Experiments in Technical Analysis

oxusmorouz · #**721** 4th February 2007, 07:37 PM

Hi Karthik,

Here is the code:

Indicator 1:
Name : "zz macd divergence binary wave bullish"
Formula : If((Mov(C,12,E) - Mov(C,26,E)-Mov(Mov(C,12,E) - Mov(C,26,E),9,E) > Ref(Mov(C,12,E) - Mov(C,26,E)-Mov(Mov(C,12,E) - Mov(C,26,E),9,E),-5) AND H < Ref(H,-5)) AND (Mov(C,12,E) - Mov(C,26,E)-Mov(Mov(C,12,E) - Mov(C,26,E),9,E) > Ref(Mov(C,12,E) - Mov(C,26,E)-Mov(Mov(C,12,E) - Mov(C,26,E),9,E),-4) AND H < Ref(H,-4)),1,0)

Indicator 2:
Name: "zz macd divergence binary wave bearish"
Formula: If(((Mov(C,12,E) - Mov(C,26,E)) - Mov((Mov(C,12,E) - Mov(C,26,E)),9,E)
< Ref((Mov(C,12,E) - Mov(C,26,E)) - Mov((Mov(C,12,E) - Mov(C,26,E)),9,E),-5) AND C > Ref(C,-5)) AND(Ref((Mov(C,12,E) - Mov(C,26,E)) - Mov((Mov(C,12,E) - Mov(C,26,E)),9,E),-1) < Ref((Mov(C,12,E) - Mov(C,26,E)) - Mov((Mov(C,12,E) - Mov(C,26,E)),9,E),-6)AND Ref(C,-1) > Ref(C,-6)),1,0)

----> Expert :
In highlights :
1) Bullish divergence: Fml( "zz MACD divergence binary wave bullish") = 1 (colour = green)
2) Bearish divergence: Fml( "zz macd divergence binary wave bearish") = 1 (colour = red)

In Symbols:
1) Buy :
Fml( "zz MACD divergence binary wave bullish") = 1 AND C > O AND C > Ref(C,-1)
2) Sell:
Fml( "zz macd divergence binary wave bearish") = 1 AND C < O AND C < Ref(C,-1)

This was coded about 2months back when I started off with metastock coding. You would find terrible coding techniques used here and hence, you must excuse me.

However, here is the MACDH formula with customizable time frames (created today):

FMA:= Input("FastMA",1,500,12);
SMA:= Input("SlowMA",1,500,26);
Sig:= Input("Signal",1,500,9);

a:= Mov(C,FMA,E);
b:= Mov(C,SMA,E);
d:= a - b ;
f:= Mov(a-b,Sig,E);
d-f;
0

BTW, I'd also recommend you visiting www.metastocktools.com
Jose's MACDH divergence toolkit is considered superb but it costs E400 I guess.

Oxy

kkseal · #**722** 4th February 2007, 08:33 PM

Quote:

Originally Posted by oxusmorouz

Hello Kalyan,
The slope of the MACDH would be more than 45* usually when the corrective rally is short but sharp and the succeeding rally is sharp. This is an implication that the market has corrected too much in a very short period of time...
Putting it in other words, are we not perhaps assuming directly or indirectly that the traditional elliot wave principle of a 3 wave correction holds good? (Sorry if I'm crapping here) This setup would probably go for a sixer when there is another retracement and another sharp rally. I have posted a chart which decipts the same

Oxy

I was simply trying to figure out whether the chances of failure of EMA crossovers was less when the slope (or rather it's angle) of the corresponding MACDH was greater than 45 degrees - (nothing to do with Elliott but perhaps something to do with GANN). This seems to hold (specially when measured at the point of crossover) - even in the chart you've posted.

In any case i'm not looking to apply it in the MACDH of prices.

Regards,
Kalyan.

oxusmorouz · #**723** 4th February 2007, 08:47 PM

Improved MACDH divergence kit

Download Link:
http://www.************/file/10030008...ivergence.html

Metastock coding for metastock 8 users:

Indicator 1: zz MACDH
Formula:
FMA:= Input("FastMA",1,500,12);
SMA:= Input("SlowMA",1,500,26);
Sig:= Input("Signal",1,500,9);

a:= Mov(C,FMA,E);
b:= Mov(C,SMA,E);
d:= a - b ;
f:= Mov(a-b,Sig,E);

d-f;
0

Indicator 2: zz MACDH Divergence
Formula:
If( H = HHV(H,60) AND Fml( "zz MACDH") = LLV(Fml("zz MACDH"),40),-1,
If(L = LLV(L,25) AND Fml("zz MACDH") = HHV(Fml("zz MACDH"),25),1,0));

Now go to expert advisor ---> new (name it MACDH Divergence kit) ---> symbols:

Symbol 1: Buy
Formula: Fml( "zz MACDH divergence") = 1 AND C > O AND C > Ref(C,-1)

Symbol 2: Sell
Formula: Fml( "zz MACDH divergence") = -1 AND C < O AND C < Ref(C,-1)

Go to highlights:
Highlight 1: Bullish divergence
Formula: Fml( "zz MACDH divergence") = 1

Highlight 2: Bearish Divergence
Formula: Fml( "zz MACDH divergence") = -1

A hectic process, I must say.
I have enclosed a couple of charts.
Note : Trend kills this system (generates too many false trades when in an established trend). The system does not kill the trend

Oxy

oxusmorouz · #**724** 4th February 2007, 09:11 PM

Kalyan, I was just trying to establish a link between EWs and MACDH divergence ,LOL...as absurd as it may sound, I strongly believe they are derived from the same premise --> a 3 wave correction

karthikmarar · #**725** 5th February 2007, 12:46 AM

Quote:

Originally Posted by kkseal

Hi Karthik,

Great going on the MACD. Just a few questions on the MACD Histogram part

1) Would it make any difference if we plot the difference between the MACD & the Signal as a line plot instead of a histogram (as it is, 'histogram' in this case is i feel a bit of a misnomer as there's no frequency involved; it's more of a bar graph The line i suggest would be one connecting the peaks of the bars)

Hi kalyan

Actually it does not make a difference how you plot it as long as you can interpret it properly. But the Histogram presents a easier to visualize and understand picture especially in case of divergences. Also the histogram lends itself to some nice color coding to make the understanding easier and quicker.

Quote:

2) Is there any way of calculating the slope of the MACD histogram? Would the LRS of the difference between MACD & Signal be good enough? I've observed that a slope above 45 degrees works quite reliably (whereas anything lower than 45 would be suspect). This, if it proves reliable in an overwhelming majority of cases, can be made use of in Anant's system.

Yes you can do the LR on the difference between the MACD and the Signal line. Also you can use Ananth’s innovative idea of BB on the Histograms as well as the LR of Histogram. But the inherent problem in using this method is the whipsaws. I have been experimenting a lot on Ananth’s method and realized that the whipsaws will be a inherent problem in this method which can be only reduced by coupling it with another indicator.. More of it later

Right now I am not in a position to comment on the Chaikin Oscillator.

Enclosing a sample chart showing the MACD with Histogram, LR of close and LR of the Histogram.

Regards

Karthik

uma_k · #**726** 5th February 2007, 11:40 AM

Hi Anant,

No new post from you for long time. Are you working on any new twist to Experiments in Technical Analysis.

Regards

*** Uma ***

kkseal · #**727** 5th February 2007, 01:25 PM

Thanks, Karthik.

Regards,
Kalyan.

kkseal · #**728** 5th February 2007, 01:31 PM

Quote:

Originally Posted by oxusmorouz

Kalyan, I was just trying to establish a link between EWs and MACDH divergence ,LOL...as absurd as it may sound, I strongly believe they are derived from the same premise --> a 3 wave correction

Could be. If prices rise & fall in 5 & 3 wave patterns then it will get reflected in the MACD as well since both reflect the same price movements.

asnavale · #**729** 5th February 2007, 07:26 PM

Quote:

Originally Posted by karthikmarar

Hi kalyan

.....................

Yes you can do the LR on the difference between the MACD and the Signal line. Also you can use Ananth’s innovative idea of BB on the Histograms as well as the LR of Histogram. But the inherent problem in using this method is the whipsaws. I have been experimenting a lot on Ananth’s method and realized that the whipsaws will be a inherent problem in this method which can be only reduced by coupling it with another indicator.. More of it later

...............................

Regards

Karthik

Hi Karthik,

You are right. The Histogram method has inherent problem of whipsaws. I have already used the Hi-Lo threshold and MA threshold to reduce them. I have described the two options in an earlier post in this thread. You might have gone through that. I was expecting some comments from you on this. Although there was reduction in whipsaws the two alternatives had some other problems in specific cases as I have pointed out in that post. The MA threshold still gives a few more whipsaws compared to Hi-Lo threshold.

I am working on another alternative. It appears to give better performance than all the earlier variants. I will post it in a day or two. Your comments and guidance will be very useful for refining the system.

Regards

-Anant

kkseal · #**730** 5th February 2007, 09:59 PM

Quote:

Originally Posted by asnavale

Hi Karthik,

You are right. The Histogram method has inherent problem of whipsaws. I have already used the Hi-Lo threshold and MA threshold to reduce them. I have described the two options in an earlier post in this thread. You might have gone through that. I was expecting some comments from you on this. Although there was reduction in whipsaws the two alternatives had some other problems in specific cases as I have pointed out in that post. The MA threshold still gives a few more whipsaws compared to Hi-Lo threshold.

I am working on another alternative. It appears to give better performance than all the earlier variants. I will post it in a day or two. Your comments and guidance will be very useful for refining the system.

Regards

-Anant

Hi Anant,

Will be eagerly awaiting your post. I had read your earlier post but really didn't know what to say as both the thresholds seemed to have their own share of problems.

Your efforts are commendable and i'm sure there's some way to fix this (to a reasonable degree); but at the same time any filter should not leave out large chunks of profitable moves - i'm sure you'll agree to that.

Regards,
Kalyan.

P.S.: This whipsaw problem (not just for your system but more or less all MA based systems) somehow keeps playing at the back of my mind all day. I'm sure you or Karthik or both will succeed where my limited intelligence fails.

Similar Threads for: Experiments in Technical Analysis
Thread	Thread Starter	Forum	Replies	Last Post
Technical Analysis Course	snowy	Technical Analysis	28	7th November 2008 07:37 PM
Technical analysis course	sh50	Seminars & Tutors	18	10th June 2007 03:46 PM
Hello everybody.Technical Analysis	vijay chauhan	Introductions	2	30th May 2007 03:09 PM
Technical Analysis	crosscurrency	Forex	0	30th October 2005 11:00 PM