Best Practices for Algorithmic Trading & Risk Management

There is great value in creating a framework for facilitating dynamic combination of different models and algorithms for testing. The main challenge in creating such a system in the algorithmic trading world is maintaining an acceptable performance profile in the face of increased dynamic execution. The ability to do thorough back-testing is a necessity for all algorithmic trading. The core capabilities of such an infrastructure are tick generation and counterparty simulation, which must, by definition, be interdependent. Integrating proof of concepts and early versions of algorithms into the same environment as the real implementations enable modelers to make use of capabilities of the sandbox for testing and scenario analysis.