Discussion of article "The algorithm of ticks’ generation within the strategy tester of the MetaTrader 5 terminal" - page 7
You are missing trading opportunities:
- Free trading apps
- Over 8,000 signals for copying
- Economic news for exploring financial markets
Registration
Log in
You agree to website policy and terms of use
If you do not have an account, please register
I disagree with this statement:
Used intelligently, the MT4 open prices only model is as accurate as the all ticks model.
The following approach will create an EA which backtests almost identically in the open prices only model.
This approach also tends to produce the most robust strategies, since forex prices approach completely random behaviour at tick level.
Paul
Thus, summing up to the sixth page (though the branch, of course, will go further:) we can say first of all to Prival'y:
Dear Prival, though you think on your wave, teak (no offence :), but it is clear that metaquots have chosen the most balanced
way of development.
In this variant.
there is much less traffic (which is important for all participants, both clients and servers in general!),
really, saves space on a screw (strange as it may seem, but useful),
ticks walk only inside minutes (very beautiful), and...
filtering, which doesn't give us exact ticks, no matter how it is done...
For Metaquotes, here's what we can say:
In the 'optimisation option' we can add (in time) such an item, like /'by high-lows, but take tick history from the file...<graal.tikitiki>'//.
And let the person who needs it have this history himself, either from his broker or from a neighbour's broker. I have to!...:)
And also, in essence.
Please do not forget about the concept.
Now we are talking only about ticks, but what about SPRED, STACK, finally?
Maybe we should think in advance about the situation when it will be possible (necessary!) to test strategies on the FUND?
I propose the same situation again, but with the proviso 'take the glass' from the file. It's more complicated, but you are in the development stage, right?
Perhaps I should have sent these messages to tech support, but here they are more on topic.
However. It's a pleasure to work with you.
Regards, Alexey.
P.S. You should have a rest...:).
Metaquotes position their platform as the most advanced platform for the development of mechanical trading systems and what they have done is both surprising and impressive, but... How can you create such a system if the conditions for testing the strategy and its real work are different? If in this case Tick is not a certain superfluous definition, but the very essence on the basis of which you enter or exit the market. In fact, the speech is about that it is impossible to trust the tested one.... And for diversion of eyes to talk about the bet, spread.....
Metaquotes position their platform as the most advanced platform for the development of mechanical trading systems, and the done and surprising and impressive, but ... How can you create such a system, if the conditions for testing the strategy and its real work are different? If in this case Tick - is not some superfluous definition, but just the very essence on the basis of which the entry into the market or exit from it. In fact, the speech is about that it is impossible to believe the tested one.... And for diversion of eyes to talk about the bet, spread.....
Why is it impossible to believe?
2010.09.17 07:28:17 Checking Holes (EURUSD,M1) Number of holes in history 49 for last 24 hours
2010.09.17 07:28:17 Checking Holes ( EURUSD,M1) Indicator started on EURUSD symbol 1 min.
2010.09.17 07:28:17 Hole Check (EURUSD,M1) Test Start Date 2010.09.16 05:16:00
2010.09.17 07:28:17 Hole Check (EURUSD,M1) Hole USDCAD 2010.09.17 05:21:00
2010.09.17 07:28:17 Hole Check (EURUSD,M1) Hole USDCAD 2010.09.17 05:02:00
2010.09.17 07:28:17 Hole Check (EURUSD,M1) Hole USDJPY 2010.09.17 04:06:00
2010.09.17 07:28:17 Hole Check (EURUSD,M1) Hole GBPUSD 2010.09.17 02:19:00
2010.09.17 07:28:17 Hole Check (EURUSD,M1) Hole GBPUSD 2010.09.17 01:47:00
2010.09.17 07:28:17 Hole Check (EURUSD,M1) Hole GBPUSD 2010.09.17 00:41:00
Why can't it be trusted?
2010.09.17 07:29:28 Spread Check (EURUSD,M1) Number of bars that have no spread 2540
2010.09.17 07:29:28 Spread Check (EURUSD,M1) For EURUSD symbol max=102 min=5 MOJ=17.92398711982532 RMS=2.720447767317922 MOJ+3*RMS=26.08533042177909 count=997502
2010.09.17 07:29:28 Spread Check (EURUSD,M1) Number of analysed bars 1000042
Prival, you are incorrigible - you take a common situation and try to convince others that it is a problem.
Skipping single minutes because there are no prices. It happens all the time at night.
The tester does not generate missing minutes.
I can't understand why we can't use the tick history received in the terminal in the tester? Why generate it. ?
I am also concerned about the absence of tick chart, why is it not connected?
I can't understand why we can't use the tick history received in the terminal in the tester? Why generate it. ?
I am also worried about the absence of tick chart, why is it not connected?
About the history
It has already been said several times that the server does NOT store tick history for reasons related including the large amount of information.
Nobody forbids you to collect ticks yourself and accumulate them (if discs allow it, even for all symbols for 10 years)....
Regarding the chart
In the market review there is a tick chart (not very big but there is). I am sure the developers do not plan to add any other tick charts (as the officially recognised minimum period is 1M).
About the story.
It has already been said several times that the server does NOT store the history for reasons related to the large amount of information.
Nobody forbids you to collect ticks yourself and accumulate them (if discs allow it, even for all symbols for 10 years)...
Regarding the schedule
In the market overview there is a tick chart (not very big but there is). I am sure the developers do not plan to add any other tick charts (as the officially recognised minimum period is 1M).
It is all clear, for those who test on the period of 10 years tick history is not necessary. It is not stored on the server, but it is accumulated in the client. And for me, for example, it is very important to test on weekly or even daily tick history! The starting strategy is based on the number of ticks per minute, their speed, etc.
A tick chart is useless in a market review, it cannot be expanded even on the full screen, let alone viewed for a day. And most importantly, if you could open 4 tick charts simultaneously on the desktop of the terminal for different currency pairs, then when trading manually, you can immediately visually see when you need to open, you can see the moment of the beginning of unidirectional movement of pairs.
Prival, you are incorrigible - you take a common situation and try to convince others that it is a problem.
Skipping single minutes because there are no prices. It happens all the time at night.
The tester does not generate missing minutes.
I'm afraid you're the one who can't fix it.
1. No price or no tick?. The price is there, but the way of building a bar (until a new tick comes, the bar is not built) leads to the fact that there are holes in the history. Although at this moment it is safe to enter the market. You can't enter the market in the tester. It is an artificial hole. Once again - there was a price, there was no tick (there was no price change). You are substituting these concepts.
2. Okay with holes, where did the spread go? The indicator is attached, anyone can run it and make sure that the history is still not all right.
For example, here is the statistics of the study of the quality of the story from 13.08.08, so in memory I left while I was in the ban.
13.08.2010 22:14
Spread check (AUDUSD,M1)
Number of bars with no spread 83323
13.08.2010 22:14
Spread Check (AUDUSD,M1)
For the symbol AUDUSD max=50 min=15 MOJ=24.42753579773531 SCO=5.172441632855508 MOJ+3*SCO=39.94486069630183 count=16691
13.08.2010 22:14
Spread Check (AUDUSD,M1)
Number of analysed bars 100014
Out of 100 thousand bars 83 thousand bars have no ASK.
Don't answer a bad question that the tester doesn't care what ASK was there, it just simulates it and takes the figure from the ceiling. For the tester has ASK at that moment, I checked it. But there is no ASK in the history of minutes....
H.Y. You can't treat the price like that. Here you can listen to what happens on the stock exchange when there is no price (no ask or bid) http://blogberg.ru/blog/11057.html and in the history of MT this situation is everywhere....