Julia Buchegger
Home: Technische Universität Wien
Host: University of California, Irvine
Topic: Yule-Walker Estimator and Adaptive Learning in Macroeconomic Model
Julia Buchegger is a Master’s student in the “Statistics - Probability - Mathematics in Economics” program at TU Wien, currently on a research stay at the University of California, Irvine. Her thesis titled "Yule-Walker Estimator and Adaptive Learning in Macroeconomic Models" examines how different adaptive learning rules (constant-gain least squares and Yule-Walker learning) shape inflation and output dynamics in a New Keynesian setting.
Using Monte Carlo simulations, she compares the stability of expectations and the volatility of key macro variables under each learning mechanism, including sensitivity to shocks and parameters such as learning gains and price flexibility. The project highlights Yule-Walker learning as a stable-by-construction benchmark that avoids artificial projection facilities, helping to locate the true source of instability.
The project helps disentangle whether observed volatility stems primarily from the policy configuration itself or from the expectations algorithm agents use. The goal is to clarify when particular learning mechanisms promote stability and when alternative designs may perform better - evidence that can inform future monetary-policy calibration and communication.
Julia Buchegger