BREAKMETRICS · Forecasting Economic Time Series in Changing Environments

The general objective of this project is to improve economic forecasts with evolutionary econometric models, i.e. models that adapt to structural changes in the economic environment. They are also called change-point models. The need for these adaptive models is obvious in the light of the current economic downturn.Economic forecasting is essential for decision making with respect to fiscal and monetary policy, public spending, and investment. The three research objectives are theoretical and empirical. First, we will study and develop change-point models for time series. In fact, it is crucial to understand why and which change-point models seem to forecast well. Second, we will develop efficient Bayesian inference methods and software. Finally, we will provide a large variety of forecasting applications using European macroeconomic and financial data. To transfer knowledge to a forecasters, software and explanations to implement the models will be made available via a website.The contributions of this proposal will be substantial for academic researchers, and users of forecasting models in the public sector (for example the EU Commission, the European Central Bank, and national institutions) and the private sector. A better understanding of the economic and financial system together with more reliable forecasts of its variables should help the conduct of finer public policies and the design of more efficient and reliable financial markets. This is especially important in view of the changes likely to be triggered by the current economic depression. The output of this proposal will be reported in the best academic journals of the field.