Discuss, backtest & implement pivot based trading systems here

trader111 · #31 2nd August 2008, 11:44 PM

Quote:

Originally Posted by kenneth

Hi Trader111,

Thank you for the excellent post giving whole picture given on "Going With The 60min Flow!!! ". At the same time I notice that your post on this thread has been deleted. Since you have understood the system perfectly Will it be possible for you to write the code as the backtest shows that it can give a profit of 25 to 35 % for 3 months after commission.

Regards
ken

Kenneth,
1. There is nothing like 'perfectly' understanding a system unless it's your own.
2. I deleted my post along with several others (abt 15) which looked like unapropriate . Anyways the intention was to point out even better than 25% returns, I was following Saint's trades hence the perception. BTW also had in mind several 'scientific' members who were saying that pivots are crap.
3. You seem to understand the rigorous process of coding. Well, sure will post here when I succeed to code his system.

Thanks.

aad · #32 9th August 2008, 12:24 AM

Quote:

Originally Posted by milind

I noticed that the new pivot code is detecting only major pivots (two lowers on each side for a high pivot). Although a good start and improvement over zig function, you are going to need lot of fine tuning. Minor pivots are considered at times - e.g. PL of 4298.

-- Milind

Dear Milind,

I know it needs fine-tuning and for that purpose, I need feedback from the users, particularly followers of 60 min. flow thread started by Saint. I have seen map presented by Trader111 in that thread and this system addresses many of the issues like, entry, exit and stoploss (except position sizing). I was away for a week and was expecting feedbacks from our experienced traders i.e. whether this system works. I will be following this system for day-trading from next week and will post my feedbacks as well.

Regards,

Abhay

aad · #33 9th August 2008, 12:26 AM

Quote:

Originally Posted by kenneth

Hi Trader111,

Thank you for the excellent post giving whole picture given on "Going With The 60min Flow!!! ". At the same time I notice that your post on this thread has been deleted. Since you have understood the system perfectly Will it be possible for you to write the code as the backtest shows that it can give a profit of 25 to 35 % for 3 months after commission.

Regards
ken

Dear Kenneth,

Did you try backtesting the latest system ? What were your observations ? Where it is lacking, needs improvements ? Where it is working correctly as expected ? I am awaiting your feedback.

Regards,

Abhay

kenneth · #34 9th August 2008, 12:43 AM

Yes I did testthe system. On hourly BT profits are around 34% for 88periods. But the trades have to be continueous and haveto be in futures 1 lot only.

On 5 minute BT it shows a loss of 6% with .03%brokerage. Probably you have not taken the brokerage for BT Please check at your end.Then if the trades are continueous they should be in futures.

Correct me if I am wrong

regards
ken

aad · #35 9th August 2008, 12:55 AM

Quote:

Originally Posted by kenneth

Yes I did testthe system. On hourly BT profits are around 34% for 88periods. But the trades have to be continueous and haveto be in futures 1 lot only.

On 5 minute BT it shows a loss of 6% with .03%brokerage. Probably you have not taken the brokerage for BT Please check at your end.Then if the trades are continueous they should be in futures.

Correct me if I am wrong

regards
ken

Thanks Ken for your reply.

Yes... you are right. The trades have to be continuous and so that is only possible in futures. Can you please send me Nifty Futures zipped data files for 2-3 months at least ? I will backtest on them as well. Sending you PM for email id. According to me, the profit % should be much higher that indicated by you but I will confirm this only after backtesting nifty futures data.

Regarding brokerage, I have taken 0.01% (which I pay). The same is visible under settings tab in the backtesting results i had included.

On shorter time-frames (5min), you are right. This system does not work. It works well for 15min, 30min and 60min very well. We will have to finetune entry, exit rules (and anything else ?) for shorter time-frames. May be we consider other factors such as volumes and other indicators in combination of our logic ? Pl. post comments.

My another problem is, I only have 2 months intraday data to backtest with me. I am trying to get additional data for backtesting for better analysis.

Best regards,

Abhay

uasish · #36 10th August 2008, 06:30 AM

http://www. 4shared. com /dir/1945306/b41e3d61/Intraday_Future_data.html

(Fill 3 Gaps)

Download from here.

aad · #37 10th August 2008, 02:37 PM

Quote:

Originally Posted by uasish

http://www. 4shared. com /dir/1945306/b41e3d61/Intraday_Future_data.html

(Fill 3 Gaps)

Download from here.

Appreciated. Thanks to you for pointing as well as Soumya for posting it.

Regards,

Abhay

gkpc · #38 10th August 2008, 03:08 PM

Can anybody code the Pivots for Amibroker posted here also in Metastock!!!

aad · #39 10th August 2008, 11:14 PM

Enclosed is the little modified system which shows the pivots calculated. Rest of the things remain same. Also I have commented the text showing max. shares that can be bought, etc. which clutters the chart when zoomed out.

Regarding backtesting of NiftyFO data, I have backtested it and the results were not encouraging, hardly 10% of profit.

I know actual system works hence the reason could be :

1. Pivot definition needs finetuning further.
2. Incorporation of 2 bar exit method in vertical movements which I have not done yet. Plan to incorporate the same asap but struggling to catch the correct condition at which we can call it a vertical movement.
3. I am also confused about what to do when there is a gap up / down opening.

Inputs will be most appreciated.

Regards,

Abhay

sikandar · #40 16th August 2008, 12:41 PM

Quote:

Originally Posted by aad

Enclosed is the little modified system which shows the pivots calculated. Rest of the things remain same. Also I have commented the text showing max. shares that can be bought, etc. which clutters the chart when zoomed out.

Regarding backtesting of NiftyFO data, I have backtested it and the results were not encouraging, hardly 10% of profit.

I know actual system works hence the reason could be :

1. Pivot definition needs finetuning further.
2. Incorporation of 2 bar exit method in vertical movements which I have not done yet. Plan to incorporate the same asap but struggling to catch the correct condition at which we can call it a vertical movement.
3. I am also confused about what to do when there is a gap up / down opening.

Inputs will be most appreciated.

Regards,

Abhay

HI add

Did you fine tune this system more further as you mentioned?
regards
sikandar