Has anyone created a successful automated trading system? What is your advice? - page 43

New comment
 

Totally depends on your preferred strategy.


You can always find inspiration from other people's current running strategy and try to craft something from there and change it to fit your style/preference. 

 
Lee See Hao #:

Totally depends on your preferred strategy.


You can always find inspiration from other people's current running strategy and try to craft something from there and change it to fit your style/preference. 

true
 

Yes and we too often, forget to have a look on scientific papers. Few years ago it was understandable but now with LLM : you can simply have an overview of the full litterature. No more excuse :)

It avoids to be trapped by over promising commercial EA and upskill immediately your level of trading...

We are in fantastic period for that !!! 

 
AdryDakodak #:

Yes and we too often, forget to have a look on scientific papers. Few years ago it was understandable but now with LLM : you can simply have an overview of the full litterature. No more excuse :)

It avoids to be trapped by over promising commercial EA and upskill immediately your level of trading...

We are in fantastic period for that !!! 

Just imagine having to buy a hardcopy book to start programming trading algo's. Yes, I'm that old.😂
 
True 😂 now you can scan papers, test ideas, and code in hours instead of weeks. No more excuse to fall for “magic” EAs. Still tho, info ≠ edge — most stuff is known, execution is what really matters.
 
Georgiy Merts #:

It can. For a while.

No system can work forever. Any system has periods of profit and periods of loss, and periods of loss, all other things being equal, give a greater outflow of capital than an inflow during periods of profit.

Yeah, that makes sense 👍

No system works in all conditions, and phases of loss are part of it. I think the key is not to avoid losses completely, but how the system handles those periods.

If drawdown is controlled and the system adapts a bit to changing conditions, it can still stay stable over time.

So maybe it’s less about “working forever” and more about how it behaves across different phases.

 
Vishal Verma #:

Yeah, that makes sense 👍

No system works in all conditions, and phases of loss are part of it. I think the key is not to avoid losses completely, but how the system handles those periods.

If drawdown is controlled and the system adapts a bit to changing conditions, it can still stay stable over time.

So maybe it’s less about “working forever” and more about how it behaves across different phases.

Thanks, I figured it out
 
JesonMile:
"automated trading" means different things to different people, here I mean a system that detects good trading situations and then places an order and closes later after some conditions are met.

It seems to be possible to create a profitable system, but so far I haven't had much luck. some big advantages I see in an automated system is that it can trade 24/7 and does not cause any emotion. ideally I would like to scalp on small timeframes (h1 or less), but h4 or d1 is probably easier, right?

what indicators have you had success with? are you using fixed sl and tp or more complicated closing conditions? did you write it completely yourself or use templates such as mt5 advisors? do you think many different indicators with certain weights are better than one or two or is it too complicated?

what kind of returns are you seeing and how consistent is your system?

can a breakout bot at market opening be successful?
I have one that i made but it does not trade 24/7 just during london new york overlap.
 
Lee See Hao #:

Totally depends on your preferred strategy.


You can always find inspiration from other people's current running strategy and try to craft something from there and change it to fit your style/preference. 

For me this is the best.

Watching a video of a manual traded strategy and then coding it and optimazing gives a totally new perspective of the strat.

 

I don't know if this has been descussed here, but starting from scrath, without any idea of what are you wanting to code is very stresfull.

I'll always suggest starting from a prexisting backtested strategy.

1...36373839404142434445
New comment