This paper considers a Student-t stochastic volatility (SV) model
using full Bayesian approach. A new two-stage scale mixtures representation for
the Student-t density is proposed and used to speed up the efficiency of the
Markov chain Monte Carlo sampling schemes.
