ECSight — ECSight Technology


  • Print
  • Connect
  • Email
  • Facebook
  • Twitter
  • LinkedIn
  • Google+

ECSight™ combines high-performance asset and liability modeling engines with scalable and efficient distributed computing technology to achieve the computing power needed to arrive at granular economic capital metrics over billions of valuation model points.

An enterprise-scale relational database facilitates the storage, analysis, and reporting of valuation results in sufficient detail to answer questions about how portfolio composition, market conditions, and management actions drive capital requirements and shape risk profile.

ECSight also satisfies IT, business, and regulatory demands for reliability, security, and auditability. ECSight supports full versioning of all inputs and outputs, generation of audit trails, fault tolerance through the automatic failover of compute nodes, a browser-based interface complete with industry-standard authentication and role-based logical access controls, and output that can be viewed and analysed in standard formats, including HTML, PDF, and CSV.

Data valuation and risk taxonomy

Data required for valuation are uploaded to the database via text files with simple, pre-defined formats and/or via custom ETL processes. ECSight is also able to integrate market data from a variety of external data vendors such as Bloomberg, SuperDerivatives, and MarkIT (subject to licensing arrangements).

The ECSight data model captures the relationships needed to map imported market data to the universe of risk drivers through a comprehensive risk taxonomy. The Risk Taxonomy establishes a hierarchical relationship between tens of thousands of risk factors, providing a data-driven mechanism for uniformly quantifying sensitivities to both market and actuarial risks. Because scenario definitions, asset and liability valuation models, and reporting functions are expressed in terms of this taxonomy, ECSight ensures consistency and compatibility across all components of the system.

Performance

A compiled C++ asset valuation library with support for a wide range of asset classes facilitates seriatim asset valuation, while proxy (or “lite”) modeling techniques can be employed to value complex liabilities over multivariate risk scenarios. Data throughput bottlenecks are addressed with a Microsoft SQL Server database, performance-tuned to manage the distribution of valuation inputs and collection of valuation results from multiple compute nodes running concurrently.

Scalability

ECSight is capable of leveraging multiple high-performance computing environments through Milliman GridStep technology, including the Milliman C-Squared or Microsoft HPC Server grid computing platforms for in-house processing, and Milliman GridStep Cloud Edition for on-demand access to virtually unlimited computing resources in the cloud. Integration with the cloud is vital to leveraging ECSight scalability beyond capacity constraints typically encountered when deploying to conventional data centres.

With the ability to run streamlined asset and liability valuations in parallel on many thousands of compute nodes in Microsoft Azure, ECSight can deliver analyses on the distribution of economic capital over hundreds of thousands of multivariate risk factor scenarios over hundreds of forward-projection or back-test timesteps. The ability to provision resources on-demand leads to an efficient use of computing resources and lower operating costs while avoiding the compromise on accuracy and granularity imposed by the technology constraints of less scalable solutions.

Flexibility

With a highly modular architecture, ECSight can be configured to utilize multiple asset and liability models. In addition to Radial Basis Functions (RBF) for lite modeling of liabilities, other lite model implementations can also be deployed. The use of highly decoupled interfaces between valuation models and the core system means accuracy, performance, change control, and cost considerations can be addressed independently without constraints imposed by system architecture.

Additionally, with integrated support for multiple distributed computing platforms, ECSight has the flexibility to utilize internal or cloud-based computing resources in whatever combination offers the optimal balance of performance, cost, and security.

Contact us to learn how to get started with ECSight.

Next Steps

For more information, contact: