Most brokers have a limit of MaxLot=100 (Roboforex=10,000?)
and during testing we have that martin deals are opened with the volume of the broker's MaxLot
I suggest making the parameter "Max 1st deal Lot ".
According to Martin's settings, everyone will be able to calculate it for themselves.
And with a test for a long distance, there will be a fairer result.
"Brainstorming" - it can be useful :)
For Community readers, I will duplicate some of the thoughts from the MQL chat
***
It's just when you optimize a lot of parameters in the tester: if you get a day with a drawdown, then: If the balance is 1999, the drawdown, for example, is 40%, and with a balance of 1000, it will already be 80%!
Here is the minus of the lot Lot/1000... In gradation....
%Balance or %Margin will be fairer.
***
*****
YOU WRITE: On the contrary - optimize on a fixed lot, and then stress tests.
*****
Look.
Testing for a long distance
Strategy without withdrawal of funds, with an increase in the lot (For example, Lot per 1000)
We start with $1000 and 0,01 Lot.
At a conditional moment
Fixlot: Balance 5.000
Lot/1000: Balance 600.000
A position is open
Fixlot: 0,01 Lot
Lot/1000: 6 lots (=600.000/1000*0,01 )
The price went to a loss of 1000 pips. Nally: The price of 100 pips at 0.01 Lot = 1$;
Loss of 1000 pips at 0.01 Lot = 10$;
Loss of 1000 pips at 6 Lot = 6.000$
Drawdown:
Fixlot: 0,2% (10/5.000)
Lot/1000: 1% (6.000/600.000)
The difference is 5 times.
"Brainstorming" - it can be useful :)
For Community readers, I will duplicate some of the thoughts from the MQL chat
***
It's just when you optimize a lot of parameters in the tester: if you get a day with a drawdown, then: If the balance is 1999, the drawdown, for example, is 40%, and with a balance of 1000, it will already be 80%!
Here is the minus of the lot Lot/1000... In gradation....
%Balance or %Margin will be fairer.
***
*****
YOU WRITE: On the contrary - optimize on a fixed lot, and then stress tests.
*****
Look.
Testing for a long distance
Strategy without withdrawal of funds, with an increase in the lot (For example, Lot per 1000)
We start with $1000 and 0,01 Lot.
At a conditional moment
Fixlot: Balance 5.000
Lot/1000: Balance 600.000
A position is open
Fixlot: 0,01 Lot
Lot/1000: 6 lots (=600.000/1000*0,01 )
The price went to a loss of 1000 pips. Nally: The price of 100 pips at 0.01 Lot = 1$;
Loss of 1000 pips at 0.01 Lot = 10$;
Loss of 1000 pips at 6 Lot = 6.000$
Drawdown:
Fixlot: 0,2% (10/5.000)
Lot/1000: 1% (6.000/600.000)
The difference is 5 times.
There is an error in your calculations.
6 lots per 600 000 is equal to 0.05 lots per 5 000. And DD will be the same.
There is no point in optimizations with reinvesting. You'll just get confused with zeros and lose useful info.
There is an error in your calculations.
6 lots per 600 000 is equal to 0.05 lots per 5 000. And DD will be the same.
There is no point in optimizations with reinvesting. You'll just get confused with zeros and lose useful info.
I gave an example Where There is a Fixed lot and Lot/1000. When testing at a long distance (when Balance increases) , the difference in MaxDD is noticeable!!! The table above.... It is important for me to find out large drawdowns throughout the testing area so that I can analyze them visually later!!
There is no point in optimizations with reinvesting.
You can find largest DDs with the fixed start lot. It will be even easier.
MaxLot was designed not for this purpose. It just limits the max used lot (for "Risk in %" mode, for example).
Can you guy change max lot for any deal to max lot for first deal. Because if apply max lot for any deal will work wrong with martingale.
Example: