Patterns and Technicals

[AW10]

Re: Trading only one hour in a day !!!!!

Coming back to this thread after a gap. Just want to share my opinion / observations across the posts in last 20 days..

1) Do not be afraid of loosing trades given by a proven system. If the system is backtested and it has given +ive expectancy over a period of time , then trader just needs to trust his system. His job is to blindly execute the trades without worrying about the outcome of individual trade. If he has not backtested himself then developing this level of blind trust is difficult. That is when hardwork of trading business comes in... to create, backtest, improve, personalise the system..

2) Stoploss for this system - base system that we discussed here was targeted to open the trade near mkt closing time and close the trade near /at the mkt open. So stoploss was meaningless for trade duration because market was close.
If we are planning to extend the exit time from Mkt open to next days market close, or afternoon etc, then stoploss can be considered. My approach to test stoploss will be to
a) change the exit rules section of the strategy
b) revisit the trade list from backtesting excel sheet for new exit rule. The column that had Exit Price will change now. Some of the prev trades will get stopped out now.
c) once tradelist is ready, then analyse the statistical output of the modified strategy.
d) compare the stats of original v/s modified strategy. The numbers will not lie to you.

3) If a strategy is working, then do not change it.. but clone it and create a new strategy with amendments in Rules. The moment we start looking at 3 time periods to monitor (like Ghosh Sir's option strategy), then we are no more trading 1hr a day but we are day trading. Nothing wrong with that, but then we need to go back to drawing board to the strategy and see if the underlying logic / various rules for entry, exit, stops makes sense or not. And then, backtest it and see if the strategy has positive expectancy or not.

4) Coolboy - to address your constraint of trading only between 3 to 3.30 pm period,
You will need a new strategy called "Trading 30mins a day" or "Trading last 30min".
My suggestion will be to backtest the new exits. You can take prev backtesting sheet, change the exit price from Next day open to next day close.. and boom... you have the stats in front of you.
You have made good observations that 90%days have lower close on next day. If you remember the underlying psychological reasoning that I had posted earlier, also mentions that profit booking generally come in sometime on next day and price reverses. In bear market, this reversals are strong we end up in Low Close on next day.

Happy Trading.

 

[rkkarnani]

Re: Trading only one hour in a day !!!!!

Guys please check out the back tested sheet from the day we had started using this method...Plzz download the XLS...the dates are from the bottom of the sheet and the last date would be on the top...The result is amazing..and another thing..for things not to be very complicated i have taken the close price only for buying or sellin..there fore the loss or profit would vary a bit from the result displayed..plzz adjust the xls just in case...sorry for the delay..

Thanks Kamlesh for reviving your thread!
Kamlesh, I could not understand the trades mentioned in your Excel file. Please mention in brief the rules u have followed in trading and preparing the excel sheet.

Also could not relate the OHLC levels mentione therein with Nifty Futures level available with me. The levels do not match.
Please calrify per your convenience.
Thanks again.

 

[bandlab2]

Re: Trading only one hour in a day !!!!!

Guys please check out the back tested sheet from the day we had started using this method...Plzz download the XLS...the dates are from the bottom of the sheet and the last date would be on the top...The result is amazing..and another thing..for things not to be very complicated i have taken the close price only for buying or sellin..there fore the loss or profit would vary a bit from the result displayed..plzz adjust the xls just in case...sorry for the delay..

kamalesh, thanks for the spreadshot. i am not able to understand how you derived the profit/loss column. can you explain the logic ? as per my nderstaning, you initiate the trade at 3:15 after consiering hi, lo for the day and square off the position next day open. in your case i dont know what is the exit price , is it next day open or close or what price ?

thanks again.

 

[ag_fx]

Re: Trading only one hour in a day !!!!!

Guys please check out the back tested sheet from the day we had started using this method...Plzz download the XLS...the dates are from the bottom of the sheet and the last date would be on the top...The result is amazing..and another thing..for things not to be very complicated i have taken the close price only for buying or sellin..there fore the loss or profit would vary a bit from the result displayed..plzz adjust the xls just in case...sorry for the delay..

There are 2-3 observations that I have made by following this thread so far and would like to highlight them, especially after this file of yours:

1) You have considered closing price of Day 1 and Opening Price of Day 2 as your entry and exit points. However, it must be noted its more than difficult to trade those levels. Almost impossible. A more viable approach to backtest/forward test this strategy would be to consider entry time as 3:25PM and exit time as 10:00 AM. I belive this would give a much better picture of the strategy and more practical one too.

2)Some times day hi/lo is the opening price...as it happens in gap openings...Thus, lets test this strategy on a more practical note.

3) Also, it would be good for us to inculcate position sizing and specific entry & exit rules for a robust strategy.

I am not too sure how have you calculated the PnL in your excel sheet. If you can elaborate on the same, it would be gud.

Regards

 

[pkamalesh]

Re: Trading only one hour in a day !!!!!

Hi all...just read all your posts...in the spread sheet..i have deliberately taken the exit price at the low( if short) and high(if long) of the next day...i have an observation that we cannot fully capitalise on just the opening price on the next day...so if you seniors can add more to it i thought we could extend the time frame of the trades because as you can see we have been given ample oppurtunity to book good profit at some part of the day...and another thing regarding the spread sheet..please follow the following..

1. The spread sheet has days starting from the bottom and not from the top..

2. The buy or sell signal here has been initiated as per the 70-30 rule..
3. The profit is calculated as per the diffenece in price at which bought/sold on the previous day and the exit price(which is the point near the high or the low) of the present day..The profit for the trade taken yeaterday would be booked and displayed today...

Thanks for the response and i hope you guys will keep adding a few importatn things..Bu the fact is that if we can identify the exit price properly we could book good trades..

PS: I have taken nifty spot for the calculations and not nifty futures...

 

[AW10]

Re: Trading only one hour in a day !!!!!

Thanks Ankit for your observation. Having involved in this thread from begining, I am posting my views on your observations and how they had been addressed earlier..

1) You have considered closing price of Day 1 and Opening Price of Day 2 as your entry and exit points. However, it must be noted its more than difficult to trade those levels. Almost impossible. A more viable approach to backtest/forward test this strategy would be to consider entry time as 3:25PM and exit time as 10:00 AM. I belive this would give a much better picture of the strategy and more practical one too.

Due to the limitation of availability of historical data for backtesting, this approach is fair approximation of the reality. To improve it further, we can add/deduct 5 or 10 points from Close and Open price and be a bit conservative in backtesting. i.e. for long entry, take trade entry price as 5 points above close price that means we got poor entry price.

2)Some times day hi/lo is the opening price...as it happens in gap openings...Thus, lets test this strategy on a more practical note.

I am not sure, how will this impact the strategy. We will never be able to simulate the market 100% in testing the strategy. For the trade setup of the strategy, the mkt day is defined between 10 am to 3.15 pm. Even if the low or high is made at 3.29 pm, we would have already got the same trade signal to long/short by 3.15 pm in majority of cases. (assumption here is that market does not cover 30% of days range in last 15 min which is reasoable assumption, IMO).

3) Also, it would be good for us to inculcate position sizing and specific entry & exit rules for a robust strategy.

Postions sizing can't be defined globally as account size and risk tolerance level are important factor in defining that. eg - for 90-10 setup of this strategy, max risk is about 150 pts, average risk is about 50 pts. then depeniding on my risk per trade limit (say 50pts or 500pts still falling within 2% of my acct size limit), I might take 1 contract or 10 contract. So it should be done by individuals to meet their account size requirement.

In one of my previous post, I had defined detail rules for this stretegy as sample. Certianly, people should personalise them by making adjustments to these rules and backtest it.

Happy Trading.

 

[rkkarnani]

Re: Trading only one hour in a day !!!!!

Hi all...just read all your posts...in the spread sheet..i have deliberately taken the exit price at the low( if short) and high(if long) of the next day...i have an observation that we cannot fully capitalise on just the opening price on the next day...so if you seniors can add more to it i thought we could extend the time frame of the trades because as you can see we have been given ample oppurtunity to book good profit at some part of the day...and another thing regarding the spread sheet..please follow the following..

1. The spread sheet has days starting from the bottom and not from the top..

2. The buy or sell signal here has been initiated as per the 70-30 rule..
3. The profit is calculated as per the diffenece in price at which bought/sold on the previous day and the exit price(which is the point near the high or the low) of the present day..The profit for the trade taken yeaterday would be booked and displayed today...

Thanks for the response and i hope you guys will keep adding a few importatn things..Bu the fact is that if we can identify the exit price properly we could book good trades..

PS: I have taken nifty spot for the calculations and not nifty futures...

Kamlesh, What you have projected is not achievable in actual trades. You sell today and you have shown that as being bought at the lowest rates the next day.
Two things :
One : You cannot know before hand the low of the next day hence cannot exit at that level.
Two: The level you exit may come at any time hence the veru purpose of trading for an hour only is defeated.

Taking the figures for 20th November,
O: 2634.20
H: 2634.20
L: 2502.90
C: 2553.15

You have shown the Trade as SELL, where as you are showing the difference as 38.72% which would come to 61.73 :: 38.27 ,beyond 70-30 rule. Its more likely that we shall have no trade as the close is not within 30% of the days range from the high or low.

The range is : 2634.20 - 2502.90= 131.30*30%=39.40
We go long if the price closes > 2634.20-39.40= 2594.80
We go short if the price closes < 2502.90+39.40= 2542.30
The price closes at 2553, which does not fulfil our crieteria for either Long or Short.

Have responded as per my understanding and would be happy to be contradicted and corrected.

 

[bandlab2]

Re: Trading only one hour in a day !!!!!

Hi all...just read all your posts...in the spread sheet..i have deliberately taken the exit price at the low( if short) and high(if long) of the next day...i have an observation that we cannot fully capitalise on just the opening price on the next day...so if you seniors can add more to it i thought we could extend the time frame of the trades because as you can see we have been given ample oppurtunity to book good profit at some part of the day...and another thing regarding the spread sheet..please follow the following..

1. The spread sheet has days starting from the bottom and not from the top..

2. The buy or sell signal here has been initiated as per the 70-30 rule..
3. The profit is calculated as per the diffenece in price at which bought/sold on the previous day and the exit price(which is the point near the high or the low) of the present day..The profit for the trade taken yeaterday would be booked and displayed today...

Thanks for the response and i hope you guys will keep adding a few importatn things..Bu the fact is that if we can identify the exit price properly we could book good trades..

PS: I have taken nifty spot for the calculations and not nifty futures...

kamalesh,

your open price is wrong for all the days. if you notice the prev day close and current day open are more or less same. If you apply the original rule, your net gain is 0. dude you need to pay due deligence before uploading the test results. too many flaws. data is wrong, calculations are wrong. cant believe you mentioned 880 points gain in 16 trades !!!

can you correct the open price and add a new column exit price and calculate P/L as diff between prev close and todays open. that will give a good picture on this strategy

i am still hopeful, it will give a decent gain in the long run provided you ca sleep well with the naked positions over night

 

[rkkarnani]

Re: Trading only one hour in a day !!!!!

kamalesh,

your open price is wrong for all the days. if you notice the prev day close and current day open are more or less same. If you apply the original rule, your net gain is 0. dude you need to pay due deligence before uploading the test results. too many flaws. data is wrong, calculations are wrong. cant believe you mentioned 880 points gain in 16 trades !!!

can you correct the open price and add a new column exit price and calculate P/L as diff between prev close and todays open. that will give a good picture on this strategy

i am still hopeful, it will give a decent gain in the long run provided you ca sleep well with the naked positions over night

Bandlab, Kamlsh is not exiting atthe opening rate next day. He has clarified this. If he is short he is exiting at the lowest rate achieved next day and if he is long he is exiting at the highest rate achieved next day. He has erred as pointed out by me in anticipating such trades. As per my view there are some errors in taking a position too!!!
Let us wait for Kamlesh's view on it. He may have some thing in his mind yet to be revealed.

 

[pkamalesh]

Re: Trading only one hour in a day !!!!!

kamalesh,

your open price is wrong for all the days. if you notice the prev day close and current day open are more or less same. If you apply the original rule, your net gain is 0. dude you need to pay due deligence before uploading the test results. too many flaws. data is wrong, calculations are wrong. cant believe you mentioned 880 points gain in 16 trades !!!

can you correct the open price and add a new column exit price and calculate P/L as diff between prev close and todays open. that will give a good picture on this strategy

i am still hopeful, it will give a decent gain in the long run provided you ca sleep well with the naked positions over night

Dear bandlab....Let me reiterate that i have used nifty spot and not nifty futures for the calculations...Regarding the open prices...anyways as i have already told you this was just a practical model of how this system works and not a perfect one yet...I have assumed to exit NEAR the high or near the NEAR the low the next day....This profit booking also should be a function of the amount of risk (in terms of time to be waited before booking profit) and this might vary from person to person....what i have also mentioned ealier is that we now need to find out ways to exit with the maximum profits...even at the cost of extending the time frame....

PS: dude regret the mistake of taking spot...making the calculations for nf....