EA Question for smarties - page 2

 

Shinigami

yeah everything u said sounds right. and i will tell u that im no expert in programming or trading. keeping in mind what u have said about different market conditions. its true but do u really think there is no way around this? there always is a way. even if we never find it. thats the reason i want to program this because i want to test it and learn and come up with different ideas or possible solutions. and one idea that could possibly be a work around to the obvious problems u mentioned is to trade at a higher time frame and possibly one trade a day and optimize in an even higher time frame.

i dont believe that the market moves in random ways for no reasons. there is a reason for everything. cause and effect. and in higher time frames there is a transition period it doesnt just change at a snap of a finger like a 1 min chart.

 

Archimedes

i think im thinking about it differently than u guys. i think its alot similar. its almost the same thing exept i think im optimizing at such a high time frame. and im having very few trades. it becomes totally different. yes the market changes, but think of that change as a VERY HIGH RESOLUTION AND SLOW MOTION CHANGE, even if its a quick one in real life market conditions.

and through this super high slow motion resolution i will land in a soft spot place an order. Then run the whole program once again.

 

alright let me get my information all together, so i can make more sense to u guys. ill post it here.

 

I think I understand what you are trying to say but please use "edit" button.

Now, about optimizing. Try to do a manual backtest of your theory.

First, take the EA you want optimized. Then start at 2006.01.01. Optimize two or three or how many weeks you want. Run the expert on a week or a day or the period you want. Then run another optimization. Use your new settings to run another week. You can save results of optimization in xls file for further calculations. If there is a trade that is left running, write a function that prevents the EA from opening new orders after the period you use, only trail that order and close it eventually. Add those results to xls file.

In order to be able to summarize it, you will have to use 0.1 lots. This way you'll be able to make a system that makes profit in pips. If the only way to make profit is money management, then the risk is great. If it makes profits in pips, its much safer.

After you do this optimization-running-optimization thing for a year (or at least 8 months to see how this goes with different markets) you'll be able to tell if there is any reason behind this.

I'm writing all of this because I was totally obsessed with this idea in the past. I ended up totally disappointed - settings that allow me to trade for the last year end up losing money nowadays. Settings that make good profits for last 3 months lose the whole deposit in the next month. The system needs to be flexible (robust) and not curve-fitted. I just don't want other people wasting time making yet another bicycle.

 

im gona take it ez fo now

 

elitecamper take a break from trading, think what else you can do for a living. If you still want to, come again and give it another try. Not everyone is born to be a trader, not everyone is born to be a president. Find what you like and stick to it. Maybe you won't be richer than Bill Gates but if you are happy, its much better that way.

 

Just a suggestion, your tag line 'Money management is a way to manage money, not make them!' is grammatically incorrect, it should read 'Money management is a way to manage money, not make it!'. The fact that you are completely incorrect in this assertion is a argument for another time.

 

ok im going to stick to trading and programing cuz thats what i like.

 

For anyone else intrested in walk-forward optimization here is an mqh that you can program into your EAs.

also check these links

AltaVista - Babel Fish Translation

http:// articles.mql4.com/ru/336

you need to run two terminals to make this happen.

Reason: