What is Freerisk?
Freerisk is a project with the goal of making freely available the data, algorithms and tools necessary to perform financial modeling. Although much important data is accessible from government agencies, it is neither well-integrated nor in a machine-readable format.
Why is it needed?
We believe that greater transparency and diversity of ideas is the key to allowing continued innovation in finance while reducing the risk of crises.
Traditional data and model providers require contracts and usually do not allow data to be republished in machine-readable formats. This limits the number of people who can build financial models and makes them virtually impossible to share. Since quantitative finance is a mixture of academics, professionals and enthusiasts, this lack of collaboration can hinder innovation.
So what are you building?
There are many components to Freerisk, some available today and others still in development. Among them are:
- A repository of original financial statements, annotated with normalized language, so they can be read by machines and verified by people
- A distributed modeling API, so that anyone can build, host and allow others to run a financial model
- A leaderboard system, which allows financial models to be easily backtested against arbitrary objective functions
- A system for user-generated content so people who read footnotes can add their own variables to statements
- A simple API for financial mashups
How do you hope this will be used?
The purpose of creating an open system is that it will be used in creative ways that you didn't anticipate. However, we have a few ideas:
- Models that calculate default risk better than the ratings agencies
- Better visualizations of financial data
- A variety of models to calculate earnings quality
- Methods for fraud detection in statements
- A way to test if guru-touted formulae really do "beat the market"
If you have more questions, you can contact us.