Seminar: Structural Change and Expectations in Macroeconomics and Finance (F)

The seminar studies how structural change—interpreted as parameter instability, breaks, and regime shifts—affects empirical work in macroeconomics and finance. A central workhorse approach will be forecast evaluation via forecast-error regressions (e.g., Mincer-Zarnowitz style), extended with simple and transparent tools to detect instability (split samples, rolling windows, multiple-break tests such as Bai-Perron, and Markov regime-switching). Applications include professional macro forecasts, inflation dynamics, and asset-price predictability.
Students will write an independent seminar paper within the overall theme. Typical projects fall into one of three (overlapping) tracks:

1.
Expectations and forecast evaluation: test properties of expectations using survey-based forecasts and the corresponding forecast errors (e.g. from the Survey of Professional Forecasters), including whether forecast biases or performance are stable over time.

2.
Structural change in key empirical relationships: document and interpret instability in core macro/finance relationships, such as the New Keynesian Phillips curve (inflation dynamics) or asset-price predictability (predictive regressions), using breaks/instability tools.

3.
Forecasting methods under structural change: study or implement forecasting approaches designed to perform well when the data-generating process changes over time—for example by allowing time variation, emphasizing robustness, or using model updating schemes—and evaluate their performance across subsamples or regimes.

The seminar is designed so that empirical projects can be completed with modest programming and standard econometric tools, while allowing more advanced students to pursue richer time-variation or regime-switching approaches.