Newbie questions into EA - cracking head

[Kelvin Chen]

Hello,

I am a newbie in the work on EA developing, i was wondering on some key questions on the world of EA trading;

there are 3 points i keep cracking my head on;;


Whats best 1)chart 2)timeframe 3)currency pairs and 4)Broker to run Expert Advisors on ? Is there a best ?

Indicator(s) to be used in a EA that has highest success rate without frequent re-optimization. ? is there ?

Main reasons why EA’s fail and whats the average lifespan of expert advisers


Thanks everyone who read and help ^^

 

[AsstModerator]

Broker suggestions belong in the Company Comparisons and Competitions folder.

 

[ilearn2t]

Hello Kelvin Chen

Welcome to the forum, nice to hear from someone that is willing to venture behind the scenes of trading, and look into the world of "programming".

Most if not all your questions can only be answered by you yourself, depending on how you intend to program your EA (Expert Advisor).

I myself use "Pending orders" for my EA's, so there is no need for "indicators", but if you prefer to add indicators to your project, you could study previous data from the past to try and predict the swing/change that all indicators have over a given amount of time.

If or when I feel the need to improve my EA's, I think my main factor behind improving my EA's might be better MM (Money Management), so coding in "ratio" would be my way of going.

As for "Brokers" avoid one's that will effect your EA.

Example:

If you program a quick open-close EA, avoid brokers who don't allow "scalping".
If you program one that is centered around economic data avoid brokers who don't allow news trading.

The few examples above can also play apart in your timeframe + chart + currency pair choice, depending on how you yourself code/program the EA.

Best thing about EA's lately is the introduction of the new MQ4 platform, were EA programmers must now have full/sole control over their own EA's and hopefully bring about the end of some commercial EA's out there.

Good luck
ilearn2t

 

[Pharaoh]

If there was a single best pair and time frame that could be traded with no re-optimization, this place would be called the Forex Millionaire's Army.

First, develop and test out a mechanical system until it shows promise. Then, try it on higher and lower time frames as well as on different pairs. When you find a promising candidate, optimize it - but don't OVER optimize it.

Over-optimization involves fitting all the parameters on backtests to try to achive perfect or near perfect results. There once was an EA sold that bragged about 10 years without a single loss. Within 1 week of deployment, it took a huge loss. The developer said it was a one-time fluke, until it quickly took more losses.

One of the things done to achieve that perfect backtest was setting the SL to the exact pip needed to remove all losing trades that eventually ended with a small profit.

This is easily avoided. Start out your development with reasonable risk reward and win loss ratio targets and don't be lured into aiming for 90%+ wins in exchange for a large SL and small TP.

 

[LoRio]

Take any strategy and optimize it on a specific spread size on a specific broker on a specific timeframe on a specific time.
And trade it the same way..Just check for conditions to be met , make sure its the same pair the same broker and have a spread checking part to allow / disallow trade execution.

 

[jotekfinance]

I agree with Pharoah. Just remember, 51% win ratio is beating the house (and making money) 60% and you have a career