Quantitative Trading Strategy
Our client, a small options trading firm, was developing a new quantitative trading strategy.
We made their advanced model a reality.
We integrated real-time options data from IVolatility's service, archived IVolatility options data, and real-time equities data from Bloomberg's service.
We worked closely with the model maker to demonstrate that our application accurately followed his model.
Bringing a strategy to life requires a delicate combination of robustness and sensitivity. Many types of data faults must be filtered or handled seamlessly (whether expected or signaling a violation the model's validity). Still others suggest an unexpected dynamic or boundary in the model itself.
We identified and resolved crucial internal stability issues in the model. Further, our application successfully caught real situations where vendors systematically shipped wrong data.
By making it multi-threaded, we achieved a significant performance increase over our client's expectations, harnessing four hyper-threaded processors. Other optimizations in the translation from Matlab to C++ gave further performance advantages.
The application was written in C++ for the Win32 environment.