Overarch

Risknowledge flagship tool, Overarch offers an innovative suite of risk analytics for model validation, model rectification, model selection and estimation of actual risk.

It provides a SaaS-based, flexible, rigorous, turnkey implementation of ES Ridge Backtest methodologies, allowing the backtest of Expected Shortfall, Value at Risk, and other backtestable risk measures, as well as prediction discrepancy measures for estimation of true risk, which is possible for Expected Shortfall thanks to the unique properties of the Ridge Backtest.

Overarch is compatible with any distributional assumption of a client risk model (parametric, historical simulation, MonteCarlo, etc.) and is completely agnostic with respect to the unknowable real-world distribution of events. In this sense, it is a completely model-independent technology, representing the best known model validation framework for distribution-based risk models today.

Model performance scorings, based on joint elicitability of VaR and ES, also allow for innovative model selection between multiple challenger models on the same portfolios.

Proprietary algorithms based on closed-form solutions, permit unparalleled computational speed, full precision and scalability across large portfolio trees.

Our approach to software development upholds the highest industry standards in performance, security, scalability, and efficiency. Every aspect of our platform is meticulously engineered to ensure that our clients receive thrustworthy, responsive, and resilient tools that meet today’s most demanding risk environments.