Quadra Safe Trading Strategy - I

[asnavale]

Hi Anant,

As far as data that I download from 'Getbhavcopy' for Equity & Futures, on the 17-04-2012 BNF made a High of 10658.00 which had Triggered a BUY and subsequently on the 18-04-2012 it was taken out by making a high 10674.80 thus making it active (marked on the chart by vertical green dotted line) where the BUY was placed one point above the high of 17-04-2012 @ 10659.00..... but the problem as I realised in hindsight was the 'Initial S/L' which should have been a little lower than that of 29-03-2012 instead of a point lower than that of 11-04-2012.

Reason.... that a point lower than Low of 29th would have been the correct Initial S/L is because we have a gap-up opening on the 30th March....therefore high probability of closing the Gap should have been taken into account by me....and that is where I made the grave error.....!!!

This is one of those instances of Gap-up.....which will perhaps need to be looked at closely, next time it happens....can't remember I came across such an instance while manually backtesting NF...!!!

SG

Hi SG,

The difference in the trigger dates as observed by me and you is due to the data. As I understand from your post above, you are using BANKNIFTY-I data which is always the current month. Therefore, the data for April will be for April Contract, data for March will be for March contract (because BANKNIFTY-I in March is March contract) and in February it is February contract. Therefore, when you use BANKNIFTY-I data you are combining different Near Month data.

I am also using the data downloaded from GetBhavCopy and combining BANKNIFTY-III data of February (which is April contract), BANKNIFTY-II data of March (This is again April contract) and BANKNIFTY-I data of April. Effectively I am using the three months' data of the same contract. This results in small differences in the calculation of the Averages and therefore I got the trigger on 18-4-2012 with an initial stoploss below the low of 16-4-2012. However, as I mentioned earlier the high of 18-4-2012 was never taken out and there was no trade at all.

I am attaching the chart image below with the markings.

-Anant

 

[adityasaraf007]

Hi TP.... You had earlier asked for MT4 indicator which changes the color of our MA as per its slope.... Found this one from net.... Have modified the Default settings.... It doesn't do what we require as there are 3 colors... But solves the purpose...

Drop the file in your MT4 Terminal's /experts/indicators/ folder....

There are 3 input parameters...
A. MAPeriod: This is obvious...
B. MAType: 1 for EMA, 2 for SMMA, 3 for LWMA, 4 for LSMA, 5 for SMA.... (Don't know what they mean...
C. MAAppliedPrice: 0 for Close, 1 for Open, 2 for High, 3 for Low, 4 for Median (Avg of H+L), 5 for Typical (Avg of H+L+C), 6 for Weighted (Avg of H+L+2C)....

 

[SavantGarde]

Hi SG,

The difference in the trigger dates as observed by me and you is due to the data. As I understand from your post above, you are using BANKNIFTY-I data which is always the current month. Therefore, the data for April will be for April Contract, data for March will be for March contract (because BANKNIFTY-I in March is March contract) and in February it is February contract. Therefore, when you use BANKNIFTY-I data you are combining different Near Month data.

I am also using the data downloaded from GetBhavCopy and combining BANKNIFTY-III data of February (which is April contract), BANKNIFTY-II data of March (This is again April contract) and BANKNIFTY-I data of April. Effectively I am using the three months' data of the same contract. This results in small differences in the calculation of the Averages and therefore I got the trigger on 18-4-2012 with an initial stoploss below the low of 16-4-2012. However, as I mentioned earlier the high of 18-4-2012 was never taken out and there was no trade at all.

I am attaching the chart image below with the markings.

-Anant

Hi Anant,

How can I possibly argue with your logic....

Think I will switch to using the Banknifty & NSEnifty from the Equity Bhavcopy that should take care of any aberrations in the Data that creeps in regularly from F&O Bhavcopy....!!!

SG

 

[SavantGarde]

Dear SG,
As per my data directly from NSE bhav copy directly downloaded EOD from NSE, buy trigger has happened on 16/4/12. See the chart attached.

Any reason for not considering 16/4/12 as buy trigger as all wilders MA and EMA are in order.
Data of close is EOD adjusted as per NSE and displayed in bhav copy, not last traded value.

Now I need a few more clarifications.
You are using a filter of 1 for above trigger. Any specific reason. Should it not be better it is placed on some % of ATR.
Second what is the filter below SL in case of long?
Further I am unable to understand connection of SL and gap.
Once a higher high on 13/4/12 and higher low on 11/4/12 is formed, why low of 11/4/12 ie 10060 can not be considered as SL
Can you please explain?
Or give a link to that if it is explained earlier.
Sorry for the trouble.

Hi LVG,

I really don't remember why Buy Trigger wasn't noticed on the 16th ...it could be plain oversight on my part or it is possible the WWMAs looked different at EOD of 16th......or perhaps 50 EMA was flattish....!!!

Will take up the balance queries in another post...!!!

SG

 

[timepass]

Hi LVG,

I really don't remember why Buy Trigger wasn't noticed on the 16th ...it could be plain oversight on my part or it is possible the WWMAs looked different at EOD of 16th......or perhaps 50 EMA was flattish....!!!

Will take up the balance queries in another post...!!!

SG

This reinforces my belief that some technical confirmation be defined to determine the trend of 50EMA, apart from visual confirmation

 

[SavantGarde]

Hi TP,

About the Dates of Entry & Exit.... have to still do it manually for Dailies...infact, only realised that dates were done for Monthly & Weekly....after I posted a reply to you......give me a day or two to compile the dates for Dailies.....!!!

SG

 

[timepass]

Hi TP,

About the Dates of Entry & Exit.... have to still do it manually for Dailies...infact, only realised that dates were done for Monthly & Weekly....after I posted a reply to you......give me a day or two to compile the dates for Dailies.....!!!

SG

No hurry.. another weekend or two will be ok. Actually I wanted to check the readings for indicators like KST, Stochs, RSI etc.. on the dates of the triggers.

 

[SavantGarde]

This reinforces my belief that some technical confirmation be defined to determine the trend of 50EMA, apart from visual confirmation

Hi TP,

Let me look into it .....I am still working with an experimental Histogram....which will take sometime....!!!

Will look to see if I can get the Hourly and the 5 minutes Charts directly from ODIN....Let's see how it goes tomorrow...!!!

SG

 

[Satyen]

Hi TP,

... Yess....Initial S/L in most entries is quite deep....but manual backtest of NF had indicated only couple of instances where the Initial S/L had Triggered....!!!

...!!!

SG

Dear SG Sir and Annat Sir with due respect i want to contradict this point ...

i.e something has not happened in past does not gurantee that it will not happen in future ....

I have remeber in 315 thread they have tested years of data they did not got whipsaws more than 3/4 times but now they are facing whips of 6/7 or more times which is really frustating ...(May be i am wrong but read some where in their thread )

So Sir my point is there should be a hypothetical worst case scenario that may not happenned in past but we should know what can be possible worst thing happen to the system and what are the remadies etc

Hope I am not cotradicting much but this is just a thought please consider it


Regards
Satya

 

[timepass]

Dear SG Sir and Annat Sir with due respect i want to contradict this point ...

i.e something has not happened in past does not gurantee that it will not happen in future ....

Agreed. Basically, you are re-affirming that there is no holy grail; that any strategy should be flexible enough to incorporate changes in trading practices; that a strategy may become stale over time and constantly needs tweaking. Ok.