Is this trading system worth taking live??

mechtrader

Well-Known Member
#21
mechtrader

It is completely meaningless to provide a back testing statistics of a strategy and ask question about its viability.

Anyone who has programmed few strategies will tell you that it is relatively simple task to create a strategy that produces almost picture perfect statistics and jaw dropping profit curves.

It is perfectly possible to write a strategy that never ever loses on any trade during back testing in any language. Only when a strategy is actually put into the market in real time trading, losses will start appearing ...
Couldnot agree more with your point.
Before commenting on the results i presume that all precautions are being taken while developing the strategy.
This is again true that building a strategy with optimizing and data snooping bias is a different ball game than exploiting a real edge.
Hope everyone realizes this.

MT
 

copypasteaee

Humbled by Markets
#24

revendar

Well-Known Member
#26
I'm inspired by MechTrader after reading through his thread. I have been working on a simple trend following method and have posted the backtest results for past 1 year. My setup involves the following.

1. Trading Intraday stocks in 15M TF. Selective stocks.
2. Based on candlestick,MA and few indicators.
3. Considering brokerage per trade as 100. Kept it higher considering slippages.
4. Trading Equity with no margin

I want to know about the following.

1. How much deviation one can see taking a backtested system live and what are practical implications.
2. What are the factors should I consider in addition to my system?
3. @MT: If you dont mind can you post your EQ curve so I can get an idea of the deviation from live system.



 
#27
Hi Revendar,

All the best..I am also inspired by the success of MT. And started walking along the road. Let's see if I get become successful or not..
 

mechtrader

Well-Known Member
#28
I'm inspired by MechTrader after reading through his thread. I have been working on a simple trend following method and have posted the backtest results for past 1 year. My setup involves the following.

1. Trading Intraday stocks in 15M TF. Selective stocks.
2. Based on candlestick,MA and few indicators.
3. Considering brokerage per trade as 100. Kept it higher considering slippages.
4. Trading Equity with no margin

I want to know about the following.

1. How much deviation one can see taking a backtested system live and what are practical implications.
2. What are the factors should I consider in addition to my system?
3. @MT: If you dont mind can you post your EQ curve so I can get an idea of the deviation from live system.



Ok... thats good work. Keep few points in mind.

1) Always backtest 3 years and 1 year forward test.
2) slippages and commissions are 500 per trade minimum now days in stock futures. Cant say about nifty.
3) There should be no deviation if the systems are tested properly exploiting a significant edge.
4) Money management and risk capacity is very important after successful backtests.
5) My systems are very mediocre in front of 93% win rate. I manage only 40-45%. make sure you are not forward looking.

MT
 

tradingstudent

Well-Known Member
#29
I'm inspired by MechTrader after reading through his thread. I have been working on a simple trend following method and have posted the backtest results for past 1 year. My setup involves the following.

1. Trading Intraday stocks in 15M TF. Selective stocks.
2. Based on candlestick,MA and few indicators.
3. Considering brokerage per trade as 100. Kept it higher considering slippages.
4. Trading Equity with no margin

I want to know about the following.

1. How much deviation one can see taking a backtested system live and what are practical implications.
2. What are the factors should I consider in addition to my system?
3. @MT: If you dont mind can you post your EQ curve so I can get an idea of the deviation from live system.



Ravender,

142 winning trades in a row. Am i reading this correct?
 

vijkris

Learner and Follower
#30
I'm inspired by MechTrader after reading through his thread. I have been working on a simple trend following method and have posted the backtest results for past 1 year. My setup involves the following.

1. Trading Intraday stocks in 15M TF. Selective stocks.
2. Based on candlestick,MA and few indicators.
3. Considering brokerage per trade as 100. Kept it higher considering slippages.
4. Trading Equity with no margin

I want to know about the following.

1. How much deviation one can see taking a backtested system live and what are practical implications.
2. What are the factors should I consider in addition to my system?
3. @MT: If you dont mind can you post your EQ curve so I can get an idea of the deviation from live system.



hi, if u can do "code check and profile" in afl editor under tools option , if it says that formula looks into future, then real results wil not be 93% win.
if it says formula is good, then u have holy grail.
enjoy:thumb:
 

Similar threads