Time Series Analysis
- Analyze economic data in accordance with the task, make preliminary data analysis.
- Build appropriate econometric time series models for the research question, analyze and interpret results.
- Understand limitation and relevance of the models.
- Know basic concepts of univariate time series analysis, build appropriate econometric time series models.
- Know basic concepts of multivariate time series analysis, build appropriate econometric time series models.
- Univariate time series analysis1. Stationary Time Series. Dealing with time series data. Concept of stationary and covariate-stationary series, autocorrelation and partial-autocorrelation functions, white Noise, autoregression models (AR), moving average (MA) models, ARMA models: properties, specification, estimation and forecasting. the Box–Jenkins methodology, diagnostic testing for model adequacy. Topic 2. Nonstationary Time Series. Problems arise due to nonstationary, unit roots and characteristic relations, testing nonstationary: the Dickey–Fuller and augmented Dickey–Fuller tests, KPSS test and Philip-Pearson test; series transformation: differencing, selecting order of difference and ARIMA model. Dealing with seasonal data: series decomposition into stationary and trend and(or) seasonal component, Fourier decomposition and periodogram, SARIMA models. ARCH-GARCH model to deal with nonstationarity due to non-constant dispersion. Forecasting and forecasting errors.
- Mutivariate time series analysis3. Vector Autoregression (VAR). Reduced and structural VAR Forms, model estimation, model conditions, vector AR(p) models, vector moving average models, lag specific criteria: LM test, Granger causality test, exogeneity in a VAR, the impulse-response function, forecasting with VAR: dynamic, static, stochastic and deterministic solutions. 4. Structural Vector Autoregression (SVAR). SVAR specification, comparison with reduced form VAR, structural impulse responses, Choleski decomposition, Blanchard-Quah decomposition, variance decomposition, identification strategies: recursive and non-recursive. 5. Vector Error Correction Model (VECM). The concept of cointegration and LR relations, the Engle–Granger cointegration test, the Johansen full-information maximum likelihood cointegration test, VECM specifications and estimation, lag length and causality tests, forecasting with VECM.
- Interim assessment (2 module)0.4 * Exam + 0.3 * Reports + 0.1 * Self-study work (DC) + 0.1 * Test 1 + 0.1 * Test 2
- Klaus Neusser. (2016). Time Series Econometrics. Springer. Retrieved from http://search.ebscohost.com/login.aspx?direct=true&site=eds-live&db=edsrep&AN=edsrep.b.spr.sptbec.978.3.319.32862.1
- Palma, W. (2016). Time Series Analysis. Hoboken: Wiley. Retrieved from http://search.ebscohost.com/login.aspx?direct=true&site=eds-live&db=edsebk&AN=1229817
- Bleikh, H. Y., & Young, W. (2013). Time Series Analysis and Adjustment : Measuring, Modelling and Forecasting for Business and Economics. Farnham: Routledge. Retrieved from http://search.ebscohost.com/login.aspx?direct=true&site=eds-live&db=edsebk&AN=531761
- Montgomery, D. C., Jennings, C. L., & Kulahci, M. (2015). Introduction to Time Series Analysis and Forecasting (Vol. Second edition). Hoboken, New Jersey: Wiley-Interscience. Retrieved from http://search.ebscohost.com/login.aspx?direct=true&site=eds-live&db=edsebk&AN=985114