Convertible bond pricing model based on copula approach

Pricing issues of convertible bond related to a variety of complex factors, have characteristics of uncertainty, nonlinear and exotic. Current pricing models usually adopt Pearson correlation coefficient to capture the correlation structure between the factors which may result in model specification errors when dealing with nonlinear situations. This paper applies the method of copula which is more sensitive to nonlinear correlation relationships to value convertible bond. We establish a framework of multivariate pricing model based on copula theory, and use a two-factor case in which takes account of stock price and exchange rate to illustrate procedures of deriving numerical solution. Four different copula functions (Gumbel, Frank, Clayton and Normal copulas) are used to measure the correlation structure between the factors respectively. Then, we illustrate the impact of different copula functions on the results of pricing through a numerical example.