Is MT5 strategy tester better than MT4?

New comment
 

what are your thoughts?

Is it more accurate?

I personally like the 1 min OLHC testing. It really speed up testing. 

 
The cloud testing is great!
 
doshur:

what are your thoughts?

Is it more accurate?

I personally like the 1 min OLHC testing. It really speed up testing. 

For me - better, because MQL5 language is faster itself, and more detailed report of MT5 tester. Cloud is cool, it is cheap and sometimes it is just enough to join the home computers only to speed up calculations with no need to join to global cloud. In my opinion MQL5 is faster then MQL4 for ca. 5-10 times, plus the use of 4 agents (2 my computers), so it increases calculations up to 20-40 times.
 
doshur:

what are your thoughts?

Is it more accurate?

As far as I can see the user has no control over the data that is used so repeatable tests can not be guaranteed,  there is no possibility of using tick data.  Is it faster ?  for a single run or for optimizations ?
 
Sure, it is better. Multicurrency strategies can be tested.
 

It is certainly slower for a single run,  by how much ?  maybe a factor of 4 . . .  I've just converted a very simple EA so I can do a back to back comparison.

Testing in non-visual mode for the same date range of the same pair  ( I can't use the same data as I know no way of getting the data into MT5 )  MT4 was approx 3 seconds  MT5 was 14 seconds 

 

I ran the same test on MT4 and MT5 this time in visual mode . . .  it took some time to complete.

 

MT4  -  16 minutes

MT5  -  4 hrs 44 mins (terminated early, I couldn't wait any longer)

MT4 ST  is at least 17 times faster than MT5 ST

 

this is a simple EA, it was tested on EURUSD H1  3rd Oct 2012 to 27th Feb 2013 that is approx. 5 months.  Does anyone know why the MT5 Strategy Tester is so slow ? 

 
RaptorUK:

I ran the same test on MT4 and MT5 this time in visual mode . . .  it took some time to complete.

 

MT4  -  16 minutes

MT5  -  4 hrs 44 mins (terminated early, I couldn't wait any longer)

MT4 ST  is at least 17 times faster than MT5 ST

 

this is a simple EA, it was tested on EURUSD H1  3rd Oct 2012 to 27th Feb 2013 that is approx. 5 months.  Does anyone know why the MT5 Strategy Tester is so slow ? 

That is strange. My EA backtest quite fast. All my EA are very simple.

Did you check your coding? You might be looping some conditions? 

 
RaptorUK:

I ran the same test on MT4 and MT5 this time in visual mode . . .  it took some time to complete.

 

MT4  -  16 minutes

MT5  -  4 hrs 44 mins (terminated early, I couldn't wait any longer)

MT4 ST  is at least 17 times faster than MT5 ST

 

this is a simple EA, it was tested on EURUSD H1  3rd Oct 2012 to 27th Feb 2013 that is approx. 5 months.  Does anyone know why the MT5 Strategy Tester is so slow ? 

You can not compare tester by time in visualization mode.

And you know it.

 
doshur:

That is strange. My EA backtest quite fast. All my EA are very simple.

Did you check your coding? You might be looping some conditions? 

I checked my code,  it does this for every tick . . .

if(PositionsTotal() == 0)

 if there isn't an open position it opens one with TP and SL set, if there is a position open it does nothing.

 
Renat:

You can not compare tester by time in visualization mode.

And you know it.

I'm a little confused by your response . . .  I have only used the MT5 Strategy Tester for the first time this week so I'm not sure why I would know very much about the MT5 ST.

In MT4 I verified the operation of my EAs by viewing the trades taken,  I do this by looking at the screen captures taken by the EA for each trade placed,  I can only do this using Visual mode,  why is it wrong that I should expect to be able to do the same with MT5 ? 

12
New comment