Variance Swap Premium under Stochastic Volatility and Self-Exciting Jumps

Ke Chen, Serhuang Poon

Research output: Working paper

Abstract

Abstract Based on the MCMC samples drawn from SV-Jump model, we found evidence that jumps are clustering even if we allow the variance process to be time varying. We introduced self-exciting jump intensity process to capture excited jump intensity after a realized jump. On the other hand, we are able to use the proposed model to price options and variance swap based on the characteristic function. By comparing the MCMC estimation under physical measure and option calibration under risk neutral measure, we are able to distinguish the risk premium from stochastic volatility and jump intensity. The results show that jump intensity is always overpriced but its premium decays fast, which explains the high risk premium after unexpected big movements in the market; however, stochastic volatility is much more persistent but sometime is underpriced especially in high volatility periods which explains the big loss by shorting variance swap in subprime crisis.
Original languageEnglish
Number of pages36
Publication statusPublished - May 2011

Keywords

  • Markov chain monte carlo, stochastic voaltility, self-exciting jumps, variance swap, risk premium

Fingerprint

Dive into the research topics of 'Variance Swap Premium under Stochastic Volatility and Self-Exciting Jumps'. Together they form a unique fingerprint.

Cite this