Discussion of article "The algorithm of ticks’ generation within the strategy tester of the MetaTrader 5 terminal" - page 18
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
Have you read the article we are talking about here?
It describes the method you mentioned in the MetaTrader 3 history.
The first strategy tester appeared in the MetaTrader 3 client terminal. It was a relatively simple tester by modern standards, in it the testing was performed according to three models of price development in a bar:
Four price model - the price successively passed the price Open, Low, High and Close for a bullish candle, for a bearish candle Open, High, Low and Close;
"Every 1 point" model - the 3-5-3 wave model is used, where the price successively passes three-wave, five-wave and three-wave again in 1 point increments;
In my pictures it is the other way round (the price pullback is much bigger):
Four price model - the price successively passed Open, High, Low and Close for a bullish candle, for a bearish candle Open, Low, High and Close;
In my pictures it is the opposite (the price pullback is many times bigger):
Four price model - price consistently passed Open, High, Low and Close for a bullish candle, for a bearish candle Open, Low, High and Close;
... maybe you will do it after all?
If we implement the system of tick generation proposed by you, there will be two orders of magnitude more grails for MT5. IMHO.
Here is a link to one of them: https://www.mql5.com/ru/code/244
Good afternoon,
I can't find the answer to the question on replacing tick history inMT5 tester :
"The developers have fundamentally blocked any alternative to generate ticks from the
minute (OHLC) or is it still possible to take historical data (e.g. from
http://ratedata.gaincapital.com/ ),
convert them from CSV format to HST format (e.g., with the help of
https://www.mql5.com/ru/code/8658 ) and write them to the appropriate history folder of the MT5 terminal?".
Will the tester still try to generate new ticks from the data in the replaced file or use it without conversion?
Maybe someone has already tried another algorithm (the above mentioned script https://www.mql5.com/ru/code/8658 for MT4, is there a similar one for MT5)?
It would be great if when you switch on the "broken line chart" mode, on a one-minute time frame, you could see a pseudo-type chart generated in the way described in the article, not just a linear interpolation of clones as it is now.
If the tick generation system you suggested is implemented, there will be two orders of magnitude more grails for MT5. IMHO.
Here is a link to one of them: https://www.mql5.com/ru/code/244
To implement? I did not propose to implement it, but only to add an option:
If you don't want to test on loaded (collected) ticks, please introduce as an option generation of ticks by the following algorithm for more plausibility (OHLC).
If there are more than 4 ticks, the price rollback always = High-Low, i.e. the maximum movement:
With the current algorithm, when testing on history, the price cannot move as I have indicated - rollback within a bar = 100% of possible rollback.
When you test your strategy on the history and it will be satisfactory for you, put it on the real, you will start (probably) exactly such pullbacks within one bar(pullback within a bar = 100% possible pullback), because there is no tick history and there is no possibility to test on tick history.
Accordingly - you will lose and you will not prove anything to anyone (because bars will be the same, but ticks are not recorded).
And if you add this option - it will be immediately visible during the test (at least on the history that comes with MT5). that your strategy does not work.
And whether there will be more or less grails for MT5 is absolutely not important IMHO.
I think they tweaked the phrase "the price can go as I indicated".
I repeat.
With the existing algorithm, in the process of testing on the history, the price can go as I indicated - a pullback inside the bar ~ 100% of the possible pullback.
Here for example for comparison - real ticks and those generated by the tester.
Please provide tick flows in tabular (xls, csv) form.
In such delicate matters, you can't operate with screens from which you can't understand anything. You also need a full description of testing conditions and settings.