Time series analysis in the economic sciences (Q5896953)

The textbook ``Zeitreihenanalyse in den Wirtschaftswissenschaften'' (time series analysis in economic sciences) has been published in the second edition in 2009 and it is written in German. Compared with the first edition the author has added an additional chapter concerned with state space models and the Kalman filter. The book is divided into two parts. The first part deals with univariate time series models, and the second one with multivariate time series models and analyses. Each part is self-contained and can be read separately. As the author mentions, the readers should have some background knowledge in econometrics as well as economics, and it is especially written for students applying for a Bachelor or a Master degree. As is usual in applied econometrics and time series analysis, some software programs are needed to solve the exercises and to replicate the examples given in the book. The author recommends the use of EVIEWS and/or MATLAB. The datasets used in the various examples as well as some MATLAB routines can be downloaded from the author's website \url{http://www.neusser.ch}. The book also contains some exercises and for a selected set of exercises the solutions. The first part of the book -- univariate time series analysis -- introduces the class of ARMA models, estimation procedures for these models, integrated processes augmented with the unit root tests and estimation procedures. Furthermore, a chapter is devoted to forecasting of stationary time series, another one to the estimation and interpretation of the autocorrelation and the partial autocorrelation functions. The part on univariate time series models closes with the presentation of volatility models such as the ARCH(1)-model or the GARCH\((p,q)\)-model. In the second part of the book deals with multivariate time series analysis in the time domain. Here the main topics covered here are the definition of stationary time series models, the estimation of vector-autoregressive models (VAR-models) as well as their interpretation by use of impulse-response functions and the forecast error variance decomposition. A separate chapter is solely devoted to the concept of cointegration. The author introduces the concept of cointegration and provides insights into the Beveridge-Nelson-decomposition of multivariate time series as well as insights into the derivation of a ``common trend'' representation of the time series. The final chapter of the book deals with Kalman-filtering. After the introduction of the basic parts of state space models, an example is provided using quarterly GDP data.