Lesson 3: How to Build Graybox Strategies

Lesson 3


There are two basic principles that define successful automated strategies and once we understand these principles or aspects then we can tune an automated strategy into a graybox or turn a graybox method into an automated system.

The forms are:

  1. Generalization and specialization. This has to do with how specific the pattern it looks at. A more generalized method might work across different charts or look to identify general classes of patterns versus specific patterns.
  2. Trade selectivity. This typically corresponds to trade filters to avoid taking mediocre or bad trades or getting locked into trading against powerful trends.

You can push an automated strategy into a graybox by increasing generalization factors and decreasing specialization factors. You can likewise increase trade selectivity to turn a graybox into an automated strategy or decrease trade selectivity to turn an automated strategy into a graybox.

You can switch a graybox into an automated strategy in real-time by creating buttons in BeyondBot that changes your trade selectivity degree or filters. This can be very powerful because your discretion might be able to create a "better filter". For example, you might decrease your strategy selectivity when using Restrict To Price Regions. This will enable to you to be more likely to get a signal when you have identified an area of opportunity.

There are basically, also, two types of strategy entries you will have:

  1. Edge based strategies or setups. These are more selective and will trade less frequently. Confidence in edge is higher. These strategies can be run fully automated. You may price restrict, change tier size, or restrict to times to influence with your discretion.
  2. Priors & invariants. The goal with these sort of entries is to be able to trade with high confidence when you have a good understanding of market environment. It is similar to ability if you had to ask Google-- if I know the market is in a downtrend or likely to be lower in 20-50 minutes, what's the optimal way to trade it? You are not trading an "edge" so to speak but trying to execute in an optimal way based on your read. This is critical because how many times have you as a trader or known a trader who was always right about the direction but lost money. The goal is to develop your proficiency with specific entries in the market replay and thus translate that to the live.