Blockseminar Recent developments in forecasting methods with applications to financial market interbank credit market and GDP

Recent developments in forecasting methods state that a better forecast can be achieved rather through the combination of different models than by the use of one individual model. Many theoretical and empirical studies have indeed shown that certain application of different combinations so-called -model-averaging” of individual models can dominate individual forecasts in many different areas of economics and finance. In one such model-averaging, each individual forecasting model is represented in the final forecast combination with certain weights which have to be estimated. Consequently, numerous methods have been proposed for the estimation of optimal weights in model-averaging that can be constant or varying with time. The goal of this course is to introduce students to the theory of optimal forecast combination and to allow students to become familiar with it. After that, a statistics-software (such as Matlab, R, GAUSS, STATA or EViews) will be introduced through a practical application of model-averaging so that students can carry out an empirical project on the basis of certain empirical data. In the empirical part of the course, the focus is placed on the forecasting variables from the real economy – that is mainly the GDP, financial market data (both monthly and daily data, as well as so-called intraday data) and interbank credit-market. In addition, the subprime crisis has shown that there are certain not-well-working mechanisms between the real economy (GDP) and financial market which both may also be connected over interbank credit-market. Therefore a special focus in the context of optimal forecasts will be placed on the analysis and forecasts of the effects of mechanisms that connect real economy and financial market via interbank credit-market. Students have to choose a topic for the application of model-averaging and, most at the end of the first part of the course present their research topic. In this first presentation, students need to explain which is the variable of interest as well as the methods for the optimal forecast combination for this variable. Moreover, it should be clearly presented and show how this survey would be conducted and how to apply a certain software for these purposes. Last, students shell share their thoughts with others in a sense what may show concrete benefits for theoretical or empirical research to be expected from their empirical project. The end of the semester is then followed by the second part of the course to the extent of two days with the second part of the presentation in which finale results and outlook will be presented. Until the first of August, the final seminar work should be finished and submitted. FB 07 Wirtschaftswissenschaften Uni Kassel SoSe 2013 Lehrveranstaltungspool FB 07 Dr. Jeleskovic Vahidin