I'm looking for a (scientific) case study about efficiency of different Technical Indicators?

New comment
 

Hey everyone,

 

I'm looking for a case study or paper about the efficiency of different Technical Indicators with different parameters on different charts. I googled it, but couldn't find anything useful. I would be glad about everything that comes close.

 

Thank you! 

 
Weedjo:

Hey everyone,

 I'm looking for a case study or paper about the efficiency of different Technical Indicators with different parameters on different charts. I googled it, but couldn't find anything useful. I would be glad about everything that comes close.

Efficiency of what ?  the coding ?  the use of system resources ?  you need to be more specific.
 
Weedjo:

Hey everyone,

 

I'm looking for a case study or paper about the efficiency of different Technical Indicators with different parameters on different charts. I googled it, but couldn't find anything useful. I would be glad about everything that comes close.

 

Thank you! 

One billion dollar request..

To me, as interesting as senseless..

Senseless because: you just use MT5 technologies (coding, visual testing, optimization, cloud, etc.) to make and discover yourself your own "case study" about such "efficiency".

Beside this, there is countless of material in this forum, that you have just the "embarrassing of the choice", to what you can choose your own favourite stuff, and follow/study/expand it's knowledge and experimentation.

Just reading the Articles you can find almost of all what you need..

 
RaptorUK:
Efficiency of what ?  the coding ?  the use of system resources ?  you need to be more specific.

Well, of course the performance for using it for trendidentifacation and/or using it for a buy&sell signal. 

 

 @Harvester: 

But isn't there an objective comparative study of the common indicators with parameters? 

Documentation on MQL5: Standard Constants, Enumerations and Structures / Objects Constants / Object Types
Documentation on MQL5: Standard Constants, Enumerations and Structures / Objects Constants / Object Types
  • www.mql5.com
Standard Constants, Enumerations and Structures / Objects Constants / Object Types - Documentation on MQL5
 
Weedjo:

Well, of course the performance for using it for trendidentifacation and/or using it for a buy&sell signal. 

 

 @Harvester: 

But isn't there an objective comparative study of the common indicators with parameters? 

Well this is something that i am working on right now.

To some degree, i have to agree with Harvester, but some kind of success measurement in necessary, although it might not be completely objective. 

Idea is to take a period like one year, and to run with strategy tester complete set of parameters, with SAR strategy. After that, you can compare the best, average and median result of complete test for profit & relative drawback.

The problems so far are:
1) How much sense does this have?  Well if you are checking how good an indicator is in detecting a trend, this should be satisfying
2) money management: If you run an indicator with a fixed lot (no money management) and it shows like 1000 pips captured in year, this doesn't mean nothing, because if you apply any kind of money management, the result might turn to loss, although without MM it has won 1000 pips. There are three general types of money management, and applying a different type may turn the result completely. So, how to deal with this?

 
Weedjo: I'm looking for a case study or paper about the efficiency of different Technical Indicators with different parameters on different charts. I googled it, but couldn't find anything useful. I would be glad about everything that comes close.

You didn't find it because such study will show all indicators ending up with negative efficiency. You can run this test yourself, read the descriptions of the standard-indicators and how their creator used them. Use the indicator's default values like number_of_bars, standard_dev etc. Choose a timeframe like hourly, a stoploss of about 20 and takeprofit of about 20. Run them on EurUsd since year 2000 til now.

Now if you want to turn any indicator from a negative-efficiency into a positive, all you've gotta do is parameter optimize (aka curve-fit it). Start looking for different number-of-bars, sd, timeframes, stoploss, takeprofit ... etc ... etc. And bingo any indicator can now be profitable within a back-test. Obviously, those parameters wouldn't work going into the future.

Some people think the answer is re-optimizing every so often. Some people think the answer is combining x with y with z indicators. You'll have to perform your own tests and draw your own conclusions.

 
Ubzen:

 You'll have to perform your own tests and draw your own conclusions.

In other words  (or acronym) . . .  DYOR
New comment