Beta version of MetaTrader 4 IDE including new MQL4 compiler and editor - page 19
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
Debatable.
But what kind of TS should be, that on ticks - it would give a stable profit, and on generation of ticks - stably lost? In my opinion, generation of ticks in MT5 is quite adequate, and if TS is losing on generation - it will lose on real account...
Of course, it's an amateurish opinion, I've never tried to trade bellow M15, and now I'm inclined to use D1 and even not to look lower than on H1... But, of course, I want to understand, what am I missing on ticks that is "cut off" during generation?
It doesn't really matter what size period to trade on - in real life everyone will trade on ticks, not bars.
And with other news/statistics/interventions/manipulations with sharp jumps in quotations the situation will be similar.
Take a look at the last screenshot https://www.mql5.com/ru/forum/1031/page19#comment_597854
This candlestick is on GBPUSD D1, ticks in the tester and in real on screenshots.
All stop/market orders in real will slide a lot, unlike in a tester, and other scenarios are possible https://www.mql5.com/ru/forum/1031/page18#comment_520781
The best IDE for C/C++, MQL4/MQL5 is Microsoft Visual Studio with all necessary plug-ins http://ru.wikipedia.org/wiki/Сравнение_IDE#C.2FC.2B.2B
https://www.mql5.com/ru/forum/13846/page2#comment_597651
Here is an example of what you are missing:
Within 2 seconds, Ask < Bid by more than 25 pips. And this situation is repeated 15 times! Do you get this when generating ticks?
It's as simple as that. There are fewer ticks generated in the four.
It's sad rather than simple. Here is the performance of a single run of the standard MovingAverage EA on M1 in OHLC mode:
In this mode and timeframe the number of ticks in both platforms should be the same, but even in this mode 64-bit MetaTrader5 lags 2.5 times behind its younger 32-bit brother.
Admittedly, both use multi-core CPUs. But the price for the modular architecture of the MT5 tester is too high.
It's sad rather than simple. Here is the performance of a single run of the standard MovingAverage EA on M1 in OHLC mode:
In this mode and timeframe the number of ticks in both platforms must be equal, but even in this mode the 64-bit MetaTrader5 lags 2,5 times behind its younger 32-bit brother.
And what about the numerous accusations that we even "cannot reproduce the number of ticks in 4"? Wanted? You got it.
I understand that you personally didn't want it. But you personally (and no one else) didn't stand up for the existing generation of ticks in the foursome
Use M1 OHLC - quite an adequate model. And faster than every tick in four. And if you're pipsing, every tick in five is much better than in four.
What about the numerous accusations that we can't even "reproduce the number of ticks in a foursome correctly"? Did you want it? You got it.
I understand that you personally didn't want this. But you personally (and nobody else) did not stand up for the existing generation of ticks in 4
Forgive my ignorance, but it seemed to me that on M1 in OHLC mode the number of ticks in MT4 and MT5 is strictly equal, because in both cases 4 points are taken (extract from the Basics of testing in MetaTrader5):
1 minute OHLC
Testing in All ticks mode is the most accurate of the three modes, but also the slowest. The OnTick() handler is run on every tick, and the tick volume can be quite large. For strategies that don't care about the tick sequence of price development during a bar, there is a faster and coarser simulation mode - "1 minute OHLC".
In the "1 minute OHLC" mode, the tick sequence is built using only OHLC prices of one-minute bars, the number of generated check points is significantly reduced - therefore the testing time is reduced. The OnTick() function is run on all control points that are built using OHLC prices of one-minute bars.
Forgive my ignorance, but it seemed to me that at M1 in OHLC mode the number of ticks in МТ4 and МТ5 is strictly equal, because 4 points are taken in both cases (extract from the Basics of testing in MetaTrader5 article):
1 minute OHLC
Testing in the "All ticks" mode is the most accurate of the three modes, but also the slowest. The OnTick() handler is run on every tick, and the tick volume can be quite large. For strategies that don't care about the tick sequence of price development during a bar, there is a faster and coarser simulation mode - "1 minute OHLC".
In the "1 minute OHLC" mode, the tick sequence is built using only OHLC prices of one-minute bars, the number of generated check points is significantly reduced - therefore the testing time is reduced. The OnTick() function is run on all control points that are built using OHLC prices of one-minute bars.
Yes, absolutely right. On M1 in OHLC mode the number of ticks coincides in both five and four
Well, the test time does not coincide, moreover, in the most dramatic way for MT5.
It's incorrect to use" StandardMoving Average " as an argument for tester comparison. The four and five examples of Moving Average Expert Advisor are differently designed (the fifth one is much more complex). To compare correctly the testers' speeds, you should use strictly the same structure of Expert Advisors. Of course, you should compare them. Do you have any ideas?
// Yes, I know, I once gave the same example for comparison, but I remember the difference was ten times or more, then the difference decreased.
It is not right to use "Standard Moving Average" as an argument for comparing testers. The four and five examples of Moving Average Expert Advisor are arranged differently (the fifth one is much trickier). To compare the tester speeds correctly, you must use Expert Advisors having strictly the same structure. Any ideas?
// I'm sorry, I once gave the same example for comparison, but then I remember the difference was ten times or more, and then the difference decreased.
What ideas do you need? Isn't it enough that the optimized MAexp in MT5 is slower than the non-optimized MAexp in MT4?
Well, count pi (Matemat posted the code) it will not be inlaid.
Within 2 seconds Ask < Bid by more than 25 pips. And this situation is repeated 15 ! times. Do you get this when generating ticks?
Hmmm !!!
Actually, it's strange... Not to mention, as already said - who would allow it to hold a position for 2 seconds... Well, let's assume there are such brokers, I wonder if there are brokers among them, who will allow you to make a significant profit on such a negative spread ?
If there are such brokers - it really turns out that we are missing in generation... And you could quite reasonably ask the developer for a list of brokers which allow you to work with negative spreads.