iCharts Discussions

[sumesh_sol]

IOC...EOD
========

[vinst]

Bharti airtel in flat BB10,2

[sumesh_sol]

AUROPHARMA...eod
================

[vinst]

superb compression in RIL daily chart

[shrigajanan]

Relcapital- BB10 BO on Daily charts, check out!

[sruthi_kulkrni]

BB10

[sumesh_sol]

RCOM,,Gapup open sell case

[sumesh_sol]

[shrigajanan]

Petronet appears to be a nice BB10 candidate on Daily charts

[chetan83]

[shrigajanan]

Ride it till it last, read trailing SL suggested in the strategy

[amitsaraf31]

Hi All

Backtested this strategy for Index futures for 250 days.

Results are excellent.

I did make some changes as per my comfort level and limitations.

Excellent results doesn't mean very high profits but it means reasonably good market returns in different market situation with some amount of consistency and limited drawdowns.

And yes i backtested it for intraday and swings in nifty and only intraday in Banknifty futures.

For Nifty, Swings performed better than intraday but intraday results were acceptable as well. With intraday one can afford to have smaller SL and higher leverage which swings may not permit.

I thought of producing my backtested results here but thought that it would be great if some body who wants to seriously follow this simple, non discretionary, mechanical, easy to implement strategy, would backtest it himself. It will take somewhere between 5-10 man-hours to backtest and analyse 250 day data for a script.

So if you backtest 10 scripts it is just 50 hours.

Also you may backtest various variations to this strategy ( which i am yet to do). Some variations that come to top of mind are:

1. 30min TF 8 period BB, 8 period SMA.

2. Initial stop loss at MA after entry or at opposite BB.

3. Keeping a fixed point stop loss.

4. Booking profits at several levels.

5. Rejecting some trades after particular time of the day.

6. Booking profits at several points in TIME.

And in this way various combinations can be tried to find the system with highest positive expectancy as well as which meets other vital parameters for system testing.

All the very best ...........

Amit

PS: A drop of sweat spent in a drill is a drop of blood saved in war

[p_wangdu]

[svkum]

AMITSARAF -
GREAT
will check your modifications, in my original strategy, shd be much better.

[amitsaraf31]

Based on what i have read on some god trading books and also what sounds logical.................

Why it is important to test variations (not with this nice strategy but any other system strategy that we look to trade)..................

Now when we test various variations we know the strength and weaknesses of each variation. We know what are the pluses and minuses of each ingredient into the strategy. So in future when we want to tweak we have the info, when we start out we have all the info on various variations.

Secondly, when we test these variations, we may find that one of the or few of the variations are giving bad results, this should alert us, because in diff variations if the results are too far apart that would mean that your fav parameters may well be a case of accidental curve fitting.

Also when you test data we should look to test it on maximum number of data available and also should look to backtest it on some other instruments (with similar features in terms of volumes and price movements) which we are not looking to trade.

Also we should pick some random period way back in history to check if worked.

Also we should check it in diff market scenarios like pre 2008 perios and post 2008 period or in sideways markets of 2011 and 12.

We should leave no stone unturned after all its a matter of winning a and losing.

Regards

Amit