Alpha9 EA *** Released to TSD Elite Members *** - page 96

 
newdigital:
Alpha is multi timeframe EA (MTF EA using many timeframes for the direction of the trend). So, any backtesting can not be reliable.

I don't know what you mean by this. Multi-timeframe EA can most certainly be backtested successfully. I have done so many many times before this one.

In my forward testing, I see nearly 100% correlation with my backtest results (true in multi-timeframes as well). In general I don't understand why people disparage backtesting so much.

Plus, if we really want to see how Alpha9 does across many pairs, I could work on getting it converted to MT5. That would allow us to test everything simultaneously.

 
silverpike:
I don't know what you mean by this. Multi-timeframe EA can most certainly be backtested successfully. I have done so many many times before this one.

Technically - yes. But concerning results - sometimes yes, sometimes not.

For example, to open the trade - EA is checking few indicators for open bar compare with same indicators (values) for close bar fopr W1 timeframe, D1 timeframe, H4 timeframe and so on. And you are backtesting on close bar (every tick method, right?) ... and how many timeframes are backtested by Metatrader? one timeframe only ... so ...

There are few types of EAs which backtesting can not be reliable:

- EAs coded on open bar together with close bar (Fireburd and some others);

- EAs using high/low of the bar in some calculation (MaChannel and some others);

- any MTF EAs;

and some more types.

As to trading many pairs simultaniously so we are backtesting 1 pair only for 1 timeframe ... and EA is forward traded many pairs taking onto consideration many timeframes simultaniously.

That is why I am forward testing some EAs for many month.

It is not news about backtesting - it is well-known information about backtesting and the types of the EAs.

 

newdigital, in my experience the differences caused by the higher TF hi/low errors are not significant enough to cause large differences in results. It is true that they are not the same. I suppose each person can make his own choice about what satisfies. I choose to rely on backtesting as a strong indicator for performance.

Anyways, Alpha9 is a very intriguing EA. I think I might try and convert it to MT5, because I really want to see what a multi-pair backtest looks like. I will have to start by converting the indicators first, so one step at a time.

I have to admit though: I'm disappointed more pairs don't show good performance in backtesting besides the EUR. I will keep testing and see if more can be found.

 

You can look at my forward testing results here https://www.mql5.com/en/forum/173403 (excel files as last post of this thread). As I remember - the best version is dll version and there are some more pairs which are profitable.

 

Hi I would like to try your EA but it wont let me shows that I need a library file does anyone know where I can get this file

Thanks

 
tropheus:
Hi I would like to try your EA but it wont let me shows that I need a library file does anyone know where I can get this file Thanks

Hi Tropheus,

Try here

https://www.mql5.com/en/forum/178132

post#819

 

somehow the EA does not change the lotsize in its backtesting. even though I changed it to 1.0, it keeps sticking on the 0.2 lotsize.

Any suggestions?

 

Hi downcrusher,

Change MMType to 0.

Explanation is on this post https://www.mql5.com/en/forum/178132/page2 and on this page:

https://www.mql5.com/en/forum/178132

 

Alpha9v1.21 DLL EA.

It is the performance for all trading pairs all together:

AUDUSD:

CHFJPY:

EURAUD:

GBPCHF:

 

Alpha9v1.21 DLL EA.

It is the performance for GBPUSD:

USDCAD:

USDCHF:

I am trading many pairs using one copy of metatrader:

Reason: