This series of lectures considers econometrics issues that arise in the process of forecasting. We will see that several basic econometric notions need to be reconsidered when forecasting. An important starting point when forecasting is the underlying motivation for the forecast, which can be summarised in the loss function. We then move to multistep forecasting, forecasting in changing environments, and forecast averaging. Going beyond univariate time series, we also consider forecasting with panel data and forecasting with large data sets.
Lecture 1+2: October 12, 2022 1:00 pm to 4:30 pm
Lecture 3+4: October 13, 2022 9:30 am to 12.45pm
An office hour with Andreas Pick will be offered on October 13 from 2.15 to 3.15pm
Please book your personal slot for the office hour here.