Hi Xray Sir
I have read the whole thread once again.Though I have read your thread regularly as soon as you posted.But this time I have read it again post by post & took notes.Its really a great stuff on MM,System making & risk management.Thanks a lot for this wonderful thread & a great stuff on trading aspects which we generally ignore.Its helpful to give me so many ideas on make my system rule based.
I have some query.I am doing backtesting.I have some query.
a) How many years data/ period is sufficient for intraday trading system? I.e. how much sample do we need to check that our system will work on every market condition?
b) How will I check effect of slippage on my system.I can check all parameters which you mentioned..i.e luck factor etc .But how to count slipage?
c) Shall I check all ratios & parameters on total result of each month or For total period of data.Lets us say, I am checking 5 year of data..so should I check luck factor for whole 5 years or shall I check it for every month?
I will put some more question as more testing will go on.Please bear with me
Thanks
P.S. I tried to Pm you But couldnt get through, even I am not able to see your profile.
I have read the whole thread once again.Though I have read your thread regularly as soon as you posted.But this time I have read it again post by post & took notes.Its really a great stuff on MM,System making & risk management.Thanks a lot for this wonderful thread & a great stuff on trading aspects which we generally ignore.Its helpful to give me so many ideas on make my system rule based.
I have some query.I am doing backtesting.I have some query.
a) How many years data/ period is sufficient for intraday trading system? I.e. how much sample do we need to check that our system will work on every market condition?
b) How will I check effect of slippage on my system.I can check all parameters which you mentioned..i.e luck factor etc .But how to count slipage?
c) Shall I check all ratios & parameters on total result of each month or For total period of data.Lets us say, I am checking 5 year of data..so should I check luck factor for whole 5 years or shall I check it for every month?
I will put some more question as more testing will go on.Please bear with me
Thanks
P.S. I tried to Pm you But couldnt get through, even I am not able to see your profile.
Ans for a
There is n0 specific Answer for how many years data is needed for back testing... it can be 5 years to 10 years ...what you want from back test is core point,large sample size is tested to know performance over n number of trades, since mechanical system strike rate is low..we check on large sample size,there is no system which will work for every market condition as per limited knowledge
Ans for b
Slippage is one of main factor for trading ,if instrument is liquid say NF or BNF... it will range from 8 to 32 points depending on quantity you trade...if you know depth based trading (Old way of trading in AP) or OFA mechanics you can observe your self can get to reasonable conclusion on slippage
Ans for c
Go for ratios for series wise ,back test should be on series data not on continue data..
Last edited: