What's necessary for a good system?

Crashh

Newbie
Messages
2
Likes
0
Hi ppl, I'm very new in here :p
I have a question for u, I searched all night, about trading systems, and I only found scams and nothing telling what I wanted to know, until I discovered this nice forum :D
I've done a lot of new indicators, and I test them with trading systems I do for them, and I do trading systems also. I have lots of them.
I would like to know something about testing trading systems:
What are the necessary performance data of a system in order for it to be good at your point of view?
- Do u test your systems with spread 0 or do u test them already with an discount of 2/3 pips?
This one is strange because why should we test them with a spread we don't know which will be in the future?
- Do u test systems with leverages around 50/1 and margin calls of 50% or with 1/1 leverage?
- What profits do u search in a trading system for ex, for Eur-Usd in 1 or 2 years in hourly candles?

We all know that a trading system much too overfitted to a cross, will give better results but worse results in the future, so there is no need for optimizing it in excess. But, when I develop a new indicator, I do a trading system based only in that indicator or with 1 more, and check how much he could do in those 2 years, if over-optimized.
This over-optimization is done for me just to measure how much the indicator can be "molded" into the cross because we can't have millionary results with only an indicator like bollinger bands and stochastic without something else, and even with something ele, over-optimized we can't have decent results.
Now, I got results in the thousands of millions value of % and more sometimes, in 2 years of hourly candle trades, with good ratios (I guess but I need u to tell me what's the standard accepted worldwide). So, wouldn't that single indicator, which we can take lots of profit of without the help of other famous ones, doing profit alone, be a good indicator to use on a daily basis in daytrading and not as an trading system? Or wouldn't it also be good in trading systems if he alone gives so much profit?
And in case it may be good, how can I made an indicator (and a bunch of them) like that public? Or would it have any chance to be famous? I would only be public something that ppl would like if I had a way to make it associated to me and not to anyone else (having my name on it).

Basically what I really need to know, is what values of performance does a system need to have, including what spread it will use on tests and leverage (I try to test them on 50/1 with 50% margin call to see if they adapt well).

PS: I'll show u at the end one example of one of my trading systems in action. It gave 600 trades in 2 years, only 9600 pips of profit (as u can see below) but with 50/1 leverage and 50% margin calls and no take profits he gave those orders in real time (but not the millions as u can imagine, but anyway worked as a good indicator as he worked without any other known indicator). U can see there the pips earned, the posision, on thouse hours on those 2 days, 3 wins 1 lost.

Best regards for u all,

Gonçalo Nuno

TradingSystemZZPCaldeirao.jpg
 
Non-pips system opinion

My opinion is only about trading indexes, not currencies but perhaps will be of use. Imo, to be consider a working system, it should be successful under rising, falling and trendless markets. Many systems work well in one but not the other. It should display more winners than losers, losing trades should be mild and drawdowns should also be mild.

Steve

QUOTE=Crashh]Hi ppl, I'm very new in here :p
I have a question for u, I searched all night, about trading systems, and I only found scams and nothing telling what I wanted to know, until I discovered this nice forum :D
I've done a lot of new indicators, and I test them with trading systems I do for them, and I do trading systems also. I have lots of them.
I would like to know something about testing trading systems:
What are the necessary performance data of a system in order for it to be good at your point of view?
- Do u test your systems with spread 0 or do u test them already with an discount of 2/3 pips?
This one is strange because why should we test them with a spread we don't know which will be in the future?
- Do u test systems with leverages around 50/1 and margin calls of 50% or with 1/1 leverage?
- What profits do u search in a trading system for ex, for Eur-Usd in 1 or 2 years in hourly candles?

We all know that a trading system much too overfitted to a cross, will give better results but worse results in the future, so there is no need for optimizing it in excess. But, when I develop a new indicator, I do a trading system based only in that indicator or with 1 more, and check how much he could do in those 2 years, if over-optimized.
This over-optimization is done for me just to measure how much the indicator can be "molded" into the cross because we can't have millionary results with only an indicator like bollinger bands and stochastic without something else, and even with something ele, over-optimized we can't have decent results.
Now, I got results in the thousands of millions value of % and more sometimes, in 2 years of hourly candle trades, with good ratios (I guess but I need u to tell me what's the standard accepted worldwide). So, wouldn't that single indicator, which we can take lots of profit of without the help of other famous ones, doing profit alone, be a good indicator to use on a daily basis in daytrading and not as an trading system? Or wouldn't it also be good in trading systems if he alone gives so much profit?
And in case it may be good, how can I made an indicator (and a bunch of them) like that public? Or would it have any chance to be famous? I would only be public something that ppl would like if I had a way to make it associated to me and not to anyone else (having my name on it).

Basically what I really need to know, is what values of performance does a system need to have, including what spread it will use on tests and leverage (I try to test them on 50/1 with 50% margin call to see if they adapt well).

PS: I'll show u at the end one example of one of my trading systems in action. It gave 600 trades in 2 years, only 9600 pips of profit (as u can see below) but with 50/1 leverage and 50% margin calls and no take profits he gave those orders in real time (but not the millions as u can imagine, but anyway worked as a good indicator as he worked without any other known indicator). U can see there the pips earned, the posision, on thouse hours on those 2 days, 3 wins 1 lost.

Best regards for u all,

Gonçalo Nuno

TradingSystemZZPCaldeirao.jpg
[/QUOTE]
 
Check one of my systems

Hi,

Thank you for your answer. I have many profitable trading systems, and meanwhile I discovered already how are the trading systems all over the world so I've put one of mine on the internet 2 months ago.
Here I leave you with one of them that will be online for one more month +-, I have it giving 100% mecanized trades in real time on that website from 18th July, and before that I was updating the website by hand.
It's on http://tradingsystem.pt.to. Sorry if part of it is not in english but the main site with the table is.
It gives a lot of profit as you can see, and it's working along since beggining of 2004 in real time, I have others with profit but still on testing.
I made that question before just to know how others would think about this question because I never met another person that developed trading systems.

Anyway I hope you like it while is still online, I can't have my machines wasting resources all the time so I'll put the website offline in one month +-.

PS: I know it says it makes millions % profit in 3 years at 50/1 but those leverages would be crazy to use them. I don't advice anyone trading on my system, it's there only for experimental purposes, and more than 10/1 (the leverage I used on the table on the website) is a bad option. The daily system makes profit even with a 80pips spread and the hourly with 12 pips spread, so it should handle the sllipages.

Here have a image of that system's performance in 3 years of horly candles:
varias.gif


Best regards,

Crashh
 
A system that wins around 65% of the time, even if the win size is only equal to the stop size will perform well. At about 50% winners, you wins need to average double the size of losers in order to deliver a steady linear like growth in equity.

Anything that delivers much under 50% of winners needs to be very good at picking large moves and holding onto them.

Basically anyting greater than 50% wins should see you make profits if the trade management is executed properly. Very high %winners for a mechanical trading system probably indicates overfit, but for a discretionary trader indicates mastery of their market.
 
Thanks for info

Be aware this website is very sensitive to anything which looks like someone is trying to sell a system here. I wish you luck with yours and be prepared for a legion of naysayers who say this can't be done. Ignore their static and keep going, being always your own worst critic and realist.

Crashh said:
Hi,

Thank you for your answer. I have many profitable trading systems, and meanwhile I discovered already how are the trading systems all over the world so I've put one of mine on the internet 2 months ago.
Here I leave you with one of them that will be online for one more month +-, I have it giving 100% mecanized trades in real time on that website from 18th July, and before that I was updating the website by hand.
It's on http://tradingsystem.pt.to. Sorry if part of it is not in english but the main site with the table is.
It gives a lot of profit as you can see, and it's working along since beggining of 2004 in real time, I have others with profit but still on testing.
I made that question before just to know how others would think about this question because I never met another person that developed trading systems.

Anyway I hope you like it while is still online, I can't have my machines wasting resources all the time so I'll put the website offline in one month +-.

PS: I know it says it makes millions % profit in 3 years at 50/1 but those leverages would be crazy to use them. I don't advice anyone trading on my system, it's there only for experimental purposes, and more than 10/1 (the leverage I used on the table on the website) is a bad option. The daily system makes profit even with a 80pips spread and the hourly with 12 pips spread, so it should handle the sllipages.

Here have a image of that system's performance in 3 years of horly candles:
varias.gif


Best regards,

Crashh
 
Top