CMStatistics 2022: Start Registration
View Submission - CFE
Title: Parallel computing and software integration applied to financial models calibration Authors:  Antonio Santos - University of Coimbra (Portugal) [presenting]
Abstract: Option pricing financial models usually belong to the set of testing problems associated with cutting-edge hardware and software developments happening at a fast pace, namely the ones related to parallel computing. We demonstrate how parallel computing, possible through Graphical Processing Units, integrated with state-of-the-art software that implements numerical optimization, permits real-time calibration of option pricing models within a big data setting. We develop software that can calculate options prices in parallel, allowing a vast amount to be calculated simultaneously, reducing by a factor of several hundred the computational time needed to compute a fit function. Furthermore, the computation time reduction makes it viable that software packages full-tested in implementing standard optimization algorithms can be used to solve the calibration problem.