Stop Loss and Profit Target Functions Problem

Questions about MultiCharts and user contributed studies.
cael
Posts: 89
Joined: 06 Sep 2006
Has thanked: 2 times

Stop Loss and Profit Target Functions Problem

Postby cael » 28 Feb 2008

I had a surprising problem with these functions. They work as intended of commission and slippage are zero. But if they are non-zero, there are actually added to the Amount parameter, which actually changes the strategy. :? I don't remember TS working this way before.

User avatar
Andrew Kirillov
Posts: 1589
Joined: 28 Jul 2005
Has thanked: 2 times
Been thanked: 31 times
Contact:

Postby Andrew Kirillov » 29 Feb 2008

TS doesn't take into account commission/slippage and we emulated it initially. Although we were sure it is wrong (irrational) implementation. Some guys contacted us and complained that it is silly and we decided to take slippage/commission to account. So yes it works differently.
What do you feel about it?

sylfvdk
Posts: 57
Joined: 06 Dec 2005

Postby sylfvdk » 29 Feb 2008

Wow,

This is defenetely WRONG implementation. You are changing the strategy. If I want to sell @ $10.00 with limot order - you are selling ~ @ 10.10 on backtesting to adjust to slipage?!

I would say, that is wrong. Please, find another way to make correction. For example, on Strategy Performance Report or somewhere else.

Please, DO NOT CHANGE STRATEGY. Those are our numbers.

Thanks
Victor

User avatar
Andrew Kirillov
Posts: 1589
Joined: 28 Jul 2005
Has thanked: 2 times
Been thanked: 31 times
Contact:

Postby Andrew Kirillov » 03 Mar 2008

Victor,
I disagree. If you set $10 profit target, you want to have $10 in you pocket and you shouldn't care about slippage and commission. Otherwise your strategy exit with “profit”, but actually you have a loss. It doesn’t make sense for me and other users.
If you disagree with our implementation simply use 0 in slippage/commissions settings.
We are not going to change it.

kevin kolodzy
Posts: 64
Joined: 25 Jan 2008

Postby kevin kolodzy » 03 Mar 2008

Andrew, I think that makes sense, but it does mean code and settings ported directly from TS need tweaking to adjust for the difference. If it is known that this is the case, it can be handled.

Are slippage and commissions handled the same way in MCFX for these exits?

User avatar
danilo
Posts: 170
Joined: 02 Jan 2007
Location: Italy
Has thanked: 4 times
Been thanked: 9 times

Postby danilo » 03 Mar 2008

The implementation of slippage in MC is GOOD! is the only way to simulate the spread ask/bid. When there is a trade on the market you don't know if was matched a bid or ask price. If you are a buyer and the last trade was a bid execution at $10 and the spread ask/bid is 0.10$ when you send the order at the market will be executed at $10.10. If you don't like this implementation you can always set the slippage at zero. In a lot of trading books you can read about the slippage, but also a lot of books forget to explain way there is the slippage.

cael
Posts: 89
Joined: 06 Sep 2006
Has thanked: 2 times

Postby cael » 04 Mar 2008

Andrew

Victor is right. You are actually changing the simulation. For example, when I buy I set a profit target of X and a stop loss of Y. I optimize this and come back with say X = 1 and Y = 1.

But if I run with slipage and commission of say 10 cents, my effective X and Y have become 1.1 and 1.1 respectively. This is not what the trader intended.

Commission and Slippage are numbers that help stress a system to help simulate performance under more realistic trading conditions. It should not change the intended system :!:

User avatar
Andrew Kirillov
Posts: 1589
Joined: 28 Jul 2005
Has thanked: 2 times
Been thanked: 31 times
Contact:

Postby Andrew Kirillov » 04 Mar 2008

Andrew, I think that makes sense, but it does mean code and settings ported directly from TS need tweaking to adjust for the difference. If it is known that this is the case, it can be handled.

Are slippage and commissions handled the same way in MCFX for these exits?
Kevin,
It is hard to answer your question, because it will force me to ask our software configuration manager and find out what MCFX version are you running. I suggest you to run a simple test and see. It is easier for both of us!

User avatar
Andrew Kirillov
Posts: 1589
Joined: 28 Jul 2005
Has thanked: 2 times
Been thanked: 31 times
Contact:

Postby Andrew Kirillov » 04 Mar 2008

Andrew

Victor is right. You are actually changing the simulation. For example, when I buy I set a profit target of X and a stop loss of Y. I optimize this and come back with say X = 1 and Y = 1.

But if I run with slipage and commission of say 10 cents, my effective X and Y have become 1.1 and 1.1 respectively. This is not what the trader intended.

Commission and Slippage are numbers that help stress a system to help simulate performance under more realistic trading conditions. It should not change the intended system :!:
I don't think this dispute makes sense since you can indicate and take into account commissions/slippage in back testing process.:)

cael
Posts: 89
Joined: 06 Sep 2006
Has thanked: 2 times

Postby cael » 04 Mar 2008

Andrew

But when I optimize comm and slip is not taken into account.

User avatar
Andrew Kirillov
Posts: 1589
Joined: 28 Jul 2005
Has thanked: 2 times
Been thanked: 31 times
Contact:

Postby Andrew Kirillov » 05 Mar 2008

back-testing and optimization is the same thing. So slippage and commission is taken into account.


Return to “MultiCharts”