international econometric journal
in Russian language
Many time series models, primarily various models with unobservable components, can be represented in a so called state space form. A state space model is a powerful tool that allows one to apply to the original model a wide range of standard procedures including estimation and forecasting. This essay provides a survey of this universal class of models and related procedures.
We consider goodness-of-fit tests based on the empirical process theory. There are two main obstacles in obtaining critical values for such tests: the parameter estimation effect and distribution dependence. We discuss solutions to these problems: martingale transformation and bootstrap. As an illustration we show how to test GARCH and diffusion models.
This article presents a survey of the developments of univariate GARCH models. ARCH, GARCH, EGARCH and other possible nonlinear extensions are examined. Conditions for stationarity (weak and strong) are presented. Inference and testing is presented in the quasi-maximum likelihood framework. Continuous GARCH approximations are discussed.
This article studies modeling dependence between futures and spot prices of financial indices and verifies a practical value of econometric models for futures hedging using Russian and foreign data. The dynamics of futures and spot prices is described by an error correction model, while volatilities and correlations are modeled by various multivariate GARCH models with dynamic conditional correlations of different degree of detail. The empirical investigation carried out in the article can answer questions on effectiveness of hedging strategies based on multivariate GARCH models, on similarities and differences of dependencies between futures and basic assets in Russian and foreign financial markets, and on a reasonable degree of detail in multivariate GARCH modeling.