Opinion on Backtest results

dell

Well-Known Member
#11
No my AFL does not looks into future. It uses a lagging indicator.

I tested BankNifty spot on 15min TF with a capital of 400000 using quotes from 1 Jan 2010 to 31 Dec 2013. Also I have included a transaction cost of 0.03% so you can say it is the actual profit that might come.
how many lots with have u tested ?
can u post ur trade tab snapshot of ur bactest report?
 
Last edited:

dell

Well-Known Member
#12
1) I do not know how to define lots. I did a normal backtesting.
2) Here is the HTML file. http://www.4shared .com/document/RjneVSokba/Backtest_Report_-_BNspot_-_1Ja.html
see here is problem .........
u have trade 42 , 66 ,155 etc no. of lots
so ur backtest report is wrong .......

if u want to correctly check than add this line in ur code :

setpositionsize(4,spsshares);

and in symbol information window :
in contract specification
roundlotsize=1
margin deposited =350
ticksize=0
pointvalue=25

than again backtest and share ur result
 
#13
1) I do not know how to define lots. I did a normal backtesting.
2) Here is the HTML file. http://www.4shared .com/document/RjneVSokba/Backtest_Report_-_BNspot_-_1Ja.html
Abhi,

Even if you backtest as per suggested by Dell, your results will be good. The drawdown number will be even lower than 14% which is vgood and other parameters will improved apart from net return. The winner % is also good.

Hiren
 
Last edited:
#16
Abhi,

This is a shorter period backtest. I have found that we must backtest on at least 1 years' data. The more the better as it accommodates different market conditions in our backtest period.

Hiren
 
#20
I can help you too in some coding afl but have to be good logic as i am making something for myself too may be we can exchange something and we can talk on some other source
email will be good :[email protected]
I want to talk to you about the strategy i am making
Might be good for your market too