Mission & needs

The quantitative researcher’s mission is to design investment algorithms and combine them within model portfolios.

Portfolio construction requires working from two angles:

checkbox

Performance driver identification, and

checkbox

The understanding and controlling of the risks of the portfolio as well as the correlations between the constituent strategies or assets

Designing investment models requires:
checkbox
Backtesting:

i.e., faithfully reproducing the historical behavior of portfolios by integrating management constraints and avoiding common pitfalls: look-ahead bias, survivor bias in the indices, underestimation of transaction costs and market impacts

checkbox
Easy access

Easy access and manipulation of any type of data relevant for building portfolios

checkbox
Quick testing

Quick testing of any ideas or intuitions

What StarQube offers

  • checkbox
    A NoSQL database, optimized for calculations, structured around a unique and natively “point-in-time” repository
  • checkbox
    A simplified language for handling data and building risk models or “optimization” objects by simply setting parameters
  • checkbox
    A backtest engine
  • checkbox
    Exceptional calculation speed, thanks to the combination of the NoSQL database (calculation orders of magnitude faster than a relational base) and the conic optimizer
  • checkbox
    A series of APIs that can extend the scope of possibilities (Python, Matlab, Excel, C, C++, .NET, REST, Java)
  • checkbox
    Permissions and version-control tools to share backtests with other employees while minimizing operational risks

Recommended package