312508 Zeitreihenanalyse II (V) (WiSe 2016/2017)

The lecture contains three main parts:
1. Stationary processes: These build the basis for all advanced methods in time series analysis. We will discuss the ARMA framework in depth both in the time and in the frequency domain. This will for example clarify the name "white noise". The lecture will also deal with multivariate processes.

2. Unit roots: A common deviation from stationarity in time series lies in random walk like behaviour. In many cases a time series behaves non-stationary while its first difference appears to be stationary. We will discuss how to detect such deviations from stationarity and how to handle them.

3. Panel data models: Both micro and macroeconomic data increasingly is available in the form of panels rather than cross-sections or time-series alone. In this part we will incorporate our findings from the first two parts into the panel setting.

Requirements for participation, required level

Prerequisites for the lecture are basic understanding of time series concepts such as trend, seasonal components and autoregressive processes on the level of the preliminary time series course. Alternatively the corresponding chapters in Wooldridge "Introduction to Econometrics" suffices.