how system performance can improved

dell

Well-Known Member
#1
let there is a system ,
we backtested it by bar to bar manually at eod ,
how can we tweak it for better performance ......
tweak and backtest , tweak and backtest over 15 yrs data .........................
can senior's can describe more ideas to test .........................plz

some steps that i think ,
1 see trend on daily , trade on 30 m in direction of trend
2 sl on trigger bar
3 5 points buffer point , means buy after trigger +5
4 adds on further confirmation
5 carry with 2 lots , exit one at fixed point

how to cut large loosing trades ?
how to get small loosers and let profit run for winners ...........
what other things to be kept in mind ................
 

jahan

Well-Known Member
#2
let there is a system ,
we backtested it by bar to bar manually at eod ,
how can we tweak it for better performance ......
tweak and backtest , tweak and backtest over 15 yrs data .........................
can senior's can describe more ideas to test .........................plz

some steps that i think ,
1 see trend on daily , trade on 30 m in direction of trend
2 sl on trigger bar
3 5 points buffer point , means buy after trigger +5
4 adds on further confirmation
5 carry with 2 lots , exit one at fixed point

how to cut large loosing trades ?
how to get small loosers and let profit run for winners ...........
what other things to be kept in mind ................

Hello,

see.....its depends on system and the person who is designed it.

Regards,
 

dell

Well-Known Member
#3
Hello,

see.....its depends on system and the person who is designed it.

Regards,
jahan sir , how to improve any system , what other steps r to be combined , so to robust it .....?

i am testing an system , result r good but in 15 yrs testing period , three times loss (constitutive 5 times ) , two times (cons.4 times)........
how to identify coming sessions in which theses constitutive losses r there ..........
 

johnnypareek

Well-Known Member
#4
jahan sir , how to improve any system , what other steps r to be combined , so to robust it .....?

i am testing an system , result r good but in 15 yrs testing period , three times loss (constitutive 5 times ) , two times (cons.4 times)........
how to identify coming sessions in which theses constitutive losses r there ..........
Actully it is not possible to comment without knowing method. I assume the ur sys is say buy on buy/cover and sell/short. if it is then use intinal stoploss and trailing stop loss and test. I guess it shud reduced some continued lose making trades.

johnny
 
#5
johny sir, i want hourly scan with 50ema cross the candle with rsi 70 above for buy and 50ema cross the candle with rsi 30 below for short with signals pls help me for this afl
 

jahan

Well-Known Member
#6
jahan sir , how to improve any system , what other steps r to be combined , so to robust it .....?

i am testing an system , result r good but in 15 yrs testing period , three times loss (constitutive 5 times ) , two times (cons.4 times)........
how to identify coming sessions in which theses constitutive losses r there ..........
Hello,

I think ur tweaking too much.....the statistics ur given for consecutive losses is common....infact there is no surprise if u get more consecutive losses as u stated.

i don't know what's wrong with taking consecutive losses....

the only way to get rid of these...is u need to wait until 5 or 4 consecutive failed trades happen on chart ...then u can initiate on 6th trade...this is the only way,without knowing ur strategy.in this case in entire 15 years u have traded 3+2 five times only.In this scenario also there is one thing if u go after 5 trades then u have only 3 trades...if u go after 4 trades then u end with loss bcoz 5 consecutive losses are 3 times.

so waiting for failed trades also no use....u can't get 100/100 buddy.Trading is not Academics Business.where people think about 95/100 or 90/100 or 99/100 or even they go for 100/100.

according to ur statistics the system is awesome....and if u can't able bare those losses...then i feel that ur looking for something miracle.(which can't exit's).

if u further tweak ur system...there is chance of damaging the entire system.

presently iam trading with the system having 2% chances of 13 consecutive losses in manual backtesting....and in the series i have already gone through 7 losses...and i yet to face above 13 if this happen i reevaluate my strategy.

see manual back-testing and system back-testing is totally different.

Regards,
 
Last edited:

Similar threads