Abstract We provide a general framework for using jump dependence structure, in addition to mean-variance, in portfolio selection. Weekly MSCI index returns of eleven developed and emerging markets are estimated using Markov Chain Monte Carlo (MCMC) method. We reach the same conclusion as Das and Uppal (2004), that for developed markets the omission of systemic co-jumps has little impact on portfolio selection. However, we find on the contrary that, for emerging markets, the loss in certainty equivalence derived using the wrong jump dependence assumptions is economically significant.
|Number of pages||40|
|Publication status||Published - 2013|
- Systemic, Idiosyncratic, Jump-diffusion, international portfolio diversification, Markov Chain Monte Carlo