A new risk computation framework has been introduced based upon an ultrafast dependency graph and a flexible scenario generator.
Multiple use cases are easily supported - from simple bucketed sensitivities for an individual trade, to stress testing and historical VaR on large multi-currency portfolios.
The dependency graph eliminates redundant computations - both between market ticks and between risk scenarios. Intermediate results are cached, and mathematical simplifications are employed to further accelerate curve fitting.
As well as being fast to compute, the framework is easy to implement - something which is particularly beneficial for our new clients as it allows them to be up and running faster, with seamless integration into their systems.
Dr. Gene Schupak, Suite Partner and Head of Quantitative Development emphasizes the precision behind this development, "Getting this feature right required a lot of detailed engineering work. The framework can be applied to any number of different functional requirements."
Elsewhere, our focus has been on increasing simplicity in ALib. Several new higher level functions have been added to accelerate workflows, helping users get things done in fewer steps.
Finally, the ALibPY Python interface has been enhanced to make it easier to build industrial scale applications. Memory management and garbage collection have been upgraded. Performance and error logging have been further improved.
James Baker, Product Manager adds, "we are fortunate that our customers include some very sophisticated and experienced market participants. This helps expose us to a wide variety of use cases and technical demands. The result with the risk framework is that we've been able to deliver something that is flexible and easy to use - and is immediately getting positive feedback from our users."
For more information, please visit our website at suitellc.com.