Fast Analytics Made Easy
Home / Resources / Sensitivities
The paper gives an overview of the proposed FRTB SA-CVA framework, discusses the approaches for computing CVA sensitivities, and shows how AAD can be applied
This paper introduces the use of sensitivities for managing CVA / XVA, details the complexities involved, and introduces AAD as a solution
The paper exposes the benefits of using AAD for financial model calibration using practical examples, and shows the costs-benefits trade-off of AAD vs. bumping
This white paper describes major algorithmic optimisations for XVA and illustrates the possible computational savings.
This paper explains the FRTB IMA P&L attribution test, highlights the practical challenges involved, and gives recommendations on how to overcome these.
This white paper walks through IDE integration, utilities, debugging, and profiling with the Xcelerit SDK.
This white paper gives an overview of the Xcelerit SDK's programming interface.
This paper zooms on Adjoint Algorithmic Differentiation (AAD) as an efficient and robust alternative to compute sensitivities vs. bump-and-revalue
This paper introduces the SIMM method, illustrates how AAD can be applied for computing the sensitivities, and explains dynamic SIMM for MVA simulations.
This paper reviews the complexities in credit exposure calculations, FRTB SBA, FRTB IMA, FRTB SA-CVA, SIMM, and XVA calculations
This paper introduces FRTB SBA, discusses its definition of the delta and vega sensitivities, and illustrates how AAD can be used to compute them.
This video gives an introduction into algorithmic differentiation: theory, manual implementation, live example
The two main approaches for automatic AD are covered in this video: source code transformation and operator overloading
This video gives practical guidelines on how to cope with the complexities involved when integrating AD into large quant libraries
This video demonstrates how to achieve high performance gains on multicore CPUs and GPUs for CVA calculations using the Xcelerit SDK
A tier 1 bank implemented a centralised XVA system with near real-time calculation capabilities, including sensitivities.
The tier one bank implemented fast CVA sensitivity calculation using Algorithmic Differentiation (AAD) with Xcelerit
The client computes MVA using ISDA's SIMM, implemented using Algorithmic Differentiation (AAD)
A set of popular quant finance applications designed to compare the computational performance of processors and software implementation approaches.
Impact of FRTB, ISDA SIMM, XVA regulations on computational complexity
This publication gives an overview of various computing processors: CPU, GPU, Xeon, Phi, etc. and shows how to choose between them