Hello
First of all: You have a great tool, I am now working a few weeks with it and learning. Always again surprised what I can do with NT!
Nevertheless NT performance statistics has 2 shortcomings in my point of view:
1. Already discussed in this forum: If entries per direction is more than one, then the cumulated % profit is wrong resp. misleading. An example:
1.2.2010 Go long with 100.- / 31.12.2010 Exit with 200.- (Profit: 100.-)
2.2.2010 Go long with 100.- / 31.12.2010 Exit with 200.- (Profit: 100.-)
3.2.2010 Go long with 100.- / 31.12.2010 Exit with 200.- (Profit: 100.-)
As you can see, we have hier Entries per direction = 3. NT will calculate as cumulated % profit 300%. But this is wrong, since my invested capital obviously is not 100.- but 300.-. The total profit is therefore 100%.
2. The max drawdown is somewhat misleading (If i understood it correct?).
Example:
Max consecutive down is 10%, then my strategy wins 1%, then again down 10%, then up 1%, then again down 10%, then up 1%, then again down 10%, then up 1%, then again down 10%.
NT will calculate a max drawdown of 10%. But in the above example I will have lost almost 50% of my capital. Quite an important factor when I decide to use leverage or not.
I would like to implement therefore a max drawdown defined as follows:
Max difference between a high and a low.
Therefore I would like to add some calculations in backtesting. But how is this done best?
1. As I understand from the forum it is not possible to recognize whether a strategy has finished backtestet. Is it possible to retrieve the date range the user has defined when testing? So I could check with the date, as you suggest somewhere in the forum when it ends
2. Is it possible to access the executions data? I have seen I can access the trade data, but in order to find out what is max capital needed from my account, I would need the exectutions
3. Is there an easy way to access in the code the statistics calculated in the performance tab and print to file?
thanks & regards
Andre
First of all: You have a great tool, I am now working a few weeks with it and learning. Always again surprised what I can do with NT!
Nevertheless NT performance statistics has 2 shortcomings in my point of view:
1. Already discussed in this forum: If entries per direction is more than one, then the cumulated % profit is wrong resp. misleading. An example:
1.2.2010 Go long with 100.- / 31.12.2010 Exit with 200.- (Profit: 100.-)
2.2.2010 Go long with 100.- / 31.12.2010 Exit with 200.- (Profit: 100.-)
3.2.2010 Go long with 100.- / 31.12.2010 Exit with 200.- (Profit: 100.-)
As you can see, we have hier Entries per direction = 3. NT will calculate as cumulated % profit 300%. But this is wrong, since my invested capital obviously is not 100.- but 300.-. The total profit is therefore 100%.
2. The max drawdown is somewhat misleading (If i understood it correct?).
Example:
Max consecutive down is 10%, then my strategy wins 1%, then again down 10%, then up 1%, then again down 10%, then up 1%, then again down 10%, then up 1%, then again down 10%.
NT will calculate a max drawdown of 10%. But in the above example I will have lost almost 50% of my capital. Quite an important factor when I decide to use leverage or not.
I would like to implement therefore a max drawdown defined as follows:
Max difference between a high and a low.
Therefore I would like to add some calculations in backtesting. But how is this done best?
1. As I understand from the forum it is not possible to recognize whether a strategy has finished backtestet. Is it possible to retrieve the date range the user has defined when testing? So I could check with the date, as you suggest somewhere in the forum when it ends
2. Is it possible to access the executions data? I have seen I can access the trade data, but in order to find out what is max capital needed from my account, I would need the exectutions
3. Is there an easy way to access in the code the statistics calculated in the performance tab and print to file?
thanks & regards
Andre
Comment