Defaultable bond pricing under the jump diffusion model with copula dependence structure

Siti Norafidah Mohd Ramli*, Jiwook Jang

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

1 Citation (Scopus)

Abstract

We study the pricing of a defaultable bond under various dependence structure captured by copulas. For that purpose, we use a bivariate jump-diffusion process to represent a bond issuer’s default intensity and the market short rate of interest. We assume that each jump of both variables occur simultaneously, and that their sizes are dependent. For these simultaneous jumps and their sizes, a homogeneous Poisson process and three copulas, which are a Farlie-Gumbel-Morgenstern copula, a Gaussian copula, and a Student t-copula are used, respectively. We use the joint Laplace transform of the integrated risk processes to obtain the expression of the defaultable bond price with copula-dependent jump sizes. Assuming exponential marginal distributions, we compute the zero coupon defaultable bond prices and their yields using the three copulas to illustrate the bond. We found that the bond price values are the lowest under the Student-t copula, suggesting that a dependence structure under the Student-t copula could be a suitable candidate to depict a riskier environment. Additionally, the hypothetical term structure of interest rates under the risky environment are also upward sloping, albeit with yields greater than 100%, reflecting a higher compensation required by investors to lend funds for a longer period when the financial market is volatile.
Original languageEnglish
Pages (from-to)941-952
Number of pages12
JournalSains Malaysiana
Volume49
Issue number4
DOIs
Publication statusPublished - Apr 2020

Keywords

  • Bivariate jump-diffusion model
  • credit risk
  • default intensity
  • short rate
  • zero coupon bond

Fingerprint

Dive into the research topics of 'Defaultable bond pricing under the jump diffusion model with copula dependence structure'. Together they form a unique fingerprint.

Cite this