An overview for understanding how models and strategies are built on Boosted Insights.
Table of Content
Models vs Strategies
Models are machine-learning algorithms trained to learn patterns from data and generate signals used for stock ratings, recommendations, and other insights. Having a performant model is the first step in getting started on Boosted Insights.
Once you have a model, you can then optimize the signals and set additional constraints to create Strategies. A model can power multiple strategies where stock ratings, recommendations, and other insights may differ from one another depending on their settings.
The model creation process is when we apply our capital expertise with the power of AI. Here, we define the machine learning settings, stock universe, and model inputs.
- Machine Learning Settings - These are the settings that define the parameter of how the machine learns and make predictions, which include Model Type, Goal Selection, Backtest Range, and more.
- Stock Universe - Here, you build your stock universe that the machine will analyze and rank.
- Model Inputs - Here, you select and modify the variables you want the machine to learn and make predictions using. To learn more, see our article on variables and features.
To learn more, see our guide on model creation.
Strategy creation is where you can configure your model signals to fit your requirements by setting additional optimizers, trading rules, and other constraints.
- Define Trading Rules - Here, you can set the criteria for what stocks the machine can trade (ie. market cap, long/short positions, etc). This will reflect the machine's trades, allocation, and holdings.
- Adjust Factor Exposures - You can maximize or minimize different factors to receive stocks with exposures that align with your requirements.
- Signal Optimization - You can set additional goals, exposure levels, and other criteria for model signals (ie. turnover optimization, sector/industry neutralization, etc).
Above are just some common ways we customize strategies, but there are so many more options. To learn more, see our article on strategy settings.
If you are an institutional client, your Customer Success Manager can work with you on model and strategy creation.