How to build Trading system?

how to find a system?
keep what you like, and if there is nothing that you like, search, and if you don't find it, don't trade, invest to managed account

read biology , how animal surfive, how they live, how they evolve their body to fit their surrounding, they are equiped with the most efficient tool by billion years of evolution.

actually any book will do, all can be metaphore to trading, you just need to filter it

Here is a good thread by Merlin:
http://www.forexfactory.com/showthread.php?t=4446

1. How did you test it?
I use Tradestation Omega Research Pro Suite

2. Which indicators are you using and how did you choose them?
Indicators are overrated, they can only tell you what is on the chart and even then, they lag. Not to say indicators are useless, but being more proficient with indicators will not make you a better trader.

3. What kind of result should I look for? For example I know that if I use MA whole range numbers should be used and give similar results (not over optimize)
Create the system based on logic and eyeballing charts and equity curves etc.

LOGIC: WHY does the system work?
CHARTS: If you use a 10ema as a trend indicator, what will that look like on the chart?
EQUITY CURVES: Is your system profitable because of one or two giant profits which have been curvefitted to be included in the system?

Finetune later on by optimizing if you want.

This leads to the question, how do I know if my optimised parameters are robust? Here is a simple way:

The optimised results should resemble a bell curve. Here is an example:

http://img513.imageshack.us/img513/6...llcurvejy6.jpg

4. What is a descent period of time that the system should be tested upon?
Here is a guide:
*At least 50 trades
*An out of sample set. Here you would like it to yield similar results.
*4 years. (You would not want to test a system on 1 year alone, only to find out later that every second year might be a loser)
5. Ones you started to trade with your system how did you find it? What problems did you encounter?
That's an easy one....reality! Dont lose faith in the system because reality is the only thing that you can't test for. Change, adapt and improve, but stickto the rules.

NOTE: By the way the system tested above is simple Elders' Triple Screen Trading System adapted by Nick Radge from The Chartist as such I give him credit for it.

The book "Trade Your Way to Financial Freedom" by Van Tharp has been helpful to many in learning some important concepts on how to build a system.

I found fairly good system using pivot points. What bother me is the 1 to 4 short/long ratio.
Analyzing 2 years back of the almost any stock market index around the world will not be so realistic due to the bull market we are in.

Is a system that produce this kind of results can still be reliable? should I get data of the market before 2003? will results that base on this data will mean something?? or maybe the market has changed since then?

again, thanks for the replays on this thread it helped me a lot!
Avihu

Forex like fishing. Easy & difficult at same time!
Best way:
Watch good fishermen. Imitate. Start slow (micro lots help).
Sometimes you get lucky, sometimes you hit snag, loose hook & bait. If too risky, slip & fall in water. Maybe sink, maybe swim.
Practice. Make stupid jokes on FF members area to blow off steam. Practice more.
Moral of story. It's better to learn fishing rather than buy fish. Fish slippery & hard to hold long enough for profit.

Very astute observation and similie.
Btw. there is a Comedy Club in the Members Forum which may be the place to vent humorously. If you just want to VENT... Open a new thread in the Members Forum. It will probably become the largest thread in FF.

detailed plan? I would be interested in any plan that is profitable on a consistent basis. if its art , its really not a plan.!.
I would have my own ea built in 1 mn if it uses indicators!

I'm Sorry, but I have to say this...

By all means, read as much as you can about this Sport of Fx. BUT Do Not assume that the last {or any} book You read has any connotation to the would be "Grail". That just incites authors to write more for profit.


This Game is as wide as the Planet and permutated by the population. No-one has "The Answer/Grail". But some may have what works "Most of the time"

That being said... I paid > $150.00 for the last book I bought on this subject.

the holly grail of trading is selling books and ea's about trading.!
Unless you trade very big timeframes of course or very tight scalping eas!

"I found fairly good system using pivot points. What bother me is the 1 to 4 short/long ratio."

There are 2 things to consider here:

1. Are the long entry / exit signal for the system identical to the short entry / exit signals?

If yes, then that is good.

If no, then you are probably curve-fitting.


2. Are the following performance measures for the longs similar to those of the shorts?

Win$:Loss$ ratio
Winning % of trades
Average Profit
Average Loss

If yes then that is good.

If no, then there is either a flaw in your code, or you are testing a period of data with a directional bias. You will need to test an out of sample set of data.


If you have answered yes to both the above questions, then you have either discovered an unusual market anomaly, or there is an error somewhere.

Money Management

Make computer do whole work for you building a system in 30 minutes or less:
Freeware - Forex Strategy Builder:
http://forexsb.com/overview.html

I too am attempting to create a trading system. Since Im a full time student as well I too need to make a system that works for intraday trading, mainly on the M15 chart. Basically I just try out different indicators on a demo account and see how often they work for me and I try to have a good combination of 2 or 3 indicators.

Trading is an art. If you want to make your trading career successful, you need to make a good and effective plan. A good plan can make you profitable in this market.

learn how to manage risk ---- THAT'S THE PLAN