Leveraging statistical and economic approaches for hybrid identification of large SVARs

• Applicant: Dr. Jan Prüser
• Subject Area: Statistics and Econometrics
• Term: since 2025
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 561170677

The project consists of three main research goals. The first one is to propose new combinations of established identification approaches. By blending different identification approaches in SVAR models, I leverage the strengths and mitigate the weaknesses of individual methods, leading to more robust and flexible models. This holistic approach enhances the credibility and reliability of the findings derived from SVAR analyses. The second one is to perform structural analysis with many variables. Incorporating a large set of variables in SVAR models offers significant benefits for structural analysis. By leveraging a rich information set, it is possible to improve the identification of structural shocks by enhancing model robustness and reducing the risk of an omitted variable bias. As a result, policymakers can make more informed and effective decisions. Advanced techniques like dynamic factor SVARs and Bayesian SVARs help manage large datasets, ensuring reliable insights. The third one is to conduct structural analysis for Quantile SVARs. Considering Quantile SVARs is useful because they provide a more detailed understanding of the effects of structural shocks across different points of the distribution of a variable. Unlike traditional SVAR models that focus on average effects, Structural Quantile VARs capture the heterogeneity in responses, revealing how shocks affect different quantiles (e.g., median, upper, and lower tails). This approach can uncover asymmetric effects, tail risks, and varying responses under different economic conditions. It enhances the analysis by providing a richer and more comprehensive picture of the underlying dynamics, informing more nuanced policy decisions and improving risk management strategies.

• DFG Programme: Research Grants