Int. J. of Data Science   »   2016 Vol.1, No.4

 

 

Title: Scalable system with accelerators for financial option prices estimation

 

Authors: D. Dimitrov; E. Atanassov

 

Addresses:
Institute of Information and Communication Technologies, Acad. G. Bonchev str. 25A, Sofia 1113, Bulgaria
Institute of Information and Communication Technologies, Acad. G. Bonchev str. 25A, Sofia 1113, Bulgaria

 

Abstract: In this paper we describe a production-ready enterprise service bus (ESB) system for estimation of option prices using stochastic volatility models. We present the motivation for our research, the main building blocks for our system and discuss our approach for calibration of the Heston model. The Heston model is used as a basis for modelling the evolution of asset prices. The Zato framework is used as an integration layer, while the main computations are distributed to HPC resources (GPUs and Intel Xeon Phi cards). The system can use various data sources and scales from both infrastructure and software point of view. The main advantage of the system is that by incorporating general-purpose computing on graphics processing units (GPGPU) and Intel Xeon Phi nodes it allows for the use of more accurate models that are otherwise unfeasible. This system can be useful for distributed processing of a large volume of option pricing tasks.

 

Keywords: option pricing; Heston model; GPGPU; general-purpose computing; graphics processing units; GPUs; Intel Xeon Phi; accelerators; scalable systems; option prices; price estimation; financial options; enterprise service bus; stochastic volatility models; modelling; asset prices.

 

DOI: 10.1504/IJDS.2016.081367

 

Int. J. of Data Science, 2016 Vol.1, No.4, pp.305 - 315

 

Available online: 05 Jan 2017

 

 

Editors Full text accessAccess for SubscribersPurchase this articleComment on this article