Chapter 9. Stochastic volatility model through MCEM: Departure from canonical SV model

Abstract

We analyze a discretized canonical stochastic volatility model through calibration to synthetic data as well as financial data. To achieve this end we resort to Monte Carlo EM (Chan and Ledolter (1995)) that is, the combination of EM algorithm (Dempster et al. (1977), Wu (1983)) and sequential Monte Carlo methods (Gordon et al. (1993), Doucet et al. (2001), Salmond and Smith (1993)). Finally we consider a slight departure from canonical stochastic volatility model in order to assess the robustness of the MCEM procedure. Simulations studies follow.