Abstract—Financial market models, like scenarios sequences analysis, deals with exponentially growing data volumes, e.g., stock prices and interest rates. Methods to solve these problems usually require simulations to cover a time span for multiple future periods. Trinomial tree is a technique that is based on the projection of different values to the price a share could achieve during its lifetime. This work aims to demonstrate the benefits of using Graphics Processing Units (GPU) to implement a software prototype for trinomial tree technique to price options. For this purpose, we compare the performance between GPU and Central Processing Unit (CPU) implementation’s methods.

Index Terms—trinomial trees, pricing options, parallel computing, GPU, financial market

Cite: Marcelo L. Rocha and David N. Prata, "Improving Performance for Pricing Options by Simulating Trinomial Trees," Journal of Advanced Management Science, Vol. 4, No. 3, pp. 260-264, May 2016. doi: 10.12720/joams.4.3.260-264