Queries From "Teach a Man to..." thread!!

[kenneth]

Hi Abbay,

I don't think think changing the percentage gives correct pivots to suit saints pivot trading system.

It is for a system where you sell when the next pivot is less than the HH by 2% or4% and vise versa.

Kathik and Saint please advise.

Regards
ken

[aad]

Originally Posted by kenneth

Hi Abbay,

I don't think think changing the percentage gives correct pivots to suit saints pivot trading system.

It is for a system where you sell when the next pivot is less than the HH by 2% or4% and vise versa.

Kathik and Saint please advise.

Regards
ken

I think there is some confusion and before I too get confused, I try to answer your doubts below.

Firstly about the zig() function. According to Amibroker http://www.amibroker.com/newsletter/08-2000.html

"As you can see Zig() function determines major peaks and troughs using percentage threshold given as a second parameter. The bigger threshold you specify the more significant peaks and troughs are detected. The line generated by Zig() function indentifies the main trend. "

Now zig function in our code receives its Q argument from the parameter that we select (dafault is 2). This argument is what they call '% change or % threshold'. We do not know what logic zig() function uses and how they use this argument but from the explaination given on their site as above, we can only conclude that zig() function returns an array which is sensitive to recent price when argument is a smaller number (like EMA(C,2)) and diversified consensus when argument is a bigger number (like EMA(C,5))

zig function only returns an array (of whatever, for our purpose, say closing price) based on parameter that we select. How we use it in our code to find HH and HL is entirely independent. So I don't think you are correct when you say this.

Originally Posted by kenneth

It is for a system where you sell when the next pivot is less than the HH by 2% or4% and vise versa.

To explain it further, in our code (where we define conditions for HH and LH), we compare Z with its previous value (Z,-1) and again compare (Z,-1) with (Z,-2) to know if (Z,-1) is higher or lower than Z and (Z,-2). Based on these conditions returning 'true' we then move further to compare whether peak() function returns true of false. In case of HH, we compare recent peak for Z is geater than its 2nd recent peak. If both these conditions are 'true' then HH gets 'true' flag or else 'false'.... similar logic goes on for LH, LL, HL code.

Edit : Also see http://www.traderji.com/amibroker/14...rough-zig.html for some more details.

I hope I have answered without adding to your confusion....

Regards,

Abhay

[mahesh2007]

Originally Posted by aad

Dear Mahesh,

As written in the previous post, you will have to adjust the %change parameter to see the correct pivots. With default 2% in price change, it gives you many false pivots. With parameter set as 4, you will get the correct results.

To change the parameter, right click chart, click on parameters and edit the field % change in pivot indicator to 4. This will give you the result you wanted.

Satyam chart with 4% price change is attached for your reference.

Regards,

Abhay

Yep! % was set to 1%,which gave the false result.
regards
mahesh

[mahesh2007]

Dear saint,

one more query,

while tracing pivots we should give emphasis to candle(top/bottom) than the shadows.
Right or wrong?
beacause a high wave candle or some long doji's may give unusual results.

please clarify.

regards

mahesh

[kenneth]

Originally Posted by aad

I think there is some confusion and before I too get confused, I try to answer your doubts below.

Abhay

Thanks Abbay. No confusion my friend. I was only saying that that AFL with Zig() function is not the one to identify Pivots to trade according to saints system

Look at the following thread and AFL


HELP NEEDED: What's wrong with the PlotShapes in this AFL (William's Third wise Man

--------------------------------------------------------------------------------

OHLC are all arrays. You dont need to, and, should not use it as array. Just try the following :

tmp = (High > Ref(High,-1)) AND (High > Ref(High,-2)) AND (High > Ref(High,1)) AND (High > Ref(High,2));
PlotShapes( IIf( tmp, shapeUpArrow, shapeNone ), colorGreen,0, H,10 );
Regards
ken

[mahesh2007]

kenneth, abhay,
my 2 cents-

I think we should think like this.

Pivot is the central point from where price pattern changes it's direction.
Lows are pivot lows and highs are pivot highs.

LL means "Lower pivot low" and HL means "higher pivot low"
similarly, HH means "higher pivot high" and LH means 'lower "pivot high".

'lower pivot low' indicates this 'pivot low' is lower than the previous 'pivot low'.

similarly "pivot high".

when % is set to 1,we get more "pivot high's" & "pivot low's"

beacause every "pivot high" which is more by 1% than the previous one, appears and the same for "pivot low".
this also gives some discrepancies.

when we set % to 4(++), we get less "pivot high's" and "pivot low's".

[prabhjeetrana]

Saint I was just going through the charts of Nifty and according to my view I found out the method where we trade dailies with weekly as a perspective seem to be working better than the method where we dont keep any perspective charts like you are teaching in thread GOING WITH 60 MIN. FLOW. Though I have not calculated the data but thats what the case looks like. So my question is :-

1. Do you think that trading both sides of a trend without perspective charts especially with brokerages, occasional whipsaws, etc. increasing the cost of trade let us have profits from this bidirectional strategy ??

2. My second question is regarding the entry. Suppose we are trading the Nifty daily with weekly as perspective. Now as far as I can make out the pivot high for weekly is at 4215 (spot), so the trend change signal will come at 4216, do we directly make the entry on daily charts or wait for new uptrend to start before making the entry .

And also looks to me that Pivot low is same on weekly and daily charts right now so how do we decide the stoploss??

Saint please take your own time to reply, I dont want you to be tired soon.

Regards

[aad]

Originally Posted by kenneth

Thanks Abbay. No confusion my friend. I was only saying that that AFL with Zig() function is not the one to identify Pivots to trade according to saints system

Are you sure ? Agreed that this forms only one ingradient of Saint's system. But zig() function is used for pattern recognition and we are very well able to identify the pivots using this code. Once again, there are some 'whipsaws' when we use lower values for '% change' (this is a confusing parameter... we are trying to corelate this parameter with '% change' of adjacent candles or adjacent pivots which is not the case actually) and smoothened values when higher values of '% change' is used.

Your post has given me a work for this weekend... I will try to design an automated trading system (I mean signalling system) based on this. Will post if any help is required.

Edit : Yes... I am not able to identify ALL pivots everytime (especially during intraday charts). So does not look like a straight forward task...

Look at the following thread and AFL

HELP NEEDED: What's wrong with the PlotShapes in this AFL (William's Third wise Man

--------------------------------------------------------------------------------

OHLC are all arrays. You dont need to, and, should not use it as array. Just try the following :

tmp = (High > Ref(High,-1)) AND (High > Ref(High,-2)) AND (High > Ref(High,1)) AND (High > Ref(High,2));
PlotShapes( IIf( tmp, shapeUpArrow, shapeNone ), colorGreen,0, H,10 );
Regards
ken

Thanks. Following this thread closely.

Regards,

Abhay

[prabhjeetrana]

Originally Posted by aad

Are you sure ? Agreed that this forms only one ingradient of Saint's system. But zig() function is used for pattern recognition and we are very well able to identify the pivots using this code. Once again, there are some 'whipsaws' when we use lower values for '% change' (this is a confusing parameter... we are trying to corelate this parameter with '% change' of adjacent candles or adjacent pivots which is not the case actually) and smoothened values when higher values of '% change' is used.

Your post has given me a work for this weekend... I will try to design an automated trading system (I mean signalling system) based on this. Will post if any help is required.

Edit : Yes... I am not able to identify ALL pivots everytime (especially during intraday charts). So does not look like a straight forward task...



Thanks. Following this thread closely.

Regards,

Abhay

Aad I dont why there is so much fuss about all these % pivots and zig zag function. I myself was using this idea sometime back with 5% pivots only but later found out that using larger % makes us leave too much of money on the table as we get signals sometime after very long time and we miss the trend change signals given by smaller 5 pivots. While using smaller % pivots the problem is that we keep on getting whipsawed too many times, believe me I have tried all these methods and they make much lesser money than Saint's original method.

BTW, Trading pivots in itself is so satisfying that I wont want to miss that part and added advantage is that we are making more money using classical signals. And tell me who has himself not got the time to track a few stocks in his portfolio, I do have.

This mechanical system's only use I think is that we can backtest various stocks which give best results with pivots methods, thats it though I even doubt that this purpose also gets solved

[rajatheroyal]

Dear saint sir iam a very big fan of yours.I have one doubt which only you can clarify. according to ur system, we should look at the weekly chart of a particular company and we should buy and only buy if the weekly chart is in a up trend and vice versa for down trend.My question is, should we buy if a weekly chart is in a down trend but the daily chart shows a strong double bottom signal?i have attached the chart of banknifty which shows a weekly down trend but the daily chart of the same may give a double bottom buy signal soon, should i buy?
My question is should i buy if the double bottom breaches the trough line