Roberto Castro - Academia.edu (original) (raw)
Uploads
Papers by Roberto Castro
This is a one semester version of what was originally a two-course sequence in time series econom... more This is a one semester version of what was originally a two-course sequence in time series econometrics that included Econ 557b. The course provides an introduction to time series methods in econometrics covering stationary series, aspects of trend behavior, detrending mechanisms and their properties, unit root theory, cointegrated system approaches, realized volatility and quarticity, Wold and BN decompositions, model selection, nonlinear nonstationary models and methods, spatial density asymptotics, and long memory modeling. Both time domain and frequency domain methods are discussed, and Bayesian as well as classical approaches are included. The treatment relies on asymptotic theory for linear processes, martingales and martingale approximations. We overview a large literature and not all topics are treated in the same depth. Theory, computations and some empirical applications are discussed. Most classes are divided into two parts, one dealing with theory and the other with empirics.
This is a one semester version of what was originally a two-course sequence in time series econom... more This is a one semester version of what was originally a two-course sequence in time series econometrics that included Econ 557b. The course provides an introduction to time series methods in econometrics covering stationary series, aspects of trend behavior, detrending mechanisms and their properties, unit root theory, cointegrated system approaches, realized volatility and quarticity, Wold and BN decompositions, model selection, nonlinear nonstationary models and methods, spatial density asymptotics, and long memory modeling. Both time domain and frequency domain methods are discussed, and Bayesian as well as classical approaches are included. The treatment relies on asymptotic theory for linear processes, martingales and martingale approximations. We overview a large literature and not all topics are treated in the same depth. Theory, computations and some empirical applications are discussed. Most classes are divided into two parts, one dealing with theory and the other with empirics.