADXcellence” deals with different strategies of trading with ADX. Then there a commercial product called Super ADX which the sellers claim is a leading indicator though the ADX by itself is a lagging indicator. Also I do not know how much of it is based on real ADX other than they have an indicator called supporting ADX. Anyway soon I will release a clone of the so called Super ADX.  I also had released a Gaussian smoothed ADX called KADX which was more sensitive and smoother than the conventional ADX. ADX is a useful indicator to measure the strength of a trend. ADX by itself is a non directional Indicator and does not indicate if the trend is up or down. We use two other indicators called the +DI and -DI to reveal which trend direction. I will not go into further details as most of you, readers of this blog, will be familiar with the ADX indicator.

As many of you may be aware the importance I provide for the volume in my analysis. Any move of a stock which is not aided by volume will not last. Volume is the fuel for any trend. So naturally any trend strength indicator without consideration of the volume cannot provide the real picture. The ADX calculation totally ignores volume and purely based on the price movement.  Price action combined with volume would be definitely provided a more realistic picture. This thought has inspired to experiment with the ADX to include the volume aspect.  The basis for the ADX calculation are the difference between the High of a day and the High of the previous day and also the difference between the low of the day and low of the previous day. In other words it uses the momentum of the Highs and Low. This momentum is fuelled by volume. So we will add a volume factor to this momentum. To calculate the volume factor we take the ratio of the current volume and the average volume. We will bias the momentum of the Highs and Lows with this volume factor. The remaining of the calculation will be based on these biased values and the final ADX, +DI and –DI values will be biased by volume. The result of this is quite evident from the chart below. The resultant +DI and –DI are much more responsive to the volume and they clearly indicate which moves are volume driven and which are not.  The ADX also quickly increase in values when the volume increases indicating increased trend strength.