I would like to start a journal for developing my automated strategy, for accountability, progress review and external input. It is something quite different for me, usually I keep everything close to my chest, but I have been reading more about successful people (both traders and in business) and a common theme is sharing ideas & success. Those who were not afraid to share ideas often seem to generate more ideas. Obviously with an automated strategy at some point you need to keep your secrete herbs and spices to yourself, but I can worry about that as things progress.
So my automated strategy concepts are initially going to be based of Price Action concepts and strategies.
My expectation is not to out preform an experienced PA trader who can use discretion to see movements that would otherwise be missed by ridged rules of an automated strategy, but to win over time by adhering to those rules and allowing all trades to take place, hopefully allowing my perceived edge to come through.
With this, I believe a smaller edge can be capitalized on which would be more difficult for the discretionary trader as one mistake or broken rule can eat into profits.
So my primary reasons for creating an automated PA trading strategy;
To take all available trades during a specified time
Opportunity to run strategies across multiple instruments/markets simultaneously
First point is required for a strategy to truly reflect any edge believed to be present. Second point leads to the ability to diversify my risk and potentially smooth out my equity curve.
My planned development methodology will consist of;
Break down individual components of a strategy
Unit test each component on its own
Measure and learn from results
There are a few reasons for this approach;
Components will be shared across multiple strategies over time, each components needs to be trusted and proven.
Having a unit test for each component also allows future monitoring to ascertain if failure of a strategy is due to a component or the strategies edge
Improvements to an individual component can then also be distributed across all strategies.
So my next actionable step is to outline my first component with planned unit test. I hope to make each unit test a mini strategy that can be entered into the Battle of the Bots too.
Special thanks needs to go out to BigMike first for the forum, and for Kevin who I feel has shared a lot of great ideas and time into this section of the forum.