A data and digital-contracts driven method for pricing complex derivatives

Abstract

In this paper, we propose a data and digital-contracts driven (DDCD) method for pricing various complex options. The DDCD method is a combination of nonparametric and parametric methods. In general, nonparametric data driven methods use observed data as training data of a learning network directly. Different from these, in the proposed DDCD method, some European-style digital contracts (DCs) of the underlying assets are added as auxiliary information to guide the learning process of the pricing formula. The DCs can be obtained by using the observed data according to parametric methods. Thus, the DCs are actually used as the hints of the pricing formula, and then the DDCD method has superior pricing accuracy to the common data driven method in practical applications. Some Monte Carlo simulation experiments are performed and the results demonstrate that the proposed method not only has the advantages of generalization and superior accuracy, as the non-parametric method has, but also has the property of robustness to financial data with noise, as the parametric method has.