Is this trading system worth taking live??

#1
I have created a trading system and I wanted to know from the other algo traders on the community what they think of the results and what are the odds of it holding true in a real trading environment.

For the trades I have assumed that they are executed on the open of the next candle and stop losses are hit intraday and commissions and slippages all inclusive are assumed at 0.035% per trade.

The attached results are on bank nifty for nifty I get a CAGR of 18% (for the same period) and a max drawdown of 11.5%.

Would really appreciate if some people could share their inputs on the motecarlo simulation and their defination of a robust system
 

Attachments

Last edited:

mechtrader

Well-Known Member
#2
As per these results, it is definitely good.

There are alot of other pitfalls to avoid.
==> Any errors in testing or in code.
==> Over fitting of parameters.
==> OOS results.
==> Real time execution may be different.

Apart from these this system is tradeable.

Thanks
MT
 

jahan

Well-Known Member
#3
I have created a trading system and I wanted to know from the other algo traders on the community what they think of the results and what are the odds of it holding true in a real trading environment.

For the trades I have assumed that they are executed on the open of the next candle and stop losses are hit intraday and commissions and slippages all inclusive are assumed at 0.35% per trade.

The attached results are on bank nifty for nifty I get a CAGR of 18% (for the same period) and a max drawdown of 11.5%.

Would really appreciate if some people could share their inputs on the mot
ecarlo simulation and their defination of a robust system
Hello,

Assuming your system is intraday type about 130- 140 trades per month ,if ur system holds the trades more than a day and ur exit strategy is based on intraday(live),then u have chance to experience catastrophic losses due to gap up or gap down opening.
Below 30% per annum didn't worth the effort we put in(according to me).

Regards,
 

bunti_k23

Well-Known Member
#4
rohit can u explain me how to perform monte carlo in amibroker..

thank you:)
 
#5
I have created a trading system and I wanted to know from the other algo traders on the community what they think of the results and what are the odds of it holding true in a real trading environment.

For the trades I have assumed that they are executed on the open of the next candle and stop losses are hit intraday and commissions and slippages all inclusive are assumed at 0.35% per trade.

The attached results are on bank nifty for nifty I get a CAGR of 18% (for the same period) and a max drawdown of 11.5%.

Would really appreciate if some people could share their inputs on the motecarlo simulation and their defination of a robust system
Your trading system looks awesome..your equity curve is almost linear..I am guessing some kind of trend following method but how the F you carried out montecarlo simulation in ami. If possible can you start a new thread for explaining how to carry out montecarlo simulation in ami.

regards mutual guru
 
#7
Hello,

Assuming your system is intraday type about 130- 140 trades per month ,if ur system holds the trades more than a day and ur exit strategy is based on intraday(live),then u have chance to experience catastrophic losses due to gap up or gap down opening.
Below 30% per annum didn't worth the effort we put in(according to me).

Regards,
The system is a trend following system and at times holds positions for days together. You can see in the backtesting stats that the total number of trades executed over five and a half years are 776 so on an average around 120 - 150 trades per year.

The risk of gap up or gap down is definately there and the possibility of being caught on the wrong side of a black swan event is also present but I guess that is part and parcel of all trend following strategies.
 
#9
As per these results, it is definitely good.

There are alot of other pitfalls to avoid.
==> Any errors in testing or in code.
==> Over fitting of parameters.
==> OOS results.
==> Real time execution may be different.

Apart from these this system is tradeable.

Thanks
MT
Hi MT can you elaborate on what are the things that can go wrong in real time trade or what factors in real time can lead to different results than backtesting of data for the same month beyond a normal margin (which I assume to be 0.01% per trade)
 
Last edited:

mindgames

Well-Known Member
#10
Hi,

This is a little off topic, my apologies in advance.

For the past couple of years I have been using free charts available on the web (Google, ET, Stockcharts etc.). Next, I'd like to buy some trading s/w next so that I can backtest ideas.

Can you advise me on Amibroker - costs, onetime/recurring charges, ease of testing (I'm not a techie) etc.

While I have gone through their website, it is still quite confusing. Is there any contact no. in India to get additional details?

I'll be using it only for EOD trades..

Any help would be great!

Thank you.
 

Similar threads