|
Lecturer(s)
|
-
Zapletal David, doc. Mgr. Ph.D.
|
|
Course content
|
Stochastic proces and its stationarity. Stationarity models of time series. AR, MA, ARMA models (properties, identifying, goodness of fit) Nonstationary models, unit-root tests. Seasonal models Model diagnostics and forcasting by estimated models. Linear and nonlinear volatility models. Model diagnostics and forcasting by estimated volatility models. Linear models of multivariate time series. Causality in time series. Linear models of multivariate non-stationary time series - EC model and cointegration of time series.
|
|
Learning activities and teaching methods
|
Monologic (reading, lecture, briefing), Dialogic (discussion, interview, brainstorming), Work with text (with textbook, with book), Methods of individual activities, Skills training
- Participation in classes
- 52 hours per semester
- Home preparation for classes
- 26 hours per semester
- Preparation for a credit (assessment)
- 32 hours per semester
- Preparation for an exam
- 40 hours per semester
|
|
Learning outcomes
|
The aim of the course is to provide students with an overview of methods of stochastic modeling of time series based on Box-Jenkins methodology, which find application in banking, insurance and dynamic modeling of macroeconomic indicators, with an emphasis on the application of these methods using appropriate statistical software.
A student who has successfully completed the course can: apply methods of stochastic modeling based on theBox-Jenkins methodology used for both level modeling and volatility modeling of univariate time series; apply linear models of multivariate stationary time series and understand concepts related to causality and exogenity of economic time series; apply linear models of multivariate non-stationary time series and understand the concepts related to the problem of co-integration of time series. A student who has successfully completed the course can: assess the stationarity and seasonality of the time series and transform the non-stationary time series into stationary; use statistical software to identify a suitable time series model and estimate its parameters; the diagnostic control means verify that the model for the time series is acceptable or requires further extensions, for example, conditional heteroscedasticity; to use linear and nonlinear models of volatility to model conditional heteroscedasticity of time series; identify, estimate and diagnose a multi-equation model of stationary time series and test the nature of relationships between them, eg. Granger causality; using unit root tests to assess the non-stationarity of time series and then examine and test possible cointegration relationships of the studied time series; use time series models to construct point or interval predictions. The student who has successfully completed the course is able to: decide which methods and models should be used to analyze given economic time series; to justify and defend the choice of this method in a clear and convincing way.
|
|
Prerequisites
|
unspecified
|
|
Assessment methods and criteria
|
Oral examination, Written examination, Student performance assessment
Full-time teaching Assignment - Active participation in seminars. Elaboration of assigned tasks. Examination - Passing written test with evaluation at least 60% Distance online teaching Assignment - Elaboration of assigned tasks. Examination - written + oral.
|
|
Recommended literature
|
-
Arlt, J., Arltová, M.:. Ekonomické časové řady. Praha: Grada Publishing, 2007.
-
Cipra, Tomáš. Finanční ekonometrie. Praha: Ekopress, 2013. ISBN 978-80-86929-93-4.
-
Tsay, Ruey S.. Analysis of financial time series. Hoboken: John Wiley & Sons, 2005. ISBN 0-471-69074-0.
|