Meeting your OBJECTIVES.

Discussions about the testing and simulation of mechanical trading systems using historical data and other methods. Trading Blox Customers should post Trading Blox specific questions in the Customer Support forum.
Post Reply
Dutchtrader
Roundtable Fellow
Roundtable Fellow
Posts: 58
Joined: Wed Apr 30, 2003 4:35 pm
Location: Netherlands

Meeting your OBJECTIVES.

Post by Dutchtrader » Sun May 04, 2003 6:37 pm

Hi all readers ( yes you too who still didn't register)

I started the new subject "Meeting your objectives" because I couldn't find so much attention about this subject. I see so many text how to test, what system is good, better, best. All kind of dynamic en creative MM rules. And almost everytime I see that all kind methodologies are in some way good, better, best profitable. But it seems if the researchers are not satisfied enough, making more adjustments, new techniques in MM rules and programming. Although I think that it is very good to test and improve your system, I miss their "startingpoint" for testing and simulation. What is their goal? What is their mission? One time I saw some objectives of c.f. ( for the new ones> he is a tradingexpert and original turtle ) in his post: about the drawdown, ratio's etc. For me it is much more clearer if a trader tells his objectives ( I don't have to know your capital and other private things ) When you ( and the reader ) know your objectives, it is, in my opinion, much easier to test and improve your system(s). Let me tell you my objectives as clear as possible:


1. Trading 2 different long term trend following systems.
2. Trading high liquid future markets.
3. 25% average annual return on capital.
4. max. drawdown of 30%.
5. Surviving every year one worst case scenario on one of the future markets.
6. Max. drawdown of 20% of STARTING equity. ( I hope you understand the difference between 4 and 6 )
7. Do you suggest more that I was forgotten?

So, where to start? I don't have a system. I don't have Trading Recipes or Behold. I can't programming ( VBA, C++ ).

Is there a systematic/general approach? I hope many will response. Not to solve my current objectives, because everybody have to "profit" from the advice from the forum. I am very curious, I want to learn and I am open minded. I hope that also people response when they recognize the problem for themselves or don't agree at all with my vision.

I'm not writing in my motherlanguage so I'm sorry if it is not clear enough. If so, please tell me!

Thanks in advance!

Marc

edward kim
Roundtable Knight
Roundtable Knight
Posts: 344
Joined: Sun Apr 20, 2003 2:42 pm
Location: Silicon Valley / San Jose, CA USA
Contact:

Post by edward kim » Mon May 05, 2003 3:51 am

Hi Marc,

I don't know what other people have done, but I'll share what I did when I first started

1. Decide what kind of system you want (for me it was medium to long-term trend following)

2. Go back and look at all the successful trades that I've had that were medium to long-term and evaluate what made them successful. I also went back and evaluated all the "good" losses (good meaning I had a reasonable process in determining why and how to get out of the trade, and hopefully at a minimal loss.)

3. Translated all of my thoughts and decisions into actual statistics and parameters. For example, I started buying Amgen early in 1991 and kept holding on to it because it started moving up. It was moving sideways during Desert Shield (late 1990), and started breaking out after that. I liked that kind of price action, so I just programmed my system to track stocks that broke out of 60 day highs and kept a 4xATR stop in. I didn't know what ATR was at the time, but I must have intuitively known somehow - most of the stops that I have put in when I did put in stops just happened to be about 4xATR.

4. I kept doing the first three items with paper and pencil to get a feel for the "artwork". I'm not very good at programming immediately from the top of my head - if I write down everything on pieces of scratch paper, I'll eventually have a stack of notes that I can start working from. Not writing down all my thoughts and some numbers will make the whole process seem overwhelming, while programming straight off the top might keep my thinking process one-track-minded.

5. Choose a platform you like. Since I was just starting, I used Excel. In the future, I am trying to figure out which platform I can use because Excel has it's limitations; however, I will still use Excel because the coding and rules are algorithmically pure in the sense that it would be very easy to port that into any other platform. If I went from Tradestation EasyLanguage to something like C++, I think I would have a much harder time making the conversion. When I look at my Excel sheet, I can see how the quote data gets converted to base statistics, and how these get converted into parameters. It's easy to see the transformation, and it gives you an idea of how your system is working.

6. Repeat any of the above steps as much as you can to improve your system.

Edward

Post Reply