My Blog

The Dogwood Report posted a sample mechanical trading system:

Go long if price is above the MA(50) and RSI(2) is less than 5

Exit when RSI(2) is above 70 or an 8 percent stop loss is hit

Go short if price is below the MA(50) and RSI(2) is greater than 95

Cover when RSI(2) is less than 30 or an 8 percent stop loss is hit

The results of this system were quite impressive.

Take a look at their results here:http://1.bp.blogspot.com/_-YWCgDK_Px4/SWQvNwoLXvI/AAAAAAAAC7U/DxV3aq5DZ2A/s1600-h/RSI(2)+Long+%26+Short.png

The one negative of this system is the number of trades this system creates is high. With a fixed cost broker like most people have, those can really cut into your profits.
I believe that this system is impressive and follows the philosophy that simpler is better.
It could probably be further refined to reduce the number of trades that it makes.

Play around with it yourself and see if you like it.
The original post is at:http://thedogwoodreport.blogspot.com/2009/01/wealth-lab-developer-51-experiment.html

Tradesignal: Strategy and Optimization screen displayed

I have started looking into developing a mechanical trading system and as with all things I try, I am going at it with a passion akin to that of Al Pacino in Scarface.
The task seems slightly in-surmountable because I have tried and failed to develop something successful in the past. In retrospect, I feel that I may have taken the wrong approach.

During my previous attempts, I was unaware of software such as Tradesignal. Tools such as this one allow anyone to develop custom trading algorithms with ease.
My approach was more of a do-it-yourself methodology and I was trying to re-invent the wheel.
I had once programed Bolliginer bands and moving averages into excel files using end of day data that I downloaded from  Yahoo Finance, only to realize that my plan of programing individual indicators then searching for trends then coming up with profitable combinations of them is a task with no end.
I tried to re-invent the wheel but my wheel turned out square.

This time, I will be successful. I am not going to waste time programming the math behind indicators. I will however, learn their use and research how others have succeed in using them in the past.
A program like Tradestation does the hard work for you. The indicators are pre-programmed, the back-testing algorithms are pre-programmed, the graphs look good, the interface is nice, etc etc etc. All in all, it is really a good program.

Your role in developing your mechanical trading system is simply to pick the indicators you want, use the Tradestation optimizer to select appropriate parameters for those indicators and test it.

Now that I have made it sound as easy as taking candy from a baby, here is the honest truth, it is hard. How hard? Very hard. Ok, maybe not that hard – but it’s not easy.
There are hundreds of indicators to choose from, millions of combinations to create and if you don’t really know what you’re doing, you may as well go back to programming indicators in excel.

Here is the approach that I am taking:

1. Learn a few common technical indicators and how they work
2. Backtest the use of those indicators
3. Learn a few complex indicators
4. Think up logical combinations of the indicators I know and test them in Tradesignal
5. Read about others who have had success in mechanical trading
6. Repeat 3 – 5

I am currently at step 2 and I am hoping to be ready to test some of my “strategies” in the next few months.
I will post my results on this blog so stay tuned.

Leave a comment if you have dabbled in mechincal trading and have any tips.