Misspecifying GARCH-M processes

We consider the relationships between ARCH-type and Stochastic Volatility models. A new class of volatility models, called Generalized Bilinear Stochastic Volatility, is described following an approach that transforms an initial GARCH-M process. The focus here is on the interpretation of some simulation results, with a special care devoted to model misspecification. Key Words: ARCH-type and Stochastic Volatility models; Bilinear Processes; Generalized Bilinear Stochastic Volatility; Misspecification Currently a Postdoc Visiting Scholar at Stanford University, PDP Research Lab. 1 Introduction The strong impulse that motivated the field of financial time series volatility models 1 , after the seminal work on Autoregressive Conditional Heteroskedastic (ARCH) models [11], materialized soon in a wide spectrum of methodological proposals and applications, making of this field an highly desired ground for testing statistical inferential techniques. Nevertheless, some of the aspects re..